research methods for research paper

"For example, we did a large voluntary survey of email subscribers and top readers a few years back."

While these readers gave us a long list of topics, formats, or content types they wanted to see, they sometimes engaged more with content types they didn’t select or favor as much on the surveys when we ran follow-up ‘in the wild’ tests, like A/B testing.”  

Pepsi saw similar results when it ran its iconic field experiment, “The Pepsi Challenge” for the first time in 1975.

The beverage brand set up tables at malls, beaches, and other public locations and ran a blindfolded taste test. Shoppers were given two cups of soda, one containing Pepsi, the other Coca-Cola (Pepsi’s biggest competitor). They were then asked to taste both and report which they preferred.

People overwhelmingly preferred Pepsi, and the brand has repeated the experiment multiple times over the years to the same results.

What I like: It yields qualitative and quantitative data and can make for engaging marketing content, especially in the digital age.

What I dislike: It can be very time-consuming. And, if you’re not careful, there is a high risk for scientific error.

Best for: Product testing and competitive analysis

Pro tip:  " Don’t make critical business decisions off of just one data set," advises Pamela Bump. "Use the survey, competitive intelligence, external data, or even a focus group to give you one layer of ideas or a short-list for improvements or solutions to test. Then gather your own fresh data to test in an experiment or trial and better refine your data-backed strategy."

Secondary Research

8. public domain or third-party research.

While original data is always a plus, there are plenty of external resources you can access online and even at a library when you’re limited on time or resources.

Some reputable resources you can use include:

• Pew Research Center
• McKinley Global Institute
• Relevant Global or Government Organizations (i.e United Nations or NASA)

It’s also smart to turn to reputable organizations that are specific to your industry or field. For instance, if you’re a gardening or landscaping company, you may want to pull statistics from the Environmental Protection Agency (EPA).

If you’re a digital marketing agency, you could look to Google Research or HubSpot Research . (Hey, I know them!)

What I like: You can save time on gathering data and spend more time on analyzing. You can also rest assured the data is from a source you trust.

What I dislike: You may not find data specific to your needs.

Best for: Companies under a time or resource crunch, adding factual support to content

Pro tip: Fellow HubSpotter Iskiev suggests using third-party data to inspire your original research. “Sometimes, I use public third-party data for ideas and inspiration. Once I have written my survey and gotten all my ideas out, I read similar reports from other sources and usually end up with useful additions for my own research.”

9. Buy Research

If the data you need isn’t available publicly and you can’t do your own market research, you can also buy some. There are many reputable analytics companies that offer subscriptions to access their data. Statista is one of my favorites, but there’s also Euromonitor , Mintel , and BCC Research .

What I like: Same as public domain research

What I dislike: You may not find data specific to your needs. It also adds to your expenses.

Best for: Companies under a time or resource crunch or adding factual support to content

Which marketing research method should you use?

You’re not going to like my answer, but “it depends.” The best marketing research method for you will depend on your objective and data needs, but also your budget and timeline.

My advice? Aim for a mix of quantitative and qualitative data. If you can do your own original research, awesome. But if not, don’t beat yourself up. Lean into free or low-cost tools . You could do primary research for qualitative data, then tap public sources for quantitative data. Or perhaps the reverse is best for you.

Whatever your marketing research method mix, take the time to think it through and ensure you’re left with information that will truly help you achieve your goals.

Don't forget to share this post!

Related articles.

SWOT Analysis: How To Do One [With Template & Examples]

28 Tools & Resources for Conducting Market Research

What is a Competitive Analysis — and How Do You Conduct One?

Market Research: A How-To Guide and Template

TAM, SAM & SOM: What Do They Mean & How Do You Calculate Them?

How to Run a Competitor Analysis [Free Guide]

5 Challenges Marketers Face in Understanding Audiences [New Data + Market Researcher Tips]

Causal Research: The Complete Guide

Total Addressable Market (TAM): What It Is & How You Can Calculate It

What Is Market Share & How Do You Calculate It?

Free Guide & Templates to Help Your Market Research

Marketing software that helps you drive revenue, save time and resources, and measure and optimize your investments — all on one easy-to-use platform

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Archaeology in space: The Sampling Quadrangle Assemblages Research Experiment (SQuARE) on the International Space Station. Report 1: Squares 03 and 05

Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliations Department of Art, Chapman University, Orange, CA, United States of America, Space Engineering Research Center, University of Southern California, Marina del Rey, CA, United States of America

Roles Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

Affiliation Department of History, Carleton University, Ottawa, ON, United States of America

Roles Conceptualization, Data curation, Methodology, Project administration, Supervision, Writing – review & editing

Affiliation College of Humanities, Arts and Social Sciences, Flinders University, Adelaide, Australia

Roles Software, Writing – original draft

Roles Investigation, Writing – original draft

Affiliation Archaeology Research Center, University of Southern California, Los Angeles, CA, United States of America

• Justin St. P. Walsh, 
• Shawn Graham, 
• Alice C. Gorman, 
• Chantal Brousseau, 
• Salma Abdullah

• Published: August 7, 2024
• https://doi.org/10.1371/journal.pone.0304229
• Reader Comments

Between January and March 2022, crew aboard the International Space Station (ISS) performed the first archaeological fieldwork in space, the Sampling Quadrangle Assemblages Research Experiment (SQuARE). The experiment aimed to: (1) develop a new understanding of how humans adapt to life in an environmental context for which we are not evolutionarily adapted, using evidence from the observation of material culture; (2) identify disjunctions between planned and actual usage of facilities on a space station; (3) develop and test techniques that enable archaeological research at a distance; and (4) demonstrate the relevance of social science methods and perspectives for improving life in space. In this article, we describe our methodology, which involves a creative re-imagining of a long-standing sampling practice for the characterization of a site, the shovel test pit. The ISS crew marked out six sample locations (“squares”) around the ISS and documented them through daily photography over a 60-day period. Here we present the results from two of the six squares: an equipment maintenance area, and an area near exercise equipment and the latrine. Using the photographs and an innovative webtool, we identified 5,438 instances of items, labeling them by type and function. We then performed chronological analyses to determine how the documented areas were actually used. Our results show differences between intended and actual use, with storage the most common function of the maintenance area, and personal hygiene activities most common in an undesignated area near locations for exercise and waste.

Citation: Walsh JSP, Graham S, Gorman AC, Brousseau C, Abdullah S (2024) Archaeology in space: The Sampling Quadrangle Assemblages Research Experiment (SQuARE) on the International Space Station. Report 1: Squares 03 and 05. PLoS ONE 19(8): e0304229. https://doi.org/10.1371/journal.pone.0304229

Editor: Peter F. Biehl, University of California Santa Cruz, UNITED STATES OF AMERICA

Received: March 9, 2024; Accepted: May 7, 2024; Published: August 7, 2024

Copyright: © 2024 Walsh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: JW was the recipient of funding from Chapman University’s Office of Research and Sponsored Programs to support the activities of Axiom Space as implementation partner for the research presented in this article. There are no associated grant numbers for this financial support. Axiom Space served in the role of a contractor hired by Chapman University for the purpose of overseeing logistics relating to our research. In-kind support in the form of ISS crew time and access to the space station’s facilities, also awarded to JW from the ISS National Laboratory, resulted from an unsolicited proposal, and therefore there is no opportunity title or number associated with our work. No salary was received by any of the investigators as a result of the grant support. No additional external funding was received for this study.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The International Space Station Archaeological Project (ISSAP) aims to fill a gap in social science investigation into the human experience of long-duration spaceflight [ 1 – 3 ]. As the largest, most intensively inhabited space station to date, with over 270 visitors from 23 countries during more than 23 years of continuous habitation, the International Space Station (ISS) is the ideal example of a new kind of spacefaring community—“a microsociety in a miniworld” [ 4 ]. While it is possible to interview crew members about their experiences, the value of an approach focused on material culture is that it allows identification of longer-term patterns of behaviors and associations that interlocutors are unable or even unwilling to articulate. In this respect, we are inspired by previous examples of contemporary archaeology such as the Tucson Garbage Project and the Undocumented Migration Project [ 5 – 7 ]. We also follow previous discussions of material culture in space contexts that highlight the social and cultural features of space technology [ 8 , 9 ].

Our primary goal is to identify how humans adapt to life in a new environment for which our species has not evolved, one characterized by isolation, confinement, and especially microgravity. Microgravity introduces opportunities, such as the ability to move and work in 360 degrees, and to carry out experiments impossible in full Earth gravity, but also limitations, as unrestrained objects float away. The most routine activities carried out on Earth become the focus of intense planning and technological intervention in microgravity. By extension, our project also seeks to develop archaeological techniques that permit the study of other habitats in remote, extreme, or dangerous environments [ 10 , 11 ]. Since it is too costly and difficult to visit our archaeological site in person, we have to creatively re-imagine traditional archaeological methods to answer key questions. To date, our team has studied crew-created visual displays [ 12 , 13 ], meanings and processes associated with items returned to Earth [ 14 ], distribution of different population groups around the various modules [ 15 ], and the development of machine learning (ML) computational techniques to extract data about people and places, all from historic photographs of life on the ISS [ 16 ].

From January to March 2022, we developed a new dataset through the first archaeological work conducted off-Earth. We documented material culture in six locations around the ISS habitat, using daily photography taken by the crew which we then annotated and studied as evidence for changes in archaeological assemblages of material culture over time. This was the first time such data had been captured in a way that allowed statistical analysis. Here, we present the data and results from Squares 03 and 05, the first two sample locations to be completed.

Materials and methods

Square concept and planning.

Gorman proposed the concept behind the investigation, deriving it from one of the most traditional terrestrial archaeological techniques, the shovel test pit. This method is used to understand the overall characteristics of a site quickly through sampling. A site is mapped with a grid of one-meter squares. Some of the squares are selected for initial excavation to understand the likely spatial and chronological distribution of features across the entire site. In effect, the technique is a way to sample a known percentage of the entire site systematically. In the ISS application of this method, we documented a notional stratigraphy through daily photography, rather than excavation.

Historic photography is a key dataset for the International Space Station Archaeological Project. Tens of thousands of images have been made available to us, either through publication [ 17 ], or through an arrangement with the ISS Research Integration Office, which supplied previously unpublished images from the first eight years of the station’s habitation. These photographs are informative about the relationships between people, places, and objects over time in the ISS. However, they were taken randomly (from an archaeological perspective) and released only according to NASA’s priorities and rules. Most significantly, they were not made with the purpose of answering archaeological questions. By contrast, the photographs taken during the present investigation were systematic, representative of a defined proportion of the habitat’s area, and targeted towards capturing archaeology’s primary evidence: material culture. We were interested in how objects move around individual spaces and the station, what these movements revealed about crew adherence to terrestrial planning, and the creative use of material culture to make the laboratory-like interior of the ISS more habitable.

Access to the field site was gained through approval of a proposal submitted to the Center for the Advancement of Science in Space (also known as the ISS National Laboratory [ISS NL]). Upon acceptance, Axiom Space was assigned as the Implementation Partner for carriage of the experiment according to standard procedure. No other permits were required for this work.

Experiment design

Since our work envisioned one-meter sample squares, and recognizing the use of acronyms as a persistent element of spacefaring culture, we named our payload the Sampling Quadrangle Assemblages Research Experiment (SQuARE). Permission from the ISS NL to conduct SQuARE was contingent on using equipment that was already on board the space station. SQuARE required only five items: a camera, a wide-angle lens, adhesive tape (for marking the boundaries of the sample locations), a ruler (for scale), and a color calibration card (for post-processing of the images). All of these were already present on the ISS.

Walsh performed tests on the walls of a terrestrial art gallery to assess the feasibility of creating perfect one-meter squares in microgravity. He worked on a vertical surface, using the Pythagorean theorem to determine where the corners should be located. The only additional items used for these tests were two metric measuring tapes and a pencil for marking the wall (these were also already on the ISS). While it was possible to make a square this way, it also became clear that at least two people were needed to manage holding the tape measures in position while marking the points for the corners. This was not possible in the ISS context.

Walsh and Gorman identified seven locations for the placement of squares. Five of these were in the US Orbital Segment (USOS, consisting of American, European, and Japanese modules) and two in the Russian Orbital Segment. Unfortunately, tense relations between the US and Russian governments meant we could only document areas in the USOS. The five locations were (with their SQuARE designations):

• 01—an experimental rack on the forward wall, starboard end, of the Japanese Experiment Module
• 02—an experimental rack on the forward wall, port end, of the European laboratory module Columbus
• 03—the starboard Maintenance Work Area (workstation) in the US Node 2 module
• 04—the wall area “above” (according to typical crew body orientation) the galley table in the US Node 1 module
• 05—the aft wall, center location, of the US Node 3 module

Our square selection encompassed different modules and activities, including work and leisure. We also asked the crew to select a sixth sample location based on their understanding of the experiment and what they thought would be interesting to document. They chose a workstation on the port wall of the US laboratory module, at the aft end, which they described in a debriefing following their return to Earth in June 2022 as “our central command post, like our shared office situation in the lab.” Results from the four squares not included here will appear in future publications.

Walsh worked with NASA staff to determine payload procedures, including precise locations for the placement of the tape that would mark the square boundaries. The squares could not obstruct other facilities or experiments, so (unlike in terrestrial excavations, where string is typically used to demarcate trench boundaries) only the corners of each square were marked, not the entire perimeter. We used Kapton tape due to its bright yellow-orange color, which aided visibility for the crew taking photographs and for us when cropping the images. In practice, due to space constraints, the procedures that could actually be performed by crew in the ISS context, and the need to avoid interfering with other ongoing experiments, none of the locations actually measured one square meter or had precise 90° corners like a trench on Earth.

On January 14, 2022, NASA astronaut Kayla Barron set up the sample locations, marking the beginning of archaeological work in space ( S1 Movie ). For 30 days, starting on January 21, a crew member took photos of the sample locations at approximately the same time each day; the process was repeated at a random time each day for a second 30-day period to eliminate biases. Photography ended on March 21, 2022. The crew were instructed not to move any items prior to taking the photographs. Walsh led image management, including color and barrel distortion correction, fixing the alignment of each image, and cropping them to the boundaries of the taped corners.

Data processing—Item tagging, statistics, visualizations

We refer to each day’s photo as a “context” by analogy with chronologically-linked assemblages of artifacts and installations at terrestrial archaeological sites ( S1 and S2 Datasets). As previously noted, each context represented a moment roughly 24 hours distant from the previous one, showing evidence of changes in that time. ISS mission planners attempted to schedule the activity at the same time in the first month, but there were inevitable changes due to contingencies. Remarkably, the average time between contexts in Phase 1 was an almost-perfect 24h 0m 13s. Most of the Phase 1 photos were taken between 1200 and 1300 GMT (the time zone in which life on the ISS is organized). In Phase 2, the times were much more variable, but the average time between contexts during this period was still 23h 31m 45s. The earliest Phase 2 photo was taken at 0815 GMT, and the latest at 2101. We did not identify any meaningful differences between results from the two phases.

Since the “test pits” were formed of images rather than soil matrices, we needed a tool to capture information about the identity, nature, and location of every object. An open-source image annotator platform [ 18 ] mostly suited our needs. Brousseau rebuilt the platform to work within the constraints of our access to the imagery (turning it into a desktop tool with secure access to our private server), to permit a greater range of metadata to be added to each item or be imported, to autosave, and to export the resulting annotations. The tool also had to respect privacy and security limitations required by NASA.

The platform Brousseau developed and iterated was rechristened “Rocket-Anno” ( S1 File ). For each context photograph, the user draws an outline around every object, creating a polygon; each polygon is assigned a unique ID and the user provides the relevant descriptive information, using a controlled vocabulary developed for ISS material culture by Walsh and Gorman. Walsh and Abdullah used Rocket-Anno to tag the items in each context for Squares 03 and 05. Once all the objects were outlined for every context’s photograph, the tool exported a JSON file with all of the metadata for both the images themselves and all of the annotations, including the coordinate points for every polygon ( S3 Dataset ). We then developed Python code using Jupyter “notebooks” (an interactive development environment) that ingests the JSON file and generates dataframes for various facets of the data. Graham created a “core” notebook that exports summary statistics, calculates Brainerd-Robinson coefficients of similarity, and visualizes the changing use of the square over time by indicating use-areas based on artifact types and subtypes ( S2 File ). Walsh and Abdullah also wrote detailed square notes with context-by-context discussions and interpretations of features and patterns.

We asked NASA for access to the ISS Crew Planner, a computer system that shows each astronaut’s tasks in five-minute increments, to aid with our interpretation of contexts, but were denied. As a proxy, we use another, less detailed source: the ISS Daily Summary Reports (DSRs), published on a semi-regular basis by NASA on its website [ 19 ]. Any activities mentioned in the DSRs often must be connected with a context by inference. Therefore, our conclusions are likely less precise than if we had seen the Crew Planner, but they also more clearly represent the result of simply observing and interpreting the material culture record.

The crew during our sample period formed ISS Expedition 66 (October 2021-March 2022). They were responsible for the movement of objects in the sample squares as they carried out their daily tasks. The group consisted of two Russians affiliated with Roscosmos (the Russian space agency, 26%), one German belonging to the European Space Agency (ESA, 14%), and four Americans employed by NASA (57%). There were six men (86%) and one woman (14%), approximately equivalent to the historic proportions in the ISS population (84% and 16%, respectively). The Russian crew had their sleeping quarters at the aft end of the station, in the Zvezda module. The ESA astronaut slept in the European Columbus laboratory module. The four NASA crew slept in the US Node 2 module (see below). These arrangements emphasize the national character of discrete spaces around the ISS, also evident in our previous study of population distributions [ 15 ]. Both of the sample areas in this study were located in US modules.

Square 03 was placed in the starboard Maintenance Work Area (MWA, Fig 1 ), one of a pair of workstations located opposite one another in the center of the Node 2 module, with four crew berths towards the aft and a series of five ports for the docking of visiting crew/cargo vehicles and two modules on the forward end ( Fig 2 ). Node 2 (sometimes called “Harmony”) is a connector that links the US, Japanese, and European lab modules. According to prevailing design standards when the workstation was developed, an MWA “shall serve as the primary location for servicing and repair of maximum sized replacement unit/system components” [ 20 ]. Historic images published by NASA showing its use suggested that its primary function was maintenance of equipment and also scientific work that did not require a specific facility such as a centrifuge or furnace.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

An open crew berth is visible at right. The yellow dotted line indicates the boundaries of the sample area. Credit: NASA/ISSAP.

https://doi.org/10.1371/journal.pone.0304229.g001

Credit: Tor Finseth, by permission, modified by Justin Walsh.

https://doi.org/10.1371/journal.pone.0304229.g002

Square 03 measured 90.3 cm (top) x 87.8 (left) x 89.4 (bottom) x 87.6 (right), for an area of approximately 0.79 m 2 . Its primary feature was a blue metal panel with 40 square loop-type Velcro patches arranged in four rows of ten. During daily photography, many items were attached to the Velcro patches (or held by a clip or in a resealable bag which had its own hook-type Velcro). Above and below the blue panel were additional Velcro patches placed directly on the white plastic wall surface. These patches were white, in different sizes and shapes and irregularly arranged, indicating that they had been placed on the wall in response to different needs. Some were dirty, indicating long use. The patches below the blue panel were rarely used during the sample period, but the patches above were used frequently to hold packages of wet wipes, as well as resealable bags with electrostatic dispersion kits and other items. Outside the sample area, the primary features were a crew berth to the right, and a blue metal table attached to the wall below. This table, the primary component of the MWA, “provides a rigid surface on which to perform maintenance tasks,” according to NASA [ 21 ]. It is modular and can be oriented in several configurations, from flat against the wall to horizontal ( i . e ., perpendicular to the wall). A laptop to the left of the square occasionally showed information about work happening in the area.

In the 60 context photos of Square 03, we recorded 3,608 instances of items, an average of 60.1 (median = 60.5) per context. The lowest count was 24 in context 2 (where most of the wall was hidden from view behind an opaque storage bag), and the highest was 75 in both contexts 20 and 21. For comparison between squares, we can also calculate the item densities per m 2 . The average count was 76.1/m 2 (minimum = 30, maximum = 95). The count per context ( Fig 3(A)) began much lower than average in the first three contexts because of a portable glovebag and a stowage bag that obscured much of the sample square. It rose to an above-average level which was sustained (with the exception of contexts 11 and 12, which involved the appearance of another portable glovebag) until about context 43, when the count dipped again and the area seemed to show less use. Contexts 42–59 showed below-average numbers, as much as 20% lower than previously.

(a) Count of artifacts in Square 03 over time. (b) Proportions of artifacts by function in Square 03. Credit: Rao Hamza Ali.

https://doi.org/10.1371/journal.pone.0304229.g003

74 types of items appeared at least once here, belonging to six categories: equipment (41%), office supplies (31%), electronic (17%), stowage (9%), media (1%), and food (<1%). To better understand the significance of various items in the archaeological record, we assigned them to functional categories ( Table 1 , Fig 3(B)) . 35% of artifacts were restraints, or items used for holding other things in place; 12% for tools; 9% for containers; 9% for writing items; 6% for audiovisual items; 6% for experimental items; 4% for lights; 4% for safety items; 4% for body maintenance; 4% for power items; 3% for computing items; 1% for labels; and less than 1% drinks. We could not identify a function for two percent of the items.

https://doi.org/10.1371/journal.pone.0304229.t001

One of the project goals is understanding cultural adaptations to the microgravity environment. We placed special attention on “gravity surrogates,” pieces of (often simple) technology that are used in space to replicate the terrestrial experience of things staying where they are placed. Gravity surrogates include restraints and containers. It is quite noticeable that gravity surrogates comprise close to half of all items (44%) in Square 03, while the tools category, which might have been expected to be most prominent in an area designated for maintenance, is less than one-third as large (12%). Adding other groups associated with work, such as “experiment” and “light,” only brings the total to 22%.

Square 05 (Figs 2 and 4 ) was placed in a central location on the aft wall of the multipurpose Node 3 (“Tranquility”) module. This module does not include any specific science facilities. Instead, there are two large pieces of exercise equipment, the TVIS (Treadmill with Vibration Isolation Stabilization System, on the forward wall at the starboard end), and the ARED (Advanced Resistive Exercise Device, on the overhead wall at the port end). Use of the machines forms a significant part of crew activities, as they are required to exercise for two hours each day to counteract loss of muscle mass and bone density, and enable readjustment to terrestrial gravity on their return. The Waste and Hygiene Compartment (WHC), which includes the USOS latrine, is also here, on the forward wall in the center of the module, opposite Square 05. Finally, three modules are docked at Node 3’s port end. Most notable is the Cupola, a kind of miniature module on the nadir side with a panoramic window looking at Earth. This is the most popular leisure space for the crew, who often describe the hours they spend there. The Permanent Multipurpose Module (PMM) is docked on the forward side, storing equipment, food, and trash. In previous expeditions, some crew described installing a curtain in the PMM to create a private space for changing clothes and performing body maintenance activities such as cleaning oneself [ 22 , 23 ], but it was unclear whether that continued to be its function during the expedition we observed. One crew member during our sample period posted a video on Instagram showing the PMM interior and their efforts to re-stow equipment in a bag [ 24 ]. The last space attached to Node 3 is an experimental inflatable module docked on the aft side, called the Bigelow Expandable Activity Module (BEAM), which is used for storage of equipment.

The yellow dotted line indicates the boundaries of the sample area. The ARED machine is at the far upper right, on the overhead wall. The TVIS treadmill is outside this image to the left, on the forward wall. The WHC is directly behind the photographer. Credit: NASA/ISSAP.

https://doi.org/10.1371/journal.pone.0304229.g004

Square 05 was on a mostly featureless wall, with a vertical handrail in the middle. Handrails are metal bars located throughout the ISS that are used by the crew to hold themselves in place or provide a point from which to propel oneself to another location. NASA’s most recent design standards acknowledge that “[t]hey also serve as convenient locations for temporary mounting, affixing, or restraint of loose equipment and as attachment points for equipment” [ 25 ]. The handrail in Square 05 was used as an impromptu object restraint when a resealable bag filled with other bags was squeezed between the handrail and the wall.

The Brine Processing Assembly (BPA), a white plastic box which separates water from other components of urine for treatment and re-introduction to the station’s drinkable water supply [ 26 ], was fixed to the wall outside the square boundaries at lower left. A bungee cord was attached to both sides of the box; the one on the right was connected at its other end to the handrail attachment bracket. Numerous items were attached to or wedged into this bungee cord during the survey, bringing “gravity” into being. A red plastic duct ran through the square from top center into the BPA. This duct led from the latrine via the overhead wall. About halfway through the survey period, in context 32, the duct was wrapped in Kapton tape. According to the DSR for that day, “the crew used duct tape [ sic ] to make a seal around the BPA exhaust to prevent odor permeation in the cabin” [ 27 ], revealing an aspect of the crew’s experience of this area that is captured only indirectly in the context photograph. Permanently attached to the wall were approximately 20 loop-type Velcro patches in many shapes and sizes, placed in a seemingly random pattern that likely indicates that they were put there at different times and for different reasons.

Other common items in Square 05 were a mirror, a laptop computer, and an experimental item belonging to the German space agency DLR called the Touch Array Assembly [ 28 ]. The laptop moved just three times, and only by a few centimeters each time, during the sample period. The Touch Array was a black frame enclosing three metal surfaces which were being tested for their bacterial resistance; members of the crew touched the surfaces at various moments during the sample period. Finally, and most prominent due to its size, frequency of appearance, and use (judged by its movement between context photos) was an unidentified crew member’s toiletry kit.

By contrast with Square 03, 05 was the most irregular sample location, roughly twice as wide as it was tall. Its dimensions were 111 cm (top) x 61.9 (left) x 111.4 (bottom) x 64.6 (right), for an area of approximately 0.7 m 2 , about 89% of Square 03. We identified 1,830 instances of items in the 60 contexts, an average of 30.5 (median = 32) per context. The minimum was 18 items in context 5, and the maximum was 39 in contexts 24, 51, and 52. The average item density was 43.6/m 2 (minimum = 26, maximum = 56), 57% of Square 03.

The number of items trended upward throughout the sample period ( Fig 5(A)) . The largest spike occurred in context 6 with the appearance of the toiletry kit, which stored (and revealed) a number of related items. The kit can also be linked to one of the largest dips in item count, seen from contexts 52 to 53, when it was closed (but remained in the square). Other major changes can often be attributed to the addition and removal of bungee cords, which had other items such as carabiners and brackets attached. For example, the dip seen in context 25 correlates with the removal of a bungee cord with four carabiners.

(a) Count of artifacts and average count in Square 05 over time. (b) Proportions of artifacts by function in Square 05. Credit: Rao Hamza Ali.

https://doi.org/10.1371/journal.pone.0304229.g005

41 different item types were found in Square 05, about 55% as many as in Square 03. These belonged to five different categories: equipment (63%), electronic (17%), stowage (10%), office supplies (5%), and food (2%). The distribution of function proportions was quite different in this sample location ( Table 2 and Fig 5(B)) . Even though restraints were still most prominent, making up 32% of all items, body maintenance was almost as high (30%), indicating how strongly this area was associated with the activity of cleaning and caring for oneself. Computing (8%, represented by the laptop, which seems not to have been used), power (8%, from various cables), container (7%, resealable bags and Cargo Transfer Bags), and hygiene (6%, primarily the BPA duct) were the next most common items. Experiment was the function of 4% of the items, mostly the Touch Array, which appeared in every context, followed by drink (2%) and life support (1%). Safety, audiovisual, food, and light each made up less than 1% of the functional categories.

https://doi.org/10.1371/journal.pone.0304229.t002

Tracking changes over time is critical to understanding the activity happening in each area. We now explore how the assemblages change by calculating the Brainerd-Robinson Coefficient of Similarity [ 29 , 30 ] as operationalized by Peeples [ 31 , 32 ]. This metric is used in archaeology for comparing all pairs of the contexts by the proportions of categorical artifact data, here functional type. Applying the coefficient to the SQuARE contexts enables identification of time periods for distinct activities using artifact function and frequency alone, independent of documentary or oral evidence.

Multiple phases of activities took place in the square. Moments of connected activity are visible as red clusters in contexts 0–2, 11–12, 28–32, and 41 ( Fig 6(A)) . Combining this visualization with close observation of the photos themselves, we argue that there are actually eight distinct chronological periods.

• Contexts 0–2: Period 1 (S1 Fig in S3 File ) is a three-day period of work involving a portable glovebag (contexts 0–1) and a large blue stowage bag (context 2). It is difficult to describe trends in functional types because the glovebag and stowage bag obstruct the view of many objects. Items which appear at the top of the sample area, such as audiovisual and body maintenance items, are overemphasized in the data as a result. It appears that some kind of science is happening here, perhaps medical sample collection due to the presence of several small resealable bags visible in the glovebag. The work appears particularly intense in context 1, with the positioning of the video camera and light to point into the glovebag. These items indicate observation and oversight of crew activities by ground control. A white cargo transfer bag for storage and the stowage bag for holding packing materials in the context 2 photo likely relate to the packing of a Cargo Dragon vehicle that was docked to Node 2. The Dragon departed from the ISS for Earth, full of scientific samples, equipment, and crew personal items, a little more than three hours after the context 2 photo was taken [ 33 ].
• Contexts 3–10: Period 2 (S2 Fig in S3 File ) was a “stable” eight-day period in the sample, when little activity is apparent, few objects were moved or transferred in or out the square, and the primary function of the area seems to be storage rather than work. In context 6, a large Post-It notepad appeared in the center of the metal panel with a phone number written on it. This number belonged to another astronaut, presumably indicating that someone on the ISS had been told to call that colleague on the ground (for reasons of privacy, and in accordance with NASA rules for disseminating imagery, we have blurred the number in the relevant images). In context 8, the same notepad sheet had new writing appear on it, this time reading “COL A1 L1,” the location of an experimental rack in the European lab module.
• Contexts 11–12: Period 3 (S3 Fig in S3 File ) involves a second appearance of a portable glovebag (a different one from that used in contexts 0–1, according to its serial number), this time for a known activity, a concrete hardening experiment belonging to the European Space Agency [ 34 , 35 ]. This two-day phase indicates how the MWA space can be shared with non-US agencies when required. It also demonstrates the utility of this flexible area for work beyond biology/medicine, such as material science. Oversight of the crew’s activities by ground staff is evident from the positioning of the video camera and LED light pointing into the glovebag.
• Contexts 13–27: Period 4 (S4 Fig in S3 File ) is another stable fifteen-day period, similar to Period 2. Many items continued to be stored on the aluminum panel. The LED light’s presence is a trace of the activity in Period 3 that persists throughout this phase. Only in context 25 can a movement of the lamp potentially be connected to an activity relating to one of the stored items on the wall: at least one nitrile glove was removed from a resealable bag behind the lamp. In general, the primary identifiable activity during Period 4 is storage.
• Contexts 28–32: Period 5 (S5 Fig in S3 File ), by contrast, represents a short period of five days of relatively high and diverse activity. In context 28, a Microsoft Hololens augmented reality headset appeared. According to the DSR for the previous day, a training activity called Sidekick was carried out using the headset [ 36 ]. The following day, a Saturday, showed no change in the quantity or type of objects, but many were moved around and grouped by function—adhesive tape rolls were placed together, tools were moved from Velcro patches into pouches or straightened, and writing implements were placed in a vertical orientation when previously they were tilted. Context 29 represents a cleaning and re-organization of the sample area, which is a common activity for the crew on Saturdays [ 37 ]. Finally, in context 32, an optical coherence tomography scanner—a large piece of equipment for medical research involving crew members’ eyes—appeared [ 38 ]. This device was used previously during the sample period, but on the same day as the ESA concrete experiment, so that earlier work seems to have happened elsewhere [ 39 ].
• Contexts 33–40: Period 6 (S6 Fig in S3 File ) is the third stable period, in which almost no changes are visible over eight days. The only sign of activity is a digital timer which was started six hours before the context 39 image was made and continued to run at least through context 42.
• Context 41: Period 7 (S7 Fig in S3 File ) is a single context in which medical sample collection may have occurred. Resealable bags (some holding others) appeared in the center of the image and at lower right. One of the bags at lower right had a printed label reading “Reservoir Containers.” We were not able to discern which type of reservoir containers the label refers to, although the DSR for the day mentions “[Human Research Facility] Generic Saliva Collection,” without stating the location for this work [ 40 ]. Evidence from photos of other squares shows that labeled bags could be re-used for other purposes, so our interpretation of medical activity for this context is not conclusive.
• Contexts 42–60: Period 8 (S8 Fig in S3 File ) is the last and longest period of stability and low activity—eighteen days in which no specific activity other than the storage of items can be detected. The most notable change is the appearance for the first time of a foil water pouch in the central part of the blue panel.

Visualization of Brainerd-Robinson similarity, compared context-by-context by item function, for (a) Square 03 and (b) Square 05. The more alike a pair of contexts is, the higher the coefficient value, with a context compared against itself where a value of 200 equals perfect similarity. The resulting matrix of coefficients is visualized on a scale from blue to red where blue is lowest and red is highest similarity. The dark red diagonal line indicates complete similarity, where each context is compared to itself. Dark blue represents a complete difference. Credit: Shawn Graham.

https://doi.org/10.1371/journal.pone.0304229.g006

In the standards used at the time of installation, “stowage space” was the sixth design requirement listed for the MWA after accessibility; equipment size capability; scratch-resistant surfaces; capabilities for electrical, mechanical, vacuum, and fluid support during maintenance; and the accommodation of diagnostic equipment [ 20 ]. Only capabilities for fabrication were listed lower than stowage. Yet 50 of the 60 contexts (83%) fell within stable periods where little or no activity is identifiable in Square 03. According to the sample results, therefore, this area seems to exist not for “maintenance,” but primarily for the storage and arrangement of items. The most recent update of the design standards does not mention the MWA, but states, “Stowage location of tool kits should be optimized for accessibility to workstations and/or maintenance workbenches” [ 25 ]. Our observation confirms the importance of this suggestion.

The MWA was also a flexible location for certain science work, like the concrete study or crew health monitoring. Actual maintenance of equipment was hardly in evidence in the sample (possibly contexts 25, 39, and 44), and may not even have happened at all in this location. Some training did happen here, such as review of procedures for the Electromagnetic Levitator camera (instructions for changing settings on a high-speed camera appeared on the laptop screen; the day’s DSR shows that this camera is part of the Electromagnetic Levitator facility, located in the Columbus module [ 41 ]. The training required the use of the Hololens system (context 28 DSR, cited above).

Although many item types were represented in Square 03, it became clear during data capture how many things were basically static, unmoving and therefore unused, especially certain tools, writing implements, and body maintenance items. The MWA was seen as an appropriate place to store these items. It may be the case that their presence here also indicates that their function was seen as an appropriate one for this space, but the function(s) may not be carried out—or perhaps not in this location. Actualization of object function was only visible to us when the state of the item changed—it appeared, it moved, it changed orientation, it disappeared, or, in the case of artifacts that were grouped in collections rather than found as singletons, its shape changed or it became visibly smaller/lesser. We therefore have the opportunity to explore not only actuality of object use, but also potentiality of use or function, and the meaning of that quality for archaeological interpretation [ 42 , 43 ]. This possibility is particularly intriguing in light of the archaeological turn towards recognizing the agency of objects to impact human activity [ 44 , 45 ]. We will explore these implications in a future publication.

We performed the same chronological analysis for Square 05. Fig 6(B) represents the analysis for both item types and for item functions. We identified three major phases of activity, corresponding to contexts 0–5, 6–52, and 53–59 (S9-S11 Figs in S3 File ). The primary characteristics of these phases relate to an early period of unclear associations (0–5) marked by the presence of rolls of adhesive tape and a few body maintenance items (toothpaste and toothbrush, wet wipes); the appearance of a toiletry kit on the right side of the sample area, fully open with clear views of many of the items contained within (6–52); and finally, the closure of the toiletry kit so that its contents can no longer be seen (53–59). We interpret the phases as follows:

• Contexts 0–5: In Period 1 (six days, S9 Fig in S3 File ), while items such as a mirror, dental floss picks, wet wipes, and a toothbrush held in the end of a toothpaste tube were visible, the presence of various other kinds of items confounds easy interpretation. Two rolls of duct tape were stored on the handrail in the center of the sample area, and the Touch Array and laptop appeared in the center. Little movement can be identified, apart from a blue nitrile glove that appeared in context 1 and moved left across the area until it was wedged into the bungee cord for contexts 3 and 4. The tape rolls were removed prior to context 5. A collection of resealable bags was wedged behind the handrail in context 3, remaining there until context 9. Overall, this appears to be a period characterized by eclectic associations, showing an area without a clear designated function.
• Contexts 6–52: Period 2 (S10 Fig in S3 File ) is clearly the most significant one for this location due to its duration (47 days). It was dominated by the number of body maintenance items located in and around the toiletry kit, especially a white hand towel (on which a brown stain was visible from context 11, allowing us to confirm that the same towel was present until context 46). A second towel appeared alongside the toiletry kit in context 47, and the first one was fixed at the same time to the handrail, where it remained through the end of the sample period. A third towel appeared in context 52, attached to the handrail together with the first one by a bungee cord, continuing to the end of the sample period. Individual body maintenance items moved frequently from one context to the next, showing the importance of this type of activity for this part of Node 3. For reasons that are unclear, the mirror shifted orientation from vertical to diagonal in context 22, and then was put back in a vertical orientation in context 31 (a Monday, a day which is not traditionally associated with cleaning and organization). Collections of resealable bags appeared at various times, including a large one labeled “KYNAR BAG OF ZIPLOCKS” in green marker at the upper left part of the sample area beginning of context 12 (Kynar is a non-flammable plastic material that NASA prefers for resealable bags to the generic commercial off-the-shelf variety because it is non-flammable; however, its resistance to heat makes it less desirable for creating custom sizes, so bags made from traditional but flammable low-density polyethylene still dominate on the ISS [ 14 ]). The Kynar bag contained varying numbers of bags within it over time; occasionally, it appeared to be empty. The Touch Array changed orientation on seven of 47 days in period 2, or 15% of the time (12% of all days in the survey), showing activity associated with scientific research in this area. In context 49, a life-support item, the Airborne Particulate Monitor (APM) was installed [ 46 ]. This device, which measures “real-time particulate data” to assess hazards to crew health [ 47 ], persisted through the end of the sample period.
• Contexts 53–59: Period 3 (S11 Fig in S3 File ) appears as a seven-day phase marked by low activity. Visually, the most notable feature is the closure of the toiletry kit, which led to much lower counts of body maintenance items. Hardly any of the items on the wall moved at all during this period.

While body maintenance in the form of cleaning and caring for oneself could be an expected function for an area with exercise and excretion facilities, it is worth noting that the ISS provides, at most, minimal accommodation for this activity. A description of the WHC stated, “To provide privacy…an enclosure was added to the front of the rack. This enclosure, referred to as the Cabin, is approximately the size of a typical bathroom stall and provides room for system consumables and hygiene item stowage. Space is available to also support limited hygiene functions such as hand and body washing” [ 48 ]. A diagram of the WHC in the same publication shows the Cabin without a scale but suggests that it measures roughly 2 m (h) x .75 (w) x .75 (d), a volume of approximately 1.125 m 3 . NASA’s current design standards state that the body volume of a 95th percentile male astronaut is 0.99 m 3 [ 20 ], meaning that a person of that size would take up 88% of the space of the Cabin, leaving little room for performing cleaning functions—especially if the Cabin is used as apparently intended, to also hold “system consumables and hygiene item[s]” that would further diminish the usable volume. This situation explains why crews try to adapt other spaces, such as storage areas like the PMM, for these activities instead. According to the crew debriefing statement, only one of them used the WHC for body maintenance purposes; it is not clear whether the toiletry kit belonged to that individual. But the appearance of the toiletry kit in Square 05—outside of the WHC, in a public space where others frequently pass by—may have been a response to the limitations of the WHC Cabin. It suggests a need for designers to re-evaluate affordances for body maintenance practices and storage for related items.

Although Square 03 and 05 were different sizes and shapes, comparing the density of items by function shows evidence of their usage ( Table 3 ). The typical context in Square 03 had twice as many restraints and containers, but less than one-quarter as many body maintenance items as Square 05. 03 also had many tools, lights, audiovisual equipment, and writing implements, while there were none of any of these types in 05. 05 had life support and hygiene items which were missing from 03. It appears that flexibility and multifunctionality were key elements for 03, while in 05 there was emphasis on one primary function (albeit an improvised one, designated by the crew rather than architects or ground control), cleaning and caring for one’s body, with a secondary function of housing static equipment for crew hygiene and life support.

https://doi.org/10.1371/journal.pone.0304229.t003

As this is the first time such an analysis has been performed, it is not yet possible to say how typical or unusual these squares are regarding the types of activities taking place; but they provide a baseline for eventual comparison with the other four squares and future work on ISS or other space habitats.

Some general characteristics are revealed by archaeological analysis of a space station’s material culture. First, even in a small, enclosed site, occupied by only a few people over a relatively short sample period, we can observe divergent patterns for different locations and activity phases. Second, while distinct functions are apparent for these two squares, they are not the functions that we expected prior to this research. As a result, our work fulfills the promise of the archaeological approach to understanding life in a space station by revealing new, previously unrecognized phenomena relating to life and work on the ISS. There is now systematically recorded archaeological data for a space habitat.

Squares 03 and 05 served quite different purposes. The reasons for this fact are their respective affordances and their locations relative to activity areas designated for science and exercise. Their national associations, especially the manifestation of the control wielded by NASA over its modules, also played a role in the use of certain materials, the placement of facilities, and the organization of work. How each area was used was also the result of an interplay between the original plans developed by mission planners and habitat designers (or the lack of such plans), the utility of the equipment and architecture in each location, and the contingent needs of the crew as they lived in the station. This interplay became visible in the station’s material culture, as certain areas were associated with particular behaviors, over time and through tradition—over the long duration across many crews (Node 2, location of Square 03, docked with the ISS in 2007, and Node 3, location of Square 05, docked in 2010), and during the specific period of this survey, from January to March 2022. During the crew debriefing, one astronaut said, “We were a pretty organized crew who was also pretty much on the same page about how to do things…. As time went on…we organized the lab and kind of got on the same page about where we put things and how we’re going to do things.” This statement shows how functional associations can become linked to different areas of the ISS through usage and mutual agreement. At the same time, the station is not frozen in time. Different people have divergent ideas about how and where to do things. It seems from the appearance of just one Russian item—a packet of generic wipes ( salfetky sukhiye ) stored in the toiletry kit throughout the sample period—that the people who used these spaces and carried out their functions did not typically include the ISS’s Russian crew. Enabling greater flexibility to define how spaces can be used could have a significant impact on improving crew autonomy over their lives, such as how and where to work. It could also lead to opening of all spaces within a habitat to the entire crew, which seems likely to improve general well-being.

An apparent disjunction between planned and actual usage appeared in Square 03. It is intended for maintenance as well as other kinds of work. But much of the time, there was nobody working here—a fact that is not captured by historic photos of the area, precisely because nothing is happening. The space has instead become the equivalent of a pegboard mounted on a wall in a home garage or shed, convenient for storage for all kinds of items—not necessarily items being used there—because it has an enormous number of attachment points. Storage has become its primary function. Designers of future workstations in space should consider that they might need to optimize for functions other than work, because most of the time, there might not be any work happening there. They could optimize for quick storage, considering whether to impose a system of organization, or allow users to organize as they want.

We expected from previous (though unsystematic) observation of historic photos and other research, that resealable plastic bags (combined with Velcro patches on the bags and walls) would be the primary means for creating gravity surrogates to control items in microgravity. They only comprise 7% of all items in Square 03 (256 instances). There are more than twice as many clips (572—more than 9 per context) in the sample. There were 193 instances of adhesive tape rolls, and more than 100 cable ties, but these were latent (not holding anything), representing potentiality of restraint rather than actualization. The squares showed different approaches to managing “gravity.” While Square 03 had a pre-existing structured array of Velcro patches, Square 05 showed a more expedient strategy with Velcro added in response to particular activities. Different needs require different affordances; creating “gravity” is a more nuanced endeavor than it initially appears. More work remains to be done to optimize gravity surrogates for future space habitats, because this is evidently one of the most critical adaptations that crews have to make in microgravity (44% of all items in Square 03, 39% in 05).

Square 05 is an empty space, seemingly just one side of a passageway for people going to use the lifting machine or the latrine, to look out of the Cupola, or get something out of deep storage in one of the ISS’s closets. In our survey, this square was a storage place for toiletries, resealable bags, and a computer that never (or almost never) gets used. It was associated with computing and hygiene simply by virtue of its location, rather than due to any particular facilities it possessed. It has no affordances for storage. There are no cabinets or drawers, as would be appropriate for organizing and holding crew personal items. A crew member decided that this was an appropriate place to leave their toiletry kit for almost two months. Whether this choice was appreciated or resented by fellow crew members cannot be discerned based on our evidence, but it seems to have been tolerated, given its long duration. The location of the other four USOS crew members’ toiletry kits during the sample period is unknown. A question raised by our observations is: how might a function be more clearly defined by designers for this area, perhaps by providing lockers for individual crew members to store their toiletries and towels? This would have a benefit not only for reducing clutter, but also for reducing exposure of toiletry kits and the items stored in them to flying sweat from the exercise equipment or other waste particles from the latrine. A larger compartment providing privacy for body maintenance and a greater range of motion would also be desirable.

As the first systematic collection of archaeological data from a space site outside Earth, this analysis of two areas on the ISS as part of the SQuARE payload has shown that novel insights into material culture use can be obtained, such as the use of wall areas as storage or staging posts between activities, the accretion of objects associated with different functions, and the complexity of using material replacements for gravity. These results enable better space station design and raise new questions that will be addressed through analysis of the remaining four squares.

Supporting information

S1 movie. nasa astronaut kayla barron installs the first square for the sampling quadrangle assemblages research experiment in the japanese experiment module (also known as kibo) on the international space station, january 14, 2022..

She places Kapton tape to mark the square’s upper right corner. Credit: NASA.

https://doi.org/10.1371/journal.pone.0304229.s001

S1 Dataset.

https://doi.org/10.1371/journal.pone.0304229.s002

S2 Dataset.

https://doi.org/10.1371/journal.pone.0304229.s003

S3 Dataset. The image annotations are represented according to sample square in json formatted text files.

The data is available in the ‘SQuARE-notebooks’ repository on Github.com in the ‘data’ subfolder at https://github.com/issarchaeologicalproject/SQuARE-notebooks/tree/main ; archived version of the repository is at Zenodo, DOI: 10.5281/zenodo.10654812 .

https://doi.org/10.1371/journal.pone.0304229.s004

S1 File. The ‘Rocket-Anno’ image annotation software is available on Github at https://github.com/issarchaeologicalproject/MRE-RocketAnno .

The archived version of the repository is at Zenodo, DOI: 10.5281/zenodo.10648399 .

https://doi.org/10.1371/journal.pone.0304229.s005

S2 File. The computational notebooks that process the data json files to reshape the data suitable for basic statistics as well as the computation of the Brainerd-Robinson coefficients of similarity are in the.ipynb notebook format.

The code is available in the ‘SQuARE-notebooks’ repository on Github.com in the ‘notebooks’ subfolder at https://github.com/issarchaeologicalproject/SQuARE-notebooks/tree/main ; archived version of the repository is at Zenodo, DOI: 10.5281/zenodo.10654812 . The software can be run online in the Google Colab environment ( https://colab.research.google.com ) or any system running Jupyter Notebooks ( https://jupyter.org/ ).

https://doi.org/10.1371/journal.pone.0304229.s006

https://doi.org/10.1371/journal.pone.0304229.s007

Acknowledgments

We thank Chapman University’s Office of Research and Sponsored Programs, and especially Dr. Thomas Piechota and Dr. Janeen Hill, for funding the Implementation Partner costs associated with the SQuARE payload. Chapman’s Leatherby Libraries’ Supporting Open Access Research and Scholarship (SOARS) program funded the article processing fee for this publication. Ken Savin and Ken Shields at the ISS National Laboratory gave major support by agreeing to sponsor SQuARE and providing access to ISS NL’s allocation of crew time. David Zuniga and Kryn Ambs at Axiom Space were key collaborators in managing payload logistics. NASA staff and contractors were critical to the experiment’s success, especially Kristen Fortson, Jay Weber, Crissy Canerday, Sierra Wolbert, and Jade Conway. We also gratefully acknowledge the help and resources provided by Dr. Erik Linstead, director of the Machine Learning and Affiliated Technology Lab at Chapman University. Aidan St. P. Walsh corrected the color and lens barrel distortion in all of the SQuARE imagery. Rao Hamza Ali produced charts using accessible color combinations for Figs 3 and 5 . And finally, of course, we are extremely appreciative of the efforts of the five USOS members of the Expedition 66 crew on the ISS—Kayla Barron, Raja Chari, Thomas Marshburn, Matthias Maurer, and Mark Vande Hei—who were the first archaeologists in space.

• 1. Buchli V. Extraterrestrial methods: Towards an ethnography of the ISS. In: Carroll T, Walford A, Walton S, editors. Lineages and advancements in material culture studies: Perspectives from UCL anthropology. London: Routledge; 2021, pp. 17–32.
• 2. Gorman A, Walsh J. Archaeology in a vacuum: obstacles to and solutions for developing a real space archaeology. In: Barnard H, editor. Archaeology outside the box: investigations at the edge of the discipline. Los Angeles, Cotsen Institute of Archaeology Press; 2023. pp. 131–123.
• 3. Walsh J. Adapting to space: The International Space Station Archaeological Project. In: Salazar Sutil JF, Gorman A, editors. Routledge handbook of social studies of outer space. London, Routledge; 2023. pp. 400–412. https://doi.org/10.4324/9781003280507-37
• View Article
• Google Scholar
• 6. Rathje W, Murphy C. Rubbish! The archaeology of garbage Tucson: University of Arizona Press; 2001.
• 7. De León J. The land of open graves: living and dying on the migrant trail. Berkeley, University of California Press; 2015.
• 8. Garrison Darrin A, O’Leary B, editors. Handbook of space engineering, archaeology, and heritage. Boca Raton, CRC Press; 2009.
• 9. Capelotti PJ. The human archaeology of space: Lunar, planetary, and interstellar relics of exploration. Jefferson, NC, McFarland Press; 2010.
• 11. Gorman A. Space and time through material culture: An account of space archaeology. In: Salazar Sutil JF, Gorman A, editors. Routledge handbook of social studies of outer space. London, Routledge; 2023. pp. 44–56. https://doi.org/10.4324/9781003280507-5
• 17. NASA. NASA Johnson. 2008 Aug [cited May 12 2024]. In: Flickr [Internet]. San Francisco. Available from https://www.flickr.com/photos/nasa2explore/
• 19. NASA. ISS Daily Status Reports. 2012 Mar 1 [Cited May 12 2024]. Available from: https://blogs.nasa.gov/stationreport/
• 20. NASA. Man-systems integration. STD-3000 Vol. 1. Houston, NASA Johnson; 1995, pp. 9–15, 78
• 21. NASA. Maintenance Work Area | Glenn Research Center. 2020 Mar 6 [cited May 12 2024]. Available from: https://www1.grc.nasa.gov/space/iss-research/mwa/
• 22. Cristoforetti S. Diario di un’apprendista astronauta. Milan, Le Polene; 2018. pp. 379.
• 23. Kelly S. Endurance: A year in space, a lifetime of discovery. New York, Knopf; 2017. pp. 175, 285–86.
• 24. Barron K. Instagram post, 2022 Feb 12 [cited 2024 May 12]. Available from: https://www.instagram.com/tv/CZ4pW9HJ2Wg/?igsh=ZDE1MWVjZGVmZQ==
• 25. NASA. NASA space flight human-system standard. STD-3001 Volume 1: Human integration design handbook. Rev. 1 Houston, NASA Johnson; 2014. pp. 814, 829–833.
• 27. Keeter B. ISS daily summary report– 2/21/2022. 2022 Feb 21 [cited May 12 2024]. In: NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/02/page/6/
• 28. DLR. Fingerprint research to combat harmful bacteria. 2022 Jan 18 [cited May 12 2024]. Available from: https://www.dlr.de/en/latest/news/2022/01/20220118_fingerprint-research-to-combat-harmful-bacteria
• 31. Peeples MA. R script for calculating the Brainerd-Robinson coefficient of similarity and assessing sampling error. 2011 [cited May 12 2024]. Available from: http://www.mattpeeples.net/br.html .
• 33. Garcia M. Cargo Dragon Splashes Down Ending SpaceX CRS-24 Mission. 2022 Jan 24 [cited May 12 2024]. NASA Space Station blog [Internet]. Available from: https://blogs.nasa.gov/spacestation/2022/01/24/cargo-dragon-splashes-down-ending-spacex-crs-24-mission/
• 34. ESA. Concrete Hardening | Cosmic Kiss 360°. 2022 Mar 5 [cited May 12 2024]. Available from: https://www.esa.int/ESA_Multimedia/Videos/2022/05/Concrete_Hardening_Cosmic_Kiss_360
• 35. Keeter B. ISS daily summary report– 2/01/2022. 2022 Feb 1 [cited May 12 2024]. In: NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/02/page/19/
• 36. Keeter B. ISS daily summary report– 2/17/2022. 2022 Feb 17 [cited May 12 2024]. In: NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/02/page/8/
• 37. T. Pultarova, How Do You Clean a Space Station? Astronaut Thomas Pesquet Shares Orbital Spring Cleaning Tips, Space.com, May 6, 2021. Online at https://www.space.com/space-station-cleaning-tips-astronaut-thomas-pesquet
• 38. Keeter B. ISS daily summary report– 2/22/2022. 2022 Feb 22 [cited May 12 2024]. In: NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/02/page/5/
• 39. Keeter B. ISS daily summary report– 2/02/2022. 2022 Feb 2 [cited May 12 2024]. NASA ISS On-Orbit Status Report blog [Internet]. Houston. Online at https://blogs.nasa.gov/stationreport/2022/02/page/18/
• 40. Keeter B. ISS daily summary report– 3/03/2022. 2022 Mar 3 [cited May 12 2024]. In: NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/03/page/21/
• 41. Keeter B. ISS daily summary report– 2/08/2022. 2022 Feb 8 [cited May 12 2024]. NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/02/page/15/
• 42. Aristotle of Stageira. Metaphysics, Volume I: Books 1–9, Tredennick H, translator. Loeb Classical Library 271. Cambridge, MA, Harvard University Press; 1933. pp. 429–473.
• 44. Hodder I. Entangled: An archaeology of the relationships between humans and things. Hoboken. NJ, Wiley-Blackwell; 2012.
• 45. Malafouris L., How Things Shape the Mind: A Theory of Material Engagement (MIT Press, 2016).
• 46. Keeter B. ISS daily summary report– 3/11/2022. 2022 Mar 11 [cited May 12 2024]. NASA ISS On-Orbit Status Report blog [Internet]. Houston. Available from: https://blogs.nasa.gov/stationreport/2022/03/page/15/

• Privacy Policy

Home » Research – Types, Methods and Examples

Research – Types, Methods and Examples

Table of Contents

Definition:

Research refers to the process of investigating a particular topic or question in order to discover new information , develop new insights, or confirm or refute existing knowledge. It involves a systematic and rigorous approach to collecting, analyzing, and interpreting data, and requires careful planning and attention to detail.

History of Research

The history of research can be traced back to ancient times when early humans observed and experimented with the natural world around them. Over time, research evolved and became more systematic as people sought to better understand the world and solve problems.

In ancient civilizations such as those in Greece, Egypt, and China, scholars pursued knowledge through observation, experimentation, and the development of theories. They explored various fields, including medicine, astronomy, and mathematics.

During the Middle Ages, research was often conducted by religious scholars who sought to reconcile scientific discoveries with their faith. The Renaissance brought about a renewed interest in science and the scientific method, and the Enlightenment period marked a major shift towards empirical observation and experimentation as the primary means of acquiring knowledge.

The 19th and 20th centuries saw significant advancements in research, with the development of new scientific disciplines and fields such as psychology, sociology, and computer science. Advances in technology and communication also greatly facilitated research efforts.

Today, research is conducted in a wide range of fields and is a critical component of many industries, including healthcare, technology, and academia. The process of research continues to evolve as new methods and technologies emerge, but the fundamental principles of observation, experimentation, and hypothesis testing remain at its core.

Types of Research

Types of Research are as follows:

• Applied Research : This type of research aims to solve practical problems or answer specific questions, often in a real-world context.
• Basic Research : This type of research aims to increase our understanding of a phenomenon or process, often without immediate practical applications.
• Experimental Research : This type of research involves manipulating one or more variables to determine their effects on another variable, while controlling all other variables.
• Descriptive Research : This type of research aims to describe and measure phenomena or characteristics, without attempting to manipulate or control any variables.
• Correlational Research: This type of research examines the relationships between two or more variables, without manipulating any variables.
• Qualitative Research : This type of research focuses on exploring and understanding the meaning and experience of individuals or groups, often through methods such as interviews, focus groups, and observation.
• Quantitative Research : This type of research uses numerical data and statistical analysis to draw conclusions about phenomena or populations.
• Action Research: This type of research is often used in education, healthcare, and other fields, and involves collaborating with practitioners or participants to identify and solve problems in real-world settings.
• Mixed Methods Research : This type of research combines both quantitative and qualitative research methods to gain a more comprehensive understanding of a phenomenon or problem.
• Case Study Research: This type of research involves in-depth examination of a specific individual, group, or situation, often using multiple data sources.
• Longitudinal Research: This type of research follows a group of individuals over an extended period of time, often to study changes in behavior, attitudes, or health outcomes.
• Cross-Sectional Research : This type of research examines a population at a single point in time, often to study differences or similarities among individuals or groups.
• Survey Research: This type of research uses questionnaires or interviews to gather information from a sample of individuals about their attitudes, beliefs, behaviors, or experiences.
• Ethnographic Research : This type of research involves immersion in a cultural group or community to understand their way of life, beliefs, values, and practices.
• Historical Research : This type of research investigates events or phenomena from the past using primary sources, such as archival records, newspapers, and diaries.
• Content Analysis Research : This type of research involves analyzing written, spoken, or visual material to identify patterns, themes, or messages.
• Participatory Research : This type of research involves collaboration between researchers and participants throughout the research process, often to promote empowerment, social justice, or community development.
• Comparative Research: This type of research compares two or more groups or phenomena to identify similarities and differences, often across different countries or cultures.
• Exploratory Research : This type of research is used to gain a preliminary understanding of a topic or phenomenon, often in the absence of prior research or theories.
• Explanatory Research: This type of research aims to identify the causes or reasons behind a particular phenomenon, often through the testing of theories or hypotheses.
• Evaluative Research: This type of research assesses the effectiveness or impact of an intervention, program, or policy, often through the use of outcome measures.
• Simulation Research : This type of research involves creating a model or simulation of a phenomenon or process, often to predict outcomes or test theories.

Data Collection Methods

• Surveys : Surveys are used to collect data from a sample of individuals using questionnaires or interviews. Surveys can be conducted face-to-face, by phone, mail, email, or online.
• Experiments : Experiments involve manipulating one or more variables to measure their effects on another variable, while controlling for other factors. Experiments can be conducted in a laboratory or in a natural setting.
• Case studies : Case studies involve in-depth analysis of a single case, such as an individual, group, organization, or event. Case studies can use a variety of data collection methods, including interviews, observation, and document analysis.
• Observational research : Observational research involves observing and recording the behavior of individuals or groups in a natural setting. Observational research can be conducted covertly or overtly.
• Content analysis : Content analysis involves analyzing written, spoken, or visual material to identify patterns, themes, or messages. Content analysis can be used to study media, social media, or other forms of communication.
• Ethnography : Ethnography involves immersion in a cultural group or community to understand their way of life, beliefs, values, and practices. Ethnographic research can use a range of data collection methods, including observation, interviews, and document analysis.
• Secondary data analysis : Secondary data analysis involves using existing data from sources such as government agencies, research institutions, or commercial organizations. Secondary data can be used to answer research questions, without collecting new data.
• Focus groups: Focus groups involve gathering a small group of people together to discuss a topic or issue. The discussions are usually guided by a moderator who asks questions and encourages discussion.
• Interviews : Interviews involve one-on-one conversations between a researcher and a participant. Interviews can be structured, semi-structured, or unstructured, and can be conducted in person, by phone, or online.
• Document analysis : Document analysis involves collecting and analyzing written documents, such as reports, memos, and emails. Document analysis can be used to study organizational communication, policy documents, and other forms of written material.

Data Analysis Methods

Data Analysis Methods in Research are as follows:

• Descriptive statistics : Descriptive statistics involve summarizing and describing the characteristics of a dataset, such as mean, median, mode, standard deviation, and frequency distributions.
• Inferential statistics: Inferential statistics involve making inferences or predictions about a population based on a sample of data, using methods such as hypothesis testing, confidence intervals, and regression analysis.
• Qualitative analysis: Qualitative analysis involves analyzing non-numerical data, such as text, images, or audio, to identify patterns, themes, or meanings. Qualitative analysis can be used to study subjective experiences, social norms, and cultural practices.
• Content analysis: Content analysis involves analyzing written, spoken, or visual material to identify patterns, themes, or messages. Content analysis can be used to study media, social media, or other forms of communication.
• Grounded theory: Grounded theory involves developing a theory or model based on empirical data, using methods such as constant comparison, memo writing, and theoretical sampling.
• Discourse analysis : Discourse analysis involves analyzing language use, including the structure, function, and meaning of words and phrases, to understand how language reflects and shapes social relationships and power dynamics.
• Network analysis: Network analysis involves analyzing the structure and dynamics of social networks, including the relationships between individuals and groups, to understand social processes and outcomes.

Research Methodology

Research methodology refers to the overall approach and strategy used to conduct a research study. It involves the systematic planning, design, and execution of research to answer specific research questions or test hypotheses. The main components of research methodology include:

• Research design : Research design refers to the overall plan and structure of the study, including the type of study (e.g., observational, experimental), the sampling strategy, and the data collection and analysis methods.
• Sampling strategy: Sampling strategy refers to the method used to select a representative sample of participants or units from the population of interest. The choice of sampling strategy will depend on the research question and the nature of the population being studied.
• Data collection methods : Data collection methods refer to the techniques used to collect data from study participants or sources, such as surveys, interviews, observations, or secondary data sources.
• Data analysis methods: Data analysis methods refer to the techniques used to analyze and interpret the data collected in the study, such as descriptive statistics, inferential statistics, qualitative analysis, or content analysis.
• Ethical considerations: Ethical considerations refer to the principles and guidelines that govern the treatment of human participants or the use of sensitive data in the research study.
• Validity and reliability : Validity and reliability refer to the extent to which the study measures what it is intended to measure and the degree to which the study produces consistent and accurate results.

Applications of Research

Research has a wide range of applications across various fields and industries. Some of the key applications of research include:

• Advancing scientific knowledge : Research plays a critical role in advancing our understanding of the world around us. Through research, scientists are able to discover new knowledge, uncover patterns and relationships, and develop new theories and models.
• Improving healthcare: Research is instrumental in advancing medical knowledge and developing new treatments and therapies. Clinical trials and studies help to identify the effectiveness and safety of new drugs and medical devices, while basic research helps to uncover the underlying causes of diseases and conditions.
• Enhancing education: Research helps to improve the quality of education by identifying effective teaching methods, developing new educational tools and technologies, and assessing the impact of various educational interventions.
• Driving innovation: Research is a key driver of innovation, helping to develop new products, services, and technologies. By conducting research, businesses and organizations can identify new market opportunities, gain a competitive advantage, and improve their operations.
• Informing public policy : Research plays an important role in informing public policy decisions. Policy makers rely on research to develop evidence-based policies that address societal challenges, such as healthcare, education, and environmental issues.
• Understanding human behavior : Research helps us to better understand human behavior, including social, cognitive, and emotional processes. This understanding can be applied in a variety of settings, such as marketing, organizational management, and public policy.

Importance of Research

Research plays a crucial role in advancing human knowledge and understanding in various fields of study. It is the foundation upon which new discoveries, innovations, and technologies are built. Here are some of the key reasons why research is essential:

• Advancing knowledge: Research helps to expand our understanding of the world around us, including the natural world, social structures, and human behavior.
• Problem-solving: Research can help to identify problems, develop solutions, and assess the effectiveness of interventions in various fields, including medicine, engineering, and social sciences.
• Innovation : Research is the driving force behind the development of new technologies, products, and processes. It helps to identify new possibilities and opportunities for improvement.
• Evidence-based decision making: Research provides the evidence needed to make informed decisions in various fields, including policy making, business, and healthcare.
• Education and training : Research provides the foundation for education and training in various fields, helping to prepare individuals for careers and advancing their knowledge.
• Economic growth: Research can drive economic growth by facilitating the development of new technologies and innovations, creating new markets and job opportunities.

When to use Research

Research is typically used when seeking to answer questions or solve problems that require a systematic approach to gathering and analyzing information. Here are some examples of when research may be appropriate:

• To explore a new area of knowledge : Research can be used to investigate a new area of knowledge and gain a better understanding of a topic.
• To identify problems and find solutions: Research can be used to identify problems and develop solutions to address them.
• To evaluate the effectiveness of programs or interventions : Research can be used to evaluate the effectiveness of programs or interventions in various fields, such as healthcare, education, and social services.
• To inform policy decisions: Research can be used to provide evidence to inform policy decisions in areas such as economics, politics, and environmental issues.
• To develop new products or technologies : Research can be used to develop new products or technologies and improve existing ones.
• To understand human behavior : Research can be used to better understand human behavior and social structures, such as in psychology, sociology, and anthropology.

Characteristics of Research

The following are some of the characteristics of research:

• Purpose : Research is conducted to address a specific problem or question and to generate new knowledge or insights.
• Systematic : Research is conducted in a systematic and organized manner, following a set of procedures and guidelines.
• Empirical : Research is based on evidence and data, rather than personal opinion or intuition.
• Objective: Research is conducted with an objective and impartial perspective, avoiding biases and personal beliefs.
• Rigorous : Research involves a rigorous and critical examination of the evidence and data, using reliable and valid methods of data collection and analysis.
• Logical : Research is based on logical and rational thinking, following a well-defined and logical structure.
• Generalizable : Research findings are often generalized to broader populations or contexts, based on a representative sample of the population.
• Replicable : Research is conducted in a way that allows others to replicate the study and obtain similar results.
• Ethical : Research is conducted in an ethical manner, following established ethical guidelines and principles, to ensure the protection of participants’ rights and well-being.
• Cumulative : Research builds on previous studies and contributes to the overall body of knowledge in a particular field.

Advantages of Research

Research has several advantages, including:

• Generates new knowledge: Research is conducted to generate new knowledge and understanding of a particular topic or phenomenon, which can be used to inform policy, practice, and decision-making.
• Provides evidence-based solutions : Research provides evidence-based solutions to problems and issues, which can be used to develop effective interventions and strategies.
• Improves quality : Research can improve the quality of products, services, and programs by identifying areas for improvement and developing solutions to address them.
• Enhances credibility : Research enhances the credibility of an organization or individual by providing evidence to support claims and assertions.
• Enables innovation: Research can lead to innovation by identifying new ideas, approaches, and technologies.
• Informs decision-making : Research provides information that can inform decision-making, helping individuals and organizations make more informed and effective choices.
• Facilitates progress: Research can facilitate progress by identifying challenges and opportunities and developing solutions to address them.
• Enhances understanding: Research can enhance understanding of complex issues and phenomena, helping individuals and organizations navigate challenges and opportunities more effectively.
• Promotes accountability : Research promotes accountability by providing a basis for evaluating the effectiveness of policies, programs, and interventions.
• Fosters collaboration: Research can foster collaboration by bringing together individuals and organizations with diverse perspectives and expertise to address complex issues and problems.

Limitations of Research

Some Limitations of Research are as follows:

• Cost : Research can be expensive, particularly when large-scale studies are required. This can limit the number of studies that can be conducted and the amount of data that can be collected.
• Time : Research can be time-consuming, particularly when longitudinal studies are required. This can limit the speed at which research findings can be generated and disseminated.
• Sample size: The size of the sample used in research can limit the generalizability of the findings to larger populations.
• Bias : Research can be affected by bias, both in the design and implementation of the study, as well as in the analysis and interpretation of the data.
• Ethics : Research can present ethical challenges, particularly when human or animal subjects are involved. This can limit the types of research that can be conducted and the methods that can be used.
• Data quality: The quality of the data collected in research can be affected by a range of factors, including the reliability and validity of the measures used, as well as the accuracy of the data entry and analysis.
• Subjectivity : Research can be subjective, particularly when qualitative methods are used. This can limit the objectivity and reliability of the findings.
• Accessibility : Research findings may not be accessible to all stakeholders, particularly those who are not part of the academic or research community.
• Interpretation : Research findings can be open to interpretation, particularly when the data is complex or contradictory. This can limit the ability of researchers to draw firm conclusions.
• Unforeseen events : Unexpected events, such as changes in the environment or the emergence of new technologies, can limit the relevance and applicability of research findings.

About the author

Muhammad Hassan

Researcher, Academic Writer, Web developer

You may also like

Chemist – Definition, Types and Work Area

What is Archaeology – Definition and Overview

Concept – Definition, Types and Examples

Physicist – Definition, Types and Work Area

What is Science – Definition, Methods, Types

Ethos Pathos Logos – Definition, Meanings and...

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

A review of computer vision-based crack detection methods in civil infrastructure: progress and challenges.

1. Introduction

2. crack detection combining traditional image processing methods and deep learning, 2.1. crack detection based on image edge detection and deep learning, 2.2. crack detection based on threshold segmentation and deep learning, 2.3. crack detection based on morphological operations and deep learning, 3. crack detection based on multimodal data fusion, 3.1. multi-sensor fusion, 3.2. multi-source data fusion, 4. crack detection based on image semantic understanding, 4.1. crack detection based on classification networks, 4.2. crack detection based on object detection networks, 4.3. crack detection based on segmentation networks.

5. Datasets

6. evaluation index, 7. discussion, 8. conclusions, author contributions, data availability statement, acknowledgments, conflicts of interest.

• Azimi, M.; Eslamlou, A.D.; Pekcan, G. Data-driven structural health monitoring and damage detection through deep learning: State-of-the-art review. Sensors 2020 , 20 , 2778. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Han, X.; Zhao, Z. Structural surface crack detection method based on computer vision technology. J. Build. Struct. 2018 , 39 , 418–427. [ Google Scholar ]
• Kruachottikul, P.; Cooharojananone, N.; Phanomchoeng, G.; Chavarnakul, T.; Kovitanggoon, K.; Trakulwaranont, D. Deep learning-based visual defect-inspection system for reinforced concrete bridge substructure: A case of thailand’s department of highways. J. Civ. Struct. Health Monit. 2021 , 11 , 949–965. [ Google Scholar ] [ CrossRef ]
• Gehri, N.; Mata-Falcón, J.; Kaufmann, W. Automated crack detection and measurement based on digital image correlation. Constr. Build. Mater. 2020 , 256 , 119383. [ Google Scholar ] [ CrossRef ]
• Mohan, A.; Poobal, S. Crack detection using image processing: A critical review and analysis. Alex. Eng. J. 2018 , 57 , 787–798. [ Google Scholar ] [ CrossRef ]
• Liu, Y.; Fan, J.; Nie, J.; Kong, S.; Qi, Y. Review and prospect of digital-image-based crack detection of structure surface. China Civ. Eng. J. 2021 , 54 , 79–98. [ Google Scholar ]
• Hsieh, Y.-A.; Tsai, Y.J. Machine learning for crack detection: Review and model performance comparison. J. Comput. Civ. Eng. 2020 , 34 , 04020038. [ Google Scholar ] [ CrossRef ]
• Xu, Y.; Bao, Y.; Chen, J.; Zuo, W.; Li, H. Surface fatigue crack identification in steel box girder of bridges by a deep fusion convolutional neural network based on consumer-grade camera images. Struct. Health Monit. 2019 , 18 , 653–674. [ Google Scholar ] [ CrossRef ]
• Wang, W.; Deng, L.; Shao, X. Fatigue design of steel bridges considering the effect of dynamic vehicle loading and overloaded trucks. J. Bridge Eng. 2016 , 21 , 04016048. [ Google Scholar ] [ CrossRef ]
• Zheng, K.; Zhou, S.; Zhang, Y.; Wei, Y.; Wang, J.; Wang, Y.; Qin, X. Simplified evaluation of shear stiffness degradation of diagonally cracked reinforced concrete beams. Materials 2023 , 16 , 4752. [ Google Scholar ] [ CrossRef ]
• Canny, J. A computational approach to edge detection. IEEE Trans. Pattern Anal. Mach. Intell. 1986 , PAMI-8 , 679–698. [ Google Scholar ] [ CrossRef ]
• Otsu, N. A threshold selection method from gray-level histograms. Automatica 1975 , 11 , 23–27. [ Google Scholar ] [ CrossRef ]
• Sohn, H.G.; Lim, Y.M.; Yun, K.H.; Kim, G.H. Monitoring crack changes in concrete structures. Comput.-Aided Civ. Infrastruct. Eng. 2005 , 20 , 52–61. [ Google Scholar ] [ CrossRef ]
• Wang, P.; Qiao, H.; Feng, Q.; Xue, C. Internal corrosion cracks evolution in reinforced magnesium oxychloride cement concrete. Adv. Cem. Res. 2023 , 36 , 15–30. [ Google Scholar ] [ CrossRef ]
• Loutridis, S.; Douka, E.; Trochidis, A. Crack identification in double-cracked beams using wavelet analysis. J. Sound Vib. 2004 , 277 , 1025–1039. [ Google Scholar ] [ CrossRef ]
• Fan, C.L. Detection of multidamage to reinforced concrete using support vector machine-based clustering from digital images. Struct. Control Health Monit. 2021 , 28 , e2841. [ Google Scholar ] [ CrossRef ]
• Kyal, C.; Reza, M.; Varu, B.; Shreya, S. Image-based concrete crack detection using random forest and convolution neural network. In Computational Intelligence in Pattern Recognition: Proceedings of the International Conference on Computational Intelligence in Pattern Recognition (CIPR 2021), Held at the Institute of Engineering and Management, Kolkata, West Bengal, India, on 24–25 April 2021 ; Springer: Singapore, 2022; pp. 471–481. [ Google Scholar ]
• Jia, H.; Lin, J.; Liu, J. Bridge seismic damage assessment model applying artificial neural networks and the random forest algorithm. Adv. Civ. Eng. 2020 , 2020 , 6548682. [ Google Scholar ] [ CrossRef ]
• Park, M.J.; Kim, J.; Jeong, S.; Jang, A.; Bae, J.; Ju, Y.K. Machine learning-based concrete crack depth prediction using thermal images taken under daylight conditions. Remote Sens. 2022 , 14 , 2151. [ Google Scholar ] [ CrossRef ]
• LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015 , 521 , 436–444. [ Google Scholar ] [ CrossRef ]
• Liu, Z.; Cao, Y.; Wang, Y.; Wang, W. Computer vision-based concrete crack detection using u-net fully convolutional networks. Autom. Constr. 2019 , 104 , 129–139. [ Google Scholar ] [ CrossRef ]
• Li, G.; Ma, B.; He, S.; Ren, X.; Liu, Q. Automatic tunnel crack detection based on u-net and a convolutional neural network with alternately updated clique. Sensors 2020 , 20 , 717. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chaiyasarn, K.; Buatik, A.; Mohamad, H.; Zhou, M.; Kongsilp, S.; Poovarodom, N. Integrated pixel-level cnn-fcn crack detection via photogrammetric 3d texture mapping of concrete structures. Autom. Constr. 2022 , 140 , 104388. [ Google Scholar ] [ CrossRef ]
• Li, S.; Zhao, X.; Zhou, G. Automatic pixel-level multiple damage detection of concrete structure using fully convolutional network. Comput.-Aided Civ. Infrastruct. Eng. 2019 , 34 , 616–634. [ Google Scholar ] [ CrossRef ]
• Zheng, X.; Zhang, S.; Li, X.; Li, G.; Li, X. Lightweight bridge crack detection method based on segnet and bottleneck depth-separable convolution with residuals. IEEE Access 2021 , 9 , 161649–161668. [ Google Scholar ] [ CrossRef ]
• Azouz, Z.; Honarvar Shakibaei Asli, B.; Khan, M. Evolution of crack analysis in structures using image processing technique: A review. Electronics 2023 , 12 , 3862. [ Google Scholar ] [ CrossRef ]
• Hamishebahar, Y.; Guan, H.; So, S.; Jo, J. A comprehensive review of deep learning-based crack detection approaches. Appl. Sci. 2022 , 12 , 1374. [ Google Scholar ] [ CrossRef ]
• Meng, S.; Gao, Z.; Zhou, Y.; He, B.; Djerrad, A. Real-time automatic crack detection method based on drone. Comput.-Aided Civ. Infrastruct. Eng. 2023 , 38 , 849–872. [ Google Scholar ] [ CrossRef ]
• Humpe, A. Bridge inspection with an off-the-shelf 360 camera drone. Drones 2020 , 4 , 67. [ Google Scholar ] [ CrossRef ]
• Truong-Hong, L.; Lindenbergh, R. Automatically extracting surfaces of reinforced concrete bridges from terrestrial laser scanning point clouds. Autom. Constr. 2022 , 135 , 104127. [ Google Scholar ] [ CrossRef ]
• Cusson, D.; Rossi, C.; Ozkan, I.F. Early warning system for the detection of unexpected bridge displacements from radar satellite data. J. Civ. Struct. Health Monit. 2021 , 11 , 189–204. [ Google Scholar ] [ CrossRef ]
• Bonaldo, G.; Caprino, A.; Lorenzoni, F.; da Porto, F. Monitoring displacements and damage detection through satellite MT-INSAR techniques: A new methodology and application to a case study in rome (Italy). Remote Sens. 2023 , 15 , 1177. [ Google Scholar ] [ CrossRef ]
• Zheng, Z.; Zhong, Y.; Wang, J.; Ma, A.; Zhang, L. Building damage assessment for rapid disaster response with a deep object-based semantic change detection framework: From natural disasters to man-made disasters. Remote Sens. Environ. 2021 , 265 , 112636. [ Google Scholar ] [ CrossRef ]
• Chen, X.; Zhang, X.; Ren, M.; Zhou, B.; Sun, M.; Feng, Z.; Chen, B.; Zhi, X. A multiscale enhanced pavement crack segmentation network coupling spectral and spatial information of UAV hyperspectral imagery. Int. J. Appl. Earth Obs. Geoinf. 2024 , 128 , 103772. [ Google Scholar ] [ CrossRef ]
• Liu, F.; Liu, J.; Wang, L. Deep learning and infrared thermography for asphalt pavement crack severity classification. Autom. Constr. 2022 , 140 , 104383. [ Google Scholar ] [ CrossRef ]
• Liu, S.; Han, Y.; Xu, L. Recognition of road cracks based on multi-scale retinex fused with wavelet transform. Array 2022 , 15 , 100193. [ Google Scholar ] [ CrossRef ]
• Zhang, H.; Qian, Z.; Tan, Y.; Xie, Y.; Li, M. Investigation of pavement crack detection based on deep learning method using weakly supervised instance segmentation framework. Constr. Build. Mater. 2022 , 358 , 129117. [ Google Scholar ] [ CrossRef ]
• Dorafshan, S.; Thomas, R.J.; Maguire, M. Comparison of deep convolutional neural networks and edge detectors for image-based crack detection in concrete. Constr. Build. Mater. 2018 , 186 , 1031–1045. [ Google Scholar ] [ CrossRef ]
• Munawar, H.S.; Hammad, A.W.; Haddad, A.; Soares, C.A.P.; Waller, S.T. Image-based crack detection methods: A review. Infrastructures 2021 , 6 , 115. [ Google Scholar ] [ CrossRef ]
• Chen, D.; Li, X.; Hu, F.; Mathiopoulos, P.T.; Di, S.; Sui, M.; Peethambaran, J. Edpnet: An encoding–decoding network with pyramidal representation for semantic image segmentation. Sensors 2023 , 23 , 3205. [ Google Scholar ] [ CrossRef ]
• Mo, S.; Shi, Y.; Yuan, Q.; Li, M. A survey of deep learning road extraction algorithms using high-resolution remote sensing images. Sensors 2024 , 24 , 1708. [ Google Scholar ] [ CrossRef ]
• Chen, D.; Li, J.; Di, S.; Peethambaran, J.; Xiang, G.; Wan, L.; Li, X. Critical points extraction from building façades by analyzing gradient structure tensor. Remote Sens. 2021 , 13 , 3146. [ Google Scholar ] [ CrossRef ]
• Liu, Y.; Yeoh, J.K.; Chua, D.K. Deep learning-based enhancement of motion blurred UAV concrete crack images. J. Comput. Civ. Eng. 2020 , 34 , 04020028. [ Google Scholar ] [ CrossRef ]
• Flah, M.; Nunez, I.; Ben Chaabene, W.; Nehdi, M.L. Machine learning algorithms in civil structural health monitoring: A systematic review. Arch. Comput. Methods Eng. 2021 , 28 , 2621–2643. [ Google Scholar ] [ CrossRef ]
• Li, G.; Li, X.; Zhou, J.; Liu, D.; Ren, W. Pixel-level bridge crack detection using a deep fusion about recurrent residual convolution and context encoder network. Measurement 2021 , 176 , 109171. [ Google Scholar ] [ CrossRef ]
• Ali, R.; Chuah, J.H.; Talip, M.S.A.; Mokhtar, N.; Shoaib, M.A. Structural crack detection using deep convolutional neural networks. Autom. Constr. 2022 , 133 , 103989. [ Google Scholar ] [ CrossRef ]
• Wang, H.; Li, Y.; Dang, L.M.; Lee, S.; Moon, H. Pixel-level tunnel crack segmentation using a weakly supervised annotation approach. Comput. Ind. 2021 , 133 , 103545. [ Google Scholar ] [ CrossRef ]
• Zhu, J.; Song, J. Weakly supervised network based intelligent identification of cracks in asphalt concrete bridge deck. Alex. Eng. J. 2020 , 59 , 1307–1317. [ Google Scholar ] [ CrossRef ]
• Li, Y.; Bao, T.; Xu, B.; Shu, X.; Zhou, Y.; Du, Y.; Wang, R.; Zhang, K. A deep residual neural network framework with transfer learning for concrete dams patch-level crack classification and weakly-supervised localization. Measurement 2022 , 188 , 110641. [ Google Scholar ] [ CrossRef ]
• Yang, Q.; Shi, W.; Chen, J.; Lin, W. Deep convolution neural network-based transfer learning method for civil infrastructure crack detection. Autom. Constr. 2020 , 116 , 103199. [ Google Scholar ] [ CrossRef ]
• Dais, D.; Bal, I.E.; Smyrou, E.; Sarhosis, V. Automatic crack classification and segmentation on masonry surfaces using convolutional neural networks and transfer learning. Autom. Constr. 2021 , 125 , 103606. [ Google Scholar ] [ CrossRef ]
• Abdellatif, M.; Peel, H.; Cohn, A.G.; Fuentes, R. Combining block-based and pixel-based approaches to improve crack detection and localisation. Autom. Constr. 2021 , 122 , 103492. [ Google Scholar ] [ CrossRef ]
• Dan, D.; Dan, Q. Automatic recognition of surface cracks in bridges based on 2D-APES and mobile machine vision. Measurement 2021 , 168 , 108429. [ Google Scholar ] [ CrossRef ]
• Weng, X.; Huang, Y.; Wang, W. Segment-based pavement crack quantification. Autom. Constr. 2019 , 105 , 102819. [ Google Scholar ] [ CrossRef ]
• Kao, S.-P.; Chang, Y.-C.; Wang, F.-L. Combining the YOLOv4 deep learning model with UAV imagery processing technology in the extraction and quantization of cracks in bridges. Sensors 2023 , 23 , 2572. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Li, X.; Xu, X.; He, X.; Wei, X.; Yang, H. Intelligent crack detection method based on GM-ResNet. Sensors 2023 , 23 , 8369. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Choi, Y.; Park, H.W.; Mi, Y.; Song, S. Crack detection and analysis of concrete structures based on neural network and clustering. Sensors 2024 , 24 , 1725. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Guo, J.-M.; Markoni, H.; Lee, J.-D. BARNet: Boundary aware refinement network for crack detection. IEEE Trans. Intell. Transp. Syst. 2021 , 23 , 7343–7358. [ Google Scholar ] [ CrossRef ]
• Luo, J.; Lin, H.; Wei, X.; Wang, Y. Adaptive canny and semantic segmentation networks based on feature fusion for road crack detection. IEEE Access 2023 , 11 , 51740–51753. [ Google Scholar ] [ CrossRef ]
• Ranyal, E.; Sadhu, A.; Jain, K. Enhancing pavement health assessment: An attention-based approach for accurate crack detection, measurement, and mapping. Expert Syst. Appl. 2024 , 247 , 123314. [ Google Scholar ] [ CrossRef ]
• Liu, K.; Chen, B.M. Industrial UAV-based unsupervised domain adaptive crack recognitions: From database towards real-site infrastructural inspections. IEEE Trans. Ind. Electron. 2022 , 70 , 9410–9420. [ Google Scholar ] [ CrossRef ]
• Wang, W.; Hu, W.; Wang, W.; Xu, X.; Wang, M.; Shi, Y.; Qiu, S.; Tutumluer, E. Automated crack severity level detection and classification for ballastless track slab using deep convolutional neural network. Autom. Constr. 2021 , 124 , 103484. [ Google Scholar ] [ CrossRef ]
• Xu, Z.; Zhang, X.; Chen, W.; Liu, J.; Xu, T.; Wang, Z. Muraldiff: Diffusion for ancient murals restoration on large-scale pre-training. IEEE Trans. Emerg. Top. Comput. Intell. 2024 , 8 , 2169–2181. [ Google Scholar ] [ CrossRef ]
• Bradley, D.; Roth, G. Adaptive thresholding using the integral image. J. Graph. Tools 2007 , 12 , 13–21. [ Google Scholar ] [ CrossRef ]
• Sezgin, M.; Sankur, B.l. Survey over image thresholding techniques and quantitative performance evaluation. J. Electron. Imaging 2004 , 13 , 146–168. [ Google Scholar ]
• Kapur, J.N.; Sahoo, P.K.; Wong, A.K. A new method for gray-level picture thresholding using the entropy of the histogram. Comput. Vis. Graph. Image Process. 1985 , 29 , 273–285. [ Google Scholar ] [ CrossRef ]
• Pal, N.R.; Pal, S.K. A review on image segmentation techniques. Pattern Recognit. 1993 , 26 , 1277–1294. [ Google Scholar ] [ CrossRef ]
• Flah, M.; Suleiman, A.R.; Nehdi, M.L. Classification and quantification of cracks in concrete structures using deep learning image-based techniques. Cem. Concr. Compos. 2020 , 114 , 103781. [ Google Scholar ] [ CrossRef ]
• Mazni, M.; Husain, A.R.; Shapiai, M.I.; Ibrahim, I.S.; Anggara, D.W.; Zulkifli, R. An investigation into real-time surface crack classification and measurement for structural health monitoring using transfer learning convolutional neural networks and otsu method. Alex. Eng. J. 2024 , 92 , 310–320. [ Google Scholar ] [ CrossRef ]
• He, Z.; Xu, W. Deep learning and image preprocessing-based crack repair trace and secondary crack classification detection method for concrete bridges. Struct. Infrastruct. Eng. 2024 , 20 , 1–17. [ Google Scholar ] [ CrossRef ]
• He, T.; Li, H.; Qian, Z.; Niu, C.; Huang, R. Research on weakly supervised pavement crack segmentation based on defect location by generative adversarial network and target re-optimization. Constr. Build. Mater. 2024 , 411 , 134668. [ Google Scholar ] [ CrossRef ]
• Su, H.; Wang, X.; Han, T.; Wang, Z.; Zhao, Z.; Zhang, P. Research on a U-Net bridge crack identification and feature-calculation methods based on a CBAM attention mechanism. Buildings 2022 , 12 , 1561. [ Google Scholar ] [ CrossRef ]
• Kang, D.; Benipal, S.S.; Gopal, D.L.; Cha, Y.-J. Hybrid pixel-level concrete crack segmentation and quantification across complex backgrounds using deep learning. Autom. Constr. 2020 , 118 , 103291. [ Google Scholar ] [ CrossRef ]
• Lei, Q.; Zhong, J.; Wang, C. Joint optimization of crack segmentation with an adaptive dynamic threshold module. IEEE Trans. Intell. Transp. Syst. 2024 , 25 , 6902–6916. [ Google Scholar ] [ CrossRef ]
• Lei, Q.; Zhong, J.; Wang, C.; Xia, Y.; Zhou, Y. Dynamic thresholding for accurate crack segmentation using multi-objective optimization. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, Turin, Italy, 18 September 2023 ; Springer: Cham, Switzerland, 2023; pp. 389–404. [ Google Scholar ]
• Vincent, L.; Soille, P. Watersheds in digital spaces: An efficient algorithm based on immersion simulations. IEEE Trans. Pattern Anal. Mach. Intell. 1991 , 13 , 583–598. [ Google Scholar ] [ CrossRef ]
• Huang, H.; Zhao, S.; Zhang, D.; Chen, J. Deep learning-based instance segmentation of cracks from shield tunnel lining images. Struct. Infrastruct. Eng. 2022 , 18 , 183–196. [ Google Scholar ] [ CrossRef ]
• Fan, Z.; Lin, H.; Li, C.; Su, J.; Bruno, S.; Loprencipe, G. Use of parallel resnet for high-performance pavement crack detection and measurement. Sustainability 2022 , 14 , 1825. [ Google Scholar ] [ CrossRef ]
• Kong, S.Y.; Fan, J.S.; Liu, Y.F.; Wei, X.C.; Ma, X.W. Automated crack assessment and quantitative growth monitoring. Comput.-Aided Civ. Infrastruct. Eng. 2021 , 36 , 656–674. [ Google Scholar ] [ CrossRef ]
• Dang, L.M.; Wang, H.; Li, Y.; Park, Y.; Oh, C.; Nguyen, T.N.; Moon, H. Automatic tunnel lining crack evaluation and measurement using deep learning. Tunn. Undergr. Space Technol. 2022 , 124 , 104472. [ Google Scholar ] [ CrossRef ]
• Andrushia, A.D.; Anand, N.; Lubloy, E. Deep learning based thermal crack detection on structural concrete exposed to elevated temperature. Adv. Struct. Eng. 2021 , 24 , 1896–1909. [ Google Scholar ] [ CrossRef ]
• Dang, L.M.; Wang, H.; Li, Y.; Nguyen, L.Q.; Nguyen, T.N.; Song, H.-K.; Moon, H. Deep learning-based masonry crack segmentation and real-life crack length measurement. Constr. Build. Mater. 2022 , 359 , 129438. [ Google Scholar ] [ CrossRef ]
• Nguyen, A.; Gharehbaghi, V.; Le, N.T.; Sterling, L.; Chaudhry, U.I.; Crawford, S. ASR crack identification in bridges using deep learning and texture analysis. Structures 2023 , 50 , 494–507. [ Google Scholar ] [ CrossRef ]
• Dong, C.; Li, L.; Yan, J.; Zhang, Z.; Pan, H.; Catbas, F.N. Pixel-level fatigue crack segmentation in large-scale images of steel structures using an encoder–decoder network. Sensors 2021 , 21 , 4135. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Jian, L.; Chengshun, L.; Guanhong, L.; Zhiyuan, Z.; Bo, H.; Feng, G.; Quanyi, X. Lightweight defect detection equipment for road tunnels. IEEE Sens. J. 2023 , 24 , 5107–5121. [ Google Scholar ]
• Liang, H.; Qiu, D.; Ding, K.-L.; Zhang, Y.; Wang, Y.; Wang, X.; Liu, T.; Wan, S. Automatic pavement crack detection in multisource fusion images using similarity and difference features. IEEE Sens. J. 2023 , 24 , 5449–5465. [ Google Scholar ] [ CrossRef ]
• Alamdari, A.G.; Ebrahimkhanlou, A. A multi-scale robotic approach for precise crack measurement in concrete structures. Autom. Constr. 2024 , 158 , 105215. [ Google Scholar ] [ CrossRef ]
• Liu, H.; Kollosche, M.; Laflamme, S.; Clarke, D.R. Multifunctional soft stretchable strain sensor for complementary optical and electrical sensing of fatigue cracks. Smart Mater. Struct. 2023 , 32 , 045010. [ Google Scholar ] [ CrossRef ]
• Dang, D.-Z.; Wang, Y.-W.; Ni, Y.-Q. Nonlinear autoregression-based non-destructive evaluation approach for railway tracks using an ultrasonic fiber bragg grating array. Constr. Build. Mater. 2024 , 411 , 134728. [ Google Scholar ] [ CrossRef ]
• Yan, M.; Tan, X.; Mahjoubi, S.; Bao, Y. Strain transfer effect on measurements with distributed fiber optic sensors. Autom. Constr. 2022 , 139 , 104262. [ Google Scholar ] [ CrossRef ]
• Shukla, H.; Piratla, K. Leakage detection in water pipelines using supervised classification of acceleration signals. Autom. Constr. 2020 , 117 , 103256. [ Google Scholar ] [ CrossRef ]
• Chen, X.; Zhang, X.; Li, J.; Ren, M.; Zhou, B. A new method for automated monitoring of road pavement aging conditions based on recurrent neural network. IEEE Trans. Intell. Transp. Syst. 2022 , 23 , 24510–24523. [ Google Scholar ] [ CrossRef ]
• Zhang, S.; He, X.; Xue, B.; Wu, T.; Ren, K.; Zhao, T. Segment-anything embedding for pixel-level road damage extraction using high-resolution satellite images. Int. J. Appl. Earth Obs. Geoinf. 2024 , 131 , 103985. [ Google Scholar ] [ CrossRef ]
• Park, S.E.; Eem, S.-H.; Jeon, H. Concrete crack detection and quantification using deep learning and structured light. Constr. Build. Mater. 2020 , 252 , 119096. [ Google Scholar ] [ CrossRef ]
• Yan, Y.; Mao, Z.; Wu, J.; Padir, T.; Hajjar, J.F. Towards automated detection and quantification of concrete cracks using integrated images and lidar data from unmanned aerial vehicles. Struct. Control Health Monit. 2021 , 28 , e2757. [ Google Scholar ] [ CrossRef ]
• Dong, Q.; Wang, S.; Chen, X.; Jiang, W.; Li, R.; Gu, X. Pavement crack detection based on point cloud data and data fusion. Philos. Trans. R. Soc. A 2023 , 381 , 20220165. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kim, H.; Lee, S.; Ahn, E.; Shin, M.; Sim, S.-H. Crack identification method for concrete structures considering angle of view using RGB-D camera-based sensor fusion. Struct. Health Monit. 2021 , 20 , 500–512. [ Google Scholar ] [ CrossRef ]
• Chen, J.; Lu, W.; Lou, J. Automatic concrete defect detection and reconstruction by aligning aerial images onto semantic-rich building information model. Comput.-Aided Civ. Infrastruct. Eng. 2023 , 38 , 1079–1098. [ Google Scholar ] [ CrossRef ]
• Pozzer, S.; Rezazadeh Azar, E.; Dalla Rosa, F.; Chamberlain Pravia, Z.M. Semantic segmentation of defects in infrared thermographic images of highly damaged concrete structures. J. Perform. Constr. Facil. 2021 , 35 , 04020131. [ Google Scholar ] [ CrossRef ]
• Kaur, R.; Singh, S. A comprehensive review of object detection with deep learning. Digit. Signal Process. 2023 , 132 , 103812. [ Google Scholar ] [ CrossRef ]
• Sharma, V.K.; Mir, R.N. A comprehensive and systematic look up into deep learning based object detection techniques: A review. Comput. Sci. Rev. 2020 , 38 , 100301. [ Google Scholar ] [ CrossRef ]
• Zhang, L.; Yang, F.; Zhang, Y.D.; Zhu, Y.J. Road crack detection using deep convolutional neural network. In Proceedings of the 2016 IEEE International Conference on Image Processing (ICIP), Phoenix, AZ, USA, 25–28 September 2016; IEEE: Piscataway, NJ, USA, 2016; pp. 3708–3712. [ Google Scholar ]
• Yang, C.; Chen, J.; Li, Z.; Huang, Y. Structural crack detection and recognition based on deep learning. Appl. Sci. 2021 , 11 , 2868. [ Google Scholar ] [ CrossRef ]
• Rajadurai, R.-S.; Kang, S.-T. Automated vision-based crack detection on concrete surfaces using deep learning. Appl. Sci. 2021 , 11 , 5229. [ Google Scholar ] [ CrossRef ]
• Kim, B.; Yuvaraj, N.; Sri Preethaa, K.; Arun Pandian, R. Surface crack detection using deep learning with shallow CNN architecture for enhanced computation. Neural Comput. Appl. 2021 , 33 , 9289–9305. [ Google Scholar ] [ CrossRef ]
• O’Brien, D.; Osborne, J.A.; Perez-Duenas, E.; Cunningham, R.; Li, Z. Automated crack classification for the CERN underground tunnel infrastructure using deep learning. Tunn. Undergr. Space Technol. 2023 , 131 , 104668. [ Google Scholar ]
• Chen, K.; Reichard, G.; Xu, X.; Akanmu, A. Automated crack segmentation in close-range building façade inspection images using deep learning techniques. J. Build. Eng. 2021 , 43 , 102913. [ Google Scholar ] [ CrossRef ]
• Dong, Z.; Wang, J.; Cui, B.; Wang, D.; Wang, X. Patch-based weakly supervised semantic segmentation network for crack detection. Constr. Build. Mater. 2020 , 258 , 120291. [ Google Scholar ] [ CrossRef ]
• Buatik, A.; Thansirichaisree, P.; Kalpiyapun, P.; Khademi, N.; Pasityothin, I.; Poovarodom, N. Mosaic crack mapping of footings by convolutional neural networks. Sci. Rep. 2024 , 14 , 7851. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Zhang, Y.; Zhang, L. Detection of pavement cracks by deep learning models of transformer and UNet. arXiv 2023 , arXiv:2304.12596. [ Google Scholar ] [ CrossRef ]
• Al-Huda, Z.; Peng, B.; Algburi, R.N.A.; Al-antari, M.A.; Rabea, A.-J.; Zhai, D. A hybrid deep learning pavement crack semantic segmentation. Eng. Appl. Artif. Intell. 2023 , 122 , 106142. [ Google Scholar ] [ CrossRef ]
• Shamsabadi, E.A.; Xu, C.; Rao, A.S.; Nguyen, T.; Ngo, T.; Dias-da-Costa, D. Vision transformer-based autonomous crack detection on asphalt and concrete surfaces. Autom. Constr. 2022 , 140 , 104316. [ Google Scholar ] [ CrossRef ]
• Huang, S.; Tang, W.; Huang, G.; Huangfu, L.; Yang, D. Weakly supervised patch label inference networks for efficient pavement distress detection and recognition in the wild. IEEE Trans. Intell. Transp. Syst. 2023 , 24 , 5216–5228. [ Google Scholar ] [ CrossRef ]
• Huang, G.; Huang, S.; Huangfu, L.; Yang, D. Weakly supervised patch label inference network with image pyramid for pavement diseases recognition in the wild. In Proceedings of the ICASSP 2021—2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada, 6–11 June 2021; IEEE: Piscataway, NJ, USA, 2021; pp. 7978–7982. [ Google Scholar ]
• Guo, J.-M.; Markoni, H. Efficient and adaptable patch-based crack detection. IEEE Trans. Intell. Transp. Syst. 2022 , 23 , 21885–21896. [ Google Scholar ] [ CrossRef ]
• König, J.; Jenkins, M.D.; Mannion, M.; Barrie, P.; Morison, G. Weakly-supervised surface crack segmentation by generating pseudo-labels using localization with a classifier and thresholding. IEEE Trans. Intell. Transp. Syst. 2022 , 23 , 24083–24094. [ Google Scholar ] [ CrossRef ]
• Al-Huda, Z.; Peng, B.; Algburi, R.N.A.; Al-antari, M.A.; Rabea, A.-J.; Al-maqtari, O.; Zhai, D. Asymmetric dual-decoder-U-Net for pavement crack semantic segmentation. Autom. Constr. 2023 , 156 , 105138. [ Google Scholar ] [ CrossRef ]
• Wen, T.; Lang, H.; Ding, S.; Lu, J.J.; Xing, Y. PCDNet: Seed operation-based deep learning model for pavement crack detection on 3d asphalt surface. J. Transp. Eng. Part B Pavements 2022 , 148 , 04022023. [ Google Scholar ] [ CrossRef ]
• Mishra, A.; Gangisetti, G.; Eftekhar Azam, Y.; Khazanchi, D. Weakly supervised crack segmentation using crack attention networks on concrete structures. Struct. Health Monit. 2024 , 23 , 14759217241228150. [ Google Scholar ] [ CrossRef ]
• Kompanets, A.; Pai, G.; Duits, R.; Leonetti, D.; Snijder, B. Deep learning for segmentation of cracks in high-resolution images of steel bridges. arXiv 2024 , arXiv:2403.17725. [ Google Scholar ]
• Liu, Y.; Yeoh, J.K. Robust pixel-wise concrete crack segmentation and properties retrieval using image patches. Autom. Constr. 2021 , 123 , 103535. [ Google Scholar ] [ CrossRef ]
• Girshick, R.; Donahue, J.; Darrell, T.; Malik, J. Rich feature hierarchies for accurate object detection and semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 23–28 June 2014; pp. 580–587. [ Google Scholar ]
• Girshick, R. Fast R-CNN. In Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile, 7–13 December 2015; pp. 1440–1448. [ Google Scholar ]
• Ren, S.; He, K.; Girshick, R.; Sun, J. Faster R-CNN: Towards real-time object detection with region proposal networks. In Proceedings of the Advances in Neural Information Processing Systems, Montreal, QC, Canada, 7–12 December 2015; Volume 28. [ Google Scholar ]
• He, K.; Gkioxari, G.; Dollár, P.; Girshick, R. Mask R-CNN. In Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy, 22–29 October 2017; pp. 2961–2969. [ Google Scholar ]
• Redmon, J.; Divvala, S.; Girshick, R.; Farhadi, A. You only look once: Unified, real-time object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 27–30 June 2016; pp. 779–788. [ Google Scholar ]
• Redmon, J.; Farhadi, A. YOLO9000: Better, faster, stronger. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, 21–26 July 2017; pp. 7263–7271. [ Google Scholar ]
• Redmon, J.; Farhadi, A. Yolov3: An incremental improvement. arXiv 2018 , arXiv:1804.02767. [ Google Scholar ]
• Bochkovskiy, A.; Wang, C.-Y.; Liao, H.-Y.M. Yolov4: Optimal speed and accuracy of object detection. arXiv 2020 , arXiv:2004.10934. [ Google Scholar ]
• Wang, C.-Y.; Bochkovskiy, A.; Liao, H.-Y.M. Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada, 18–22 June 2023; pp. 7464–7475. [ Google Scholar ]
• Liu, W.; Anguelov, D.; Erhan, D.; Szegedy, C.; Reed, S.; Fu, C.-Y.; Berg, A.C. SSD: Single shot multibox detector. In Proceedings of the Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, 11–14 October 2016; Part I 14. Springer: Berlin/Heidelberg, Germany, 2016; pp. 21–37. [ Google Scholar ]
• Lin, T.-Y.; Goyal, P.; Girshick, R.; He, K.; Dollár, P. Focal loss for dense object detection. In Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy, 22–29 October 2017; pp. 2980–2988. [ Google Scholar ]
• Xu, Y.; Li, D.; Xie, Q.; Wu, Q.; Wang, J. Automatic defect detection and segmentation of tunnel surface using modified mask R-CNN. Measurement 2021 , 178 , 109316. [ Google Scholar ] [ CrossRef ]
• Zhao, W.; Liu, Y.; Zhang, J.; Shao, Y.; Shu, J. Automatic pixel-level crack detection and evaluation of concrete structures using deep learning. Struct. Control Health Monit. 2022 , 29 , e2981. [ Google Scholar ] [ CrossRef ]
• Li, R.; Yu, J.; Li, F.; Yang, R.; Wang, Y.; Peng, Z. Automatic bridge crack detection using unmanned aerial vehicle and faster R-CNN. Constr. Build. Mater. 2023 , 362 , 129659. [ Google Scholar ] [ CrossRef ]
• Tran, T.S.; Nguyen, S.D.; Lee, H.J.; Tran, V.P. Advanced crack detection and segmentation on bridge decks using deep learning. Constr. Build. Mater. 2023 , 400 , 132839. [ Google Scholar ] [ CrossRef ]
• Zhang, J.; Qian, S.; Tan, C. Automated bridge crack detection method based on lightweight vision models. Complex Intell. Syst. 2023 , 9 , 1639–1652. [ Google Scholar ] [ CrossRef ]
• Ren, R.; Liu, F.; Shi, P.; Wang, H.; Huang, Y. Preprocessing of crack recognition: Automatic crack-location method based on deep learning. J. Mater. Civ. Eng. 2023 , 35 , 04022452. [ Google Scholar ] [ CrossRef ]
• Liu, Z.; Yeoh, J.K.; Gu, X.; Dong, Q.; Chen, Y.; Wu, W.; Wang, L.; Wang, D. Automatic pixel-level detection of vertical cracks in asphalt pavement based on gpr investigation and improved mask R-CNN. Autom. Constr. 2023 , 146 , 104689. [ Google Scholar ] [ CrossRef ]
• Li, Z.; Zhu, H.; Huang, M. A deep learning-based fine crack segmentation network on full-scale steel bridge images with complicated backgrounds. IEEE Access 2021 , 9 , 114989–114997. [ Google Scholar ] [ CrossRef ]
• Alipour, M.; Harris, D.K.; Miller, G.R. Robust pixel-level crack detection using deep fully convolutional neural networks. J. Comput. Civ. Eng. 2019 , 33 , 04019040. [ Google Scholar ] [ CrossRef ]
• Wang, S.; Pan, Y.; Chen, M.; Zhang, Y.; Wu, X. FCN-SFW: Steel structure crack segmentation using a fully convolutional network and structured forests. IEEE Access 2020 , 8 , 214358–214373. [ Google Scholar ] [ CrossRef ]
• Hang, J.; Wu, Y.; Li, Y.; Lai, T.; Zhang, J.; Li, Y. A deep learning semantic segmentation network with attention mechanism for concrete crack detection. Struct. Health Monit. 2023 , 22 , 3006–3026. [ Google Scholar ] [ CrossRef ]
• Sun, Y.; Yang, Y.; Yao, G.; Wei, F.; Wong, M. Autonomous crack and bughole detection for concrete surface image based on deep learning. IEEE Access 2021 , 9 , 85709–85720. [ Google Scholar ] [ CrossRef ]
• Wang, Z.; Leng, Z.; Zhang, Z. A weakly-supervised transformer-based hybrid network with multi-attention for pavement crack detection. Constr. Build. Mater. 2024 , 411 , 134134. [ Google Scholar ] [ CrossRef ]
• Chen, T.; Cai, Z.; Zhao, X.; Chen, C.; Liang, X.; Zou, T.; Wang, P. Pavement crack detection and recognition using the architecture of segNet. J. Ind. Inf. Integr. 2020 , 18 , 100144. [ Google Scholar ] [ CrossRef ]
• Bai, S.; Ma, M.; Yang, L.; Liu, Y. Pixel-wise crack defect segmentation with dual-encoder fusion network. Constr. Build. Mater. 2024 , 426 , 136179. [ Google Scholar ] [ CrossRef ]
• Wang, W.; Su, C. Semi-supervised semantic segmentation network for surface crack detection. Autom. Constr. 2021 , 128 , 103786. [ Google Scholar ] [ CrossRef ]
• Tabernik, D.; Šela, S.; Skvarč, J.; Skočaj, D. Segmentation-based deep-learning approach for surface-defect detection. J. Intell. Manuf. 2020 , 31 , 759–776. [ Google Scholar ] [ CrossRef ]
• König, J.; Jenkins, M.D.; Mannion, M.; Barrie, P.; Morison, G. Optimized deep encoder-decoder methods for crack segmentation. Digit. Signal Process. 2021 , 108 , 102907. [ Google Scholar ] [ CrossRef ]
• Wang, C.; Liu, H.; An, X.; Gong, Z.; Deng, F. Swincrack: Pavement crack detection using convolutional swin-transformer network. Digit. Signal Process. 2024 , 145 , 104297. [ Google Scholar ] [ CrossRef ]
• Lan, Z.-X.; Dong, X.-M. Minicrack: A simple but efficient convolutional neural network for pixel-level narrow crack detection. Comput. Ind. 2022 , 141 , 103698. [ Google Scholar ] [ CrossRef ]
• Salton, G. Introduction to Modern Information Retrieval ; McGraw-Hill: New York, NY, USA, 1983. [ Google Scholar ]
• Jenkins, M.D.; Carr, T.A.; Iglesias, M.I.; Buggy, T.; Morison, G. A deep convolutional neural network for semantic pixel-wise segmentation of road and pavement surface cracks. In Proceedings of the 2018 26th European Signal Processing Conference (EUSIPCO), Rome, Italy, 3–7 September 2018; IEEE: Piscataway, NJ, USA; pp. 2120–2124. [ Google Scholar ]
• Tsai, Y.-C.; Chatterjee, A. Comprehensive, quantitative crack detection algorithm performance evaluation system. J. Comput. Civ. Eng. 2017 , 31 , 04017047. [ Google Scholar ] [ CrossRef ]
• Li, H.; Wang, J.; Zhang, Y.; Wang, Z.; Wang, T. A study on evaluation standard for automatic crack detection regard the random fractal. arXiv 2020 , arXiv:2007.12082. [ Google Scholar ]

Click here to enlarge figure

Share and Cite

Yuan, Q.; Shi, Y.; Li, M. A Review of Computer Vision-Based Crack Detection Methods in Civil Infrastructure: Progress and Challenges. Remote Sens. 2024 , 16 , 2910. https://doi.org/10.3390/rs16162910

Yuan Q, Shi Y, Li M. A Review of Computer Vision-Based Crack Detection Methods in Civil Infrastructure: Progress and Challenges. Remote Sensing . 2024; 16(16):2910. https://doi.org/10.3390/rs16162910

Yuan, Qi, Yufeng Shi, and Mingyue Li. 2024. "A Review of Computer Vision-Based Crack Detection Methods in Civil Infrastructure: Progress and Challenges" Remote Sensing 16, no. 16: 2910. https://doi.org/10.3390/rs16162910

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Energy & Environmental Science

Health prognostics for lithium-ion batteries: mechanisms, methods, and prospects.

* Corresponding authors

a Department of Energy, Aalborg University, Aalborg, Denmark

b College of Mechanical and Vehicle Engineering, Chongqing University, Chongqing, China E-mail: [email protected]

c Department of Automotive and Mechatronics Engineering, Ontario Tech University, Oshawa, Canada

Lithium-ion battery aging mechanism analysis and health prognostics are of great significance for a smart battery management system to ensure safe and optimal use of the battery system. This paper provides a comprehensive review of aging mechanisms and the state-of-the-art health prognostic methods and summarizes the main challenges and research prospects for battery health prognostics. First, the complex relationships among aging mechanisms, aging modes, influencing factors, and aging types are reviewed and summarized. Then, the battery health prognostic methods are divided according to different time scales and objectives, which include the short-term state of health estimation, long-term end-of-life prediction, and degradation trajectory prediction, followed by a detailed review of each prognostic task and method. For consistency, we first provide a clear and concise description of each method, showing the similarities and peculiarities of these methods, and then review several representative works. After that, comparative evaluations are conducted. The main advantages and disadvantages of each prognostic task and prognostic method are analyzed in detail. Next, key challenges are presented by considering the specific characteristics of each prognostic task. Moreover, for each challenge, potential solutions are presented and discussed. These proposed potential solutions to the main challenges are beneficial and can be considered by researchers in their further studies. Finally, the future trends of battery health prognostics are discussed, and several new ideas for battery health prognostics are proposed.

Article information

Download Citation

Permissions.

Y. Che, X. Hu, X. Lin, J. Guo and R. Teodorescu, Energy Environ. Sci. , 2023,  16 , 338 DOI: 10.1039/D2EE03019E

This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence . You can use material from this article in other publications, without requesting further permission from the RSC, provided that the correct acknowledgement is given and it is not used for commercial purposes.

To request permission to reproduce material from this article in a commercial publication , please go to the Copyright Clearance Center request page .

If you are an author contributing to an RSC publication, you do not need to request permission provided correct acknowledgement is given.

If you are the author of this article, you do not need to request permission to reproduce figures and diagrams provided correct acknowledgement is given. If you want to reproduce the whole article in a third-party commercial publication (excluding your thesis/dissertation for which permission is not required) please go to the Copyright Clearance Center request page .

Read more about how to correctly acknowledge RSC content .

Social activity

Search articles by author, advertisements.

Research on data analysis method of oscillating water column wave energy converters

• Zhang, Yaqun
• Li, Xianhao
• Sheng, Songwei
• Fan, Zhaohui
• Liu, Jingfeng

Taking the floating oscillating water column central tube wave energy converters (OWC WECs) as the research goal, the Fourier transform (FFT), and the ensemble empirical mode decomposition method (EEMD) were used to statistically analyze the wave height, motion acceleration, velocity and displacement of the device in the physical model test from the perspective of a frequency domain. Through the Fourier transform, the energy spectrum of the incident wave and the motion of the WEC is obtained, and it is found that the peak frequency of the two is the same, that is, the period is 1.138s, but the peak energy of the former is slightly greater than that of the latter. Through the ensemble empirical mode decomposition method, the 1-8 intrinsic mode functions of the incident wave and the displacement of the device are obtained, which shows that the device efficiently captures the dominant frequency wave in the incident wave, that is, the wave height is 0.0628m, the period is 1.1s, and the motion displacement of the vertical direction obtained by the device is 0.0585mm, and the period is 1.1119s. At the same time, it absorbs low-frequency, long-period reflected waves in the pool. The analysis results of the two methods are highly consistent, and the OWC WEC shows good hydrodynamic performance, with efficient wave trapping ability and fast wave responsability. This paper's research methods and results provide a basis for the later optimization of the device and a new idea for the data processing method of the WECs.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 07 August 2024

Highest ocean heat in four centuries places Great Barrier Reef in danger

• Benjamin J. Henley   ORCID: orcid.org/0000-0003-3940-1963 1 , 2 , 3 ,
• Helen V. McGregor   ORCID: orcid.org/0000-0002-4031-2282 1 , 2 ,
• Andrew D. King   ORCID: orcid.org/0000-0001-9006-5745 4 , 5 ,
• Ove Hoegh-Guldberg   ORCID: orcid.org/0000-0001-7510-6713 6 ,
• Ariella K. Arzey 1 , 2 ,
• David J. Karoly 4 ,
• Janice M. Lough 7 ,
• Thomas M. DeCarlo   ORCID: orcid.org/0000-0003-3269-1320 8 , 9 &
• Braddock K. Linsley   ORCID: orcid.org/0000-0003-2085-0662 10  

Nature volume  632 ,  pages 320–326 ( 2024 ) Cite this article

11k Accesses

1 Citations

2751 Altmetric

Metrics details

• Climate change
• Environmental impact
• Palaeoclimate

Mass coral bleaching on the Great Barrier Reef (GBR) in Australia between 2016 and 2024 was driven by high sea surface temperatures (SST) 1 . The likelihood of temperature-induced bleaching is a key determinant for the future threat status of the GBR 2 , but the long-term context of recent temperatures in the region is unclear. Here we show that the January–March Coral Sea heat extremes in 2024, 2017 and 2020 (in order of descending mean SST anomalies) were the warmest in 400 years, exceeding the 95th-percentile uncertainty limit of our reconstructed pre-1900 maximum. The 2016, 2004 and 2022 events were the next warmest, exceeding the 90th-percentile limit. Climate model analysis confirms that human influence on the climate system is responsible for the rapid warming in recent decades. This attribution, together with the recent ocean temperature extremes, post-1900 warming trend and observed mass coral bleaching, shows that the existential threat to the GBR ecosystem from anthropogenic climate change is now realized. Without urgent intervention, the iconic GBR is at risk of experiencing temperatures conducive to near-annual coral bleaching 3 , with negative consequences for biodiversity and ecosystems services. A continuation on the current trajectory would further threaten the ecological function 4 and outstanding universal value 5 of one of Earth’s greatest natural wonders.

Similar content being viewed by others

Mesophotic coral bleaching associated with changes in thermocline depth

Atypical weather patterns cause coral bleaching on the Great Barrier Reef, Australia during the 2021–2022 La Niña

Internal tides can provide thermal refugia that will buffer some coral reefs from future global warming

Like many coral reefs globally, the World Heritage-listed GBR in Australia is under threat 4 , 6 . Mass coral bleaching, declining calcification rates 5 , 7 , outbreaks of crown-of-thorns starfish ( Acanthaster spp.) 8 , severe tropical cyclones 9 and overfishing 10 have placed compounding detrimental pressures on the reef ecosystem. Coral bleaching typically occurs when heat stress triggers the breakdown of the symbiosis between corals and their symbiotic dinoflagellates 11 . Although coral bleaching can occur locally as a result of low salinity, cold waters or pollution, regional and global mass bleaching events, in which the majority of corals in one or more regions bleach at once, are strongly associated with increasing SST linked to global warming 2 .

The first modern observations of mass coral bleaching on the GBR occurred in the 1980s, but these events were less widespread and generally less severe 3 than the bleaching events in the twenty-first century 4 . Stress bands in coral skeletal cores have provided potential evidence for pre-1980s bleaching in the GBR and Coral Sea, such as during the 1877–78 El Niño 12 . However, stress bands are evident in relatively few cores before 1980 (ref. 12 ),  suggesting that severe mass bleaching did not occur in the 1800s and most of the 1900s.

As the oceans have warmed, however, mass coral bleaching events have become increasingly lethal to corals 4 . Coral bleaching on the GBR 1 in 1998 coincided with a strong eastern-Pacific El Niño, and in 2002 with a weak El Niño. El Niño events can induce lower cloud cover and increased solar irradiance over the GBR 13 , increasing the risk of thermal stress and mass bleaching events 14 . In 2004, water temperatures were anomalously warm, and although bleaching occurred in the Coral Sea 15 , it was not widespread in the GBR, probably because there was reduced upwelling and an associated reduced influence of nutrients on symbiotic dinoflagellate expulsion 16 .

However, in the nine January–March periods from 2016 to 2024 (inclusive) there were five mass coral bleaching events on the GBR. Each was associated with high SSTs and affected large sections of the reef. GBR mass bleaching occurred in both 2016 and 2017, influenced by the presence of an El Niño event in 2016, and led to the death of at least 50% of shallow-water (depths of 5–10 m) reef-building corals 4 . Major bleaching events occurred again in quick succession in 2020 and 2022, with the accumulated heat stress for large sections of the GBR reaching levels conducive to widespread bleaching but lower levels of coral mortality 1 . The bleaching event in 2022 occurred, unusually, during a La Niña event, which is typically associated with cooler summer SSTs, higher than average rainfall and higher cloud cover on the GBR 1 . At the time of writing, researchers are assessing the impacts of the 2024 mass bleaching event.

The frequency of recent mass coral bleaching and mortality on the GBR is cause for concern. In 2021, the World Heritage Committee of the United Nations Educational, Scientific and Cultural Organization (UNESCO) drafted 17 a decision to inscribe the GBR on the List of World Heritage in Danger, stating that the reef is “facing ascertained danger”, citing recent mass coral bleaching events and insufficient progress by the State Party (Australia) in countering climate change, improving water quality and land management issues. The committee’s adopted decisions 18 have not included inscription of the ‘in danger’ status, but the draft inscription highlights the seriousness of the recent mass coral bleaching events. Authorities in Australia 5 have noted that climate change and coral bleaching have deteriorated the integrity of the outstanding universal value of the GBR, a defining feature of its World Heritage status.

Although rapidly rising SSTs are attributed to human activities with virtual certainty 19 , understanding the multi-century SST history of the GBR is critical to understanding the influence of SST on mass coral bleaching and mortality in recent decades. Putting aside a problematic attempt to do this 20 , which was discredited 21 , 22 , knowledge of the long-term context for GBR SSTs comes primarily from two multi-century reconstructions based on the geochemistry of coral cores collected from the inner shelf 23 and outer shelf 24 (Flinders Reef) in the central GBR. These reconstructions showed that SSTs in the early 2000s were not unusually high relative to levels in the past three centuries, with five-year mean SSTs (and salinities) estimated to be higher in the 1700s than in the 1900s. However, these records were limited by their relatively coarse five-year sampling resolution and their most recent data point being from the early 2000s. After these studies were published, SSTs in the GBR have continued to rise. Updated analysis of coral data from Flinders Reef provides valuable improved temporal resolution 25 , but interpretations of these records remain limited spatially.

Here, we investigate the recent high SST events in the GBR region in the context of the past four centuries. We combine a network of 22 coral Sr/Ca and δ 18 O palaeothermometer series (Supplementary Tables 1 and 2 ) located in and near to the Coral Sea region to infer spatial mean SST anomalies (SSTAs) for January–March, the months when maximum SST and thermal bleaching are most likely to occur in the Coral Sea 16 , 26 , each year from 1618 to 1995 ( Methods and Supplementary Information ). Anthropogenic climate change began and proceeded entirely within the multi-century lives of some of these massive coral colonies, offering a continuous multi-century record covering the industrial era. We use this 1618–1995 reconstruction and the available 1900–2024 instrumental data to contextualize the modern trend and rank four centuries of January–March SSTAs with greater precision than was previously possible. We then assess the degree of human influence on ocean temperatures in the region using climate model simulations run both with and without anthropogenic forcing.

The instrumental period (1900–present)

Mass coral bleaching on the GBR in 2016, 2017, 2020, 2022 and 2024 during January–March coincided with widespread warm SSTAs in the surrounding seas 1 , including the Coral Sea (Fig. 1a–e , using ERSSTv5 data 27 ). The Coral Sea and GBR have experienced a strong warming trend since 1900 (Fig. 1f ). January–March SSTAs averaged over the GBR are strongly correlated ( ρ  = 0.84, P   ≪  0.01) with those in the broader Coral Sea (Fig. 1f ), including when the long-term warming trend is removed from both time series ( ρ  = 0.69, P  < 0.01; Supplementary Fig. 4 ). Based on the strength of this correlation, we associate high January–March area-averaged Coral Sea SSTAs with increased thermal bleaching risk in the GBR.

a – e , SSTAs (using ERSSTv5 data) for January–March in the Australasian region relative to the 1961–90 average for the five recent GBR mass coral bleaching years: 2016, 2017, 2020, 2022 and 2024. The black box shows the Coral Sea region (4° S–26° S, 142° E–174° E). f , Coral Sea and GBR mean SSTAs for 1900–2024 in January–March relative to the 1961–90 average. The black vertical lines indicate the five recent GBR mass coral bleaching years.

Record temperatures were set in 2016 and 2017 in the Coral Sea, and in 2020 they peaked fractionally below the record high of 2017. The January–March of 2022 was another warm event, the fifth warmest on record at the time. Recent data (ERSSTv5) indicate that 2024 set a new record by a margin of more than 0.19 °C above the previous record for the region. The January–March mean SSTs averaged over the five mass bleaching years during the period 2016–2024 are 0.77 °C higher than the 1961–90 January–March averages in both the Coral Sea and the GBR. The multidecadal warming trend, extreme years and association between GBR and Coral Sea SSTs are similar for the HadISST 28 gridded SST dataset, with some notable differences in the 1900–40 period (Supplementary Fig. 3 ). Furthermore, analysis of modern temperature-sensitive Sr/Ca series from GBR corals for 1900–2017 provides coherent independent evidence of statistically significant multi-decadal warming trends in January–March SSTs in the central and southern GBR (Supplementary Information section  4.2 ).

A multi-century context (1618–present)

Reconstructing Coral Sea January–March SSTs from 1618 to 1995 extends the century-long instrumental record back in time by an additional three centuries (Fig. 2a and Methods ). The reconstruction (calibrated to ERSSTv5) shows that multi-decadal SST variability was a persistent feature in the past. At the centennial timescale, there is relative stability before 1900, with the exception that cooler temperatures prevailed in the 1600s. Warming during the industrial era has been evident since the early 1900s (Fig. 2a ). There is a warming trend for January–March of 0.09 °C per decade for 1900–2024 and 0.12 °C per decade for 1960–2024 (Fig. 1f ) using ERSSTv5 data. Calibrating our reconstruction to HadISST1.1 yields similar results, with some differences in the degree of pre-1900 variability at both multi-decadal and centennial timescales (Supplementary Information section  5.2.6 ).

a , Reconstructed and observed mean January–March SSTAs in the Coral Sea for 1618–2024 relative to 1961–90. Dark blue, highest skill (maximum coefficient of efficiency) reconstruction with the full proxy network; light blue, 5th–95th-percentile reconstruction uncertainty; black, observed (ERSSTv5) data. Red crosses indicate the five recent mass bleaching events. Dashed lines indicate the best estimate (highest skill, red) and 95th-percentile (pink) uncertainty bound for the maximum pre-1900 January–March SSTA. b , Central GBR SSTA for the inner shelf 23 in thick orange and outer shelf 25 (Flinders Reef) in thin orange lines; these series are aligned here (see Methods ) with modern observations of mean GBR SSTAs for January–March relative to 1961–90. Observed data are shown at annual (grey line) and five-year (black line with open circles, plotted at the centre of each five-year period and temporally aligned with the five-year coral series 23 ) resolution. Dashed lines indicate best-estimate pre-1900 January–March maxima for refs. 23 (red) and 25 (pink). Orange shading indicates 5th–95th-percentile uncertainty bounds. Red crosses indicate the five recent mass bleaching events. c , Evaluation metrics for the Coral Sea reconstruction (Supplementary Information section  3.1 ); RE, reduction of error; CE, coefficient of efficiency; Rsq-cal, R-squared in the calibration period; Rsq-ver, R-squared in the verification (evaluation) period. d , Coral data locations relative to source data region (orange box) and Coral Sea region (red box). Coral proxy metadata are given in Supplementary Tables 1 and 2 .

Our best-estimate (highest skill; Methods ) annual-resolution Coral Sea reconstruction (Fig. 2a ), using the full coral network calibrated to the ERSSTv5 instrumental data, indicates that the January–March mean SSTAs in 2016, 2017, 2020, 2022 and 2024 were, respectively, 1.50 °C, 1.54 °C, 1.53 °C, 1.46 °C and 1.73 °C above the 1618–1899 (hereafter ‘pre-1900’) reconstructed average. Using the same best-estimate reconstruction, Coral Sea January–March SSTs during these GBR mass bleaching years were five of the six warmest years the region has experienced in the past 400 years (Fig. 2a ).

By comparing the recent warm events to the reconstruction’s uncertainty range ( Methods ), we quantify, using likelihood terminology consistent with recent reports from the Intergovernmental Panel on Climate Change 19 , that the recent heat extremes in 2017, 2020 and 2024 are ‘extremely likely’ (>95th percentile; Fig. 2a ) to be higher than any January–March in the period 1618–1899. Furthermore, the heat extremes in 2016 and 2022 are (at least) ‘very likely’ (>90th percentile) to be above the pre-1900 maximum. We perform a series of tests that verify that our findings are not simply an artefact of the nature of the coral network itself (Supplementary Information section 5.2 ). In a network perturbation test, we generate 22 subsets of the reconstruction by adding proxy records incrementally in order from the highest to the lowest correlation with the target (Supplementary Information section  5.2.5 ). We confirm that 2017, 2020 and 2024 were ‘extremely likely’ (>95th percentile) to have been warmer than any year pre-1900 (using ERSSTv5 data) for all of these proxy subsets. Furthermore, in 20 of the 22 subsets, 2016 was also ‘extremely likely’ (>95th percentile), rather than ‘very likely’, to be warmer (2022 was ‘extremely likely’ in 14 of the 22 subsets). All our additional tests, including a reconstruction with only Sr/Ca coral data (thereby omitting the possibility of any non-temperature signal in δ 18 O coral on the reconstruction), achieve high reconstruction skill and confirm the extraordinary nature of recent extreme temperatures in the multi-century context (Supplementary Information section  5.2 ). Analyses using HadISST1.1 generally show lower correlations with the coral data and reconstructions with slightly warmer regional SSTs before 1900, along with more-muted centennial and multi-decadal variability in the pre-instrumental period. Nevertheless, the HadISST1.1-calibrated reconstructions show that the recent thermal extremes are well above the best estimate (highest skill) of the pre-1900 maximum of reconstructed January–March SSTAs (Supplementary Fig. 42 ). Furthermore, lower SSTAs (in the HadISST1.1 data) relative to the previous three centuries (as in our reconstructions calibrated to HadISST1.1), coupled with the recently observed mass coral bleaching events, could indicate that long-lived corals have a greater sensitivity to warming than is currently recognized.

Reconstructed regional GBR SSTAs based on a five-year-resolution, multi-century coral δ 18 O record from the central inshore GBR 23 (Fig. 2b ) show similarly strong warming since 1900 but more multi-decadal-to-centennial variability than the Coral Sea reconstruction. Recent five-year mean January–March GBR SSTAs narrowly exceed the best estimate of the maximum pre-1900 five-year mean since the early 1600s (Fig. 2b ). The averages for the five-year periods centred on 2018 and 2022 exceed the pre-1900 maximum by 0.11 °C and 0.06 °C, respectively. Results are similar using the five-year-resolution Flinders Reef (central outer shelf) 24 record (Supplementary Fig. 24 ), although its interpretation is limited by the lack of uncertainty estimates available for that record. Our Coral Sea reconstruction incorporates an updated (annual resolution) record from Flinders Reef 25 , which indicates similar centennial trends (thin orange line in Fig. 2b ) and shows that the recent high January–March SSTA events have approached the estimated local pre-1900 maximum SSTA. Although contiguous multi-century cores from within the GBR are limited in their spatial extent, twentieth-century warming is evident in these records.

The extraordinary nature of the recent Coral Sea January–March SSTs in the context of the past 400 years is further illustrated by comparing the ranked temperature anomalies (Fig. 3 ) for the combined reconstructed and instrumental period from 1618–2024, incorporating reconstruction uncertainty ( Methods ). The mass coral bleaching years of 2016, 2017, 2020, 2022 and 2024, and the heat event of 2004, stand out as the warmest events across the whole 407-year record. The warmest three years (2024, 2017 and 2020) exceed the upper uncertainty bound (95th percentile) of the warmest reconstructed January–March in the pre-1900 period (pink (upper) dashed line in Fig. 3 ); 2016, 2004 and 2022 exceed the 90th percentile bound (red (lower) dashed line in Fig. 3 ). The warming trend is clear in the association between the ascending rank of the temperature anomalies and the year (shown as the colour of the filled circles in Fig. 3 ). Despite high interannual variability, 78 of the warmest 100 January–March periods between 1618 and 2024 occurred after 1900, and the 23 warmest all occur after 1900. The warmest 20 January–March periods all occur after 1950, coinciding with accelerated global warming.

Ranked January–March SSTAs for 1618–2024 relative to 1961–90 (coloured circles) from the best-estimate (highest skill, full coral network) reconstruction (1618–1899) and instrumental (ERSSTv5) data (1900–2024). The year is indicated by the colour of the filled circles. The 5th–95th-percentile uncertainty bounds of the pre-1900 reconstructed SSTAs are shown by small grey dots. The year labels indicate the warmest six years on record, five of which were mass coral bleaching years on the GBR. The pink (upper) dashed line indicates the 95th-percentile uncertainty bound of the maximum pre-1900 reconstructed SSTA; the red (lower) dashed line indicates the 90th-percentile limit.

Assessing anthropogenic influence

Using climate model simulations from the most recent (sixth) phase of the Coupled Model Intercomparison Project 29 (CMIP6), we assess the human influence on January–March SSTAs in the Coral Sea. The model simulations are from two experiments in the Detection and Attribution Model Intercomparison Project (DAMIP) 30 . The first set of simulations represents historical climate conditions, including both the natural and human influences on the climate system over the 1850–2014 period (‘historical’; red in Fig. 4 ). The second experiment is a counterfactual climate that spans the same period and uses the same models but includes only natural influences on the climate, omitting all human influences (‘historical-natural’; blue in Fig. 4 ). The historical experiment includes anthropogenic emissions of greenhouse gases and aerosols, stratospheric ozone changes and anthropogenic land-use changes; the historical-natural experiment does not. Variations in natural climate forcings, such as from volcanic eruptions and solar variability, are incorporated in both experiments. We include models that have a transient climate response (the global mean surface-temperature anomaly at the time of a doubling of atmospheric CO 2 concentration) in the range 1.4–2.2 °C, which is deemed ‘likely’ by the science community 31 ( Methods and Supplementary Information ).

Climate-model simulations of Coral Sea January–March SSTAs relative to the 1850–1900 average for the period 1850–2014, for models within the ‘likely’ range for their transient climate response 31 . The blue line (median) and light blue shading (5th–95th-percentile limits) are from the ‘historical-natural’ climate model simulations (no anthropogenic climate forcing); the red line and light red shading are from the ‘historical’ simulations (anthropogenic influences on the climate included) using the same set of climate models. The climate-model-derived time of emergence of anthropogenic climate change, shown by the grey and black vertical lines (1976 and 1997), is when the ratio of the climate change signal to the standard deviation of noise/variability 32 across model ensemble members first rises above 1 and 2, respectively. All models are represented equally in the model ensemble.

It is only with the incorporation of anthropogenic influences on the climate that the model simulations capture the modern-era warming of the Coral Sea January–March SSTA (Fig. 4 ). The median of the historical simulations has statistically significant warming trends of 0.05 °C, 0.10 °C and 0.15 °C per decade for the periods from 1900, 1950 and 1970 to 2014, respectively; the equivalent historical-natural trends are smaller in magnitude than ±0.01 °C per decade. To further explore the centennial-scale trends, we use a bootstrap ensemble ( Methods ) of the two sets of 165-year simulations from 1850–2014. We found that 100% of the historical bootstrap ensemble has statistically significant positive trends ( Methods ) for 1900–2014, but this value is 0% for the historical-natural ensemble. The observed (ERSSTv5) mean SSTA for 2016–2024 of 0.60 °C relative to 1961–90 is warmer than any nine-year sequence in the 7,095 simulated years in the historical-natural experiments from models with transient climate responses in the ‘likely’ range 31 .

We also use the simulations to estimate the time of emergence of the anthropogenic influence on January–March Coral Sea SSTAs above the natural background variability. The anthropogenic warming signal 32 increases from near zero in 1900 to around 0.5 standard deviations of the variability (‘noise’) in 1960. The climate change signal-to-noise ratio then increases rapidly from 1960 to 2014, exceeding 1.0 in 1976, 2.0 in 1997 and around 2.8 by 2014, the end of these simulations (Fig. 4 , Methods and Supplementary Fig. 50 ). Anthropogenic impacts on the climate are virtually certain to be the primary driver of this long-term warming in the Coral Sea.

Previously, our knowledge of the SST history of the GBR and the Coral Sea region has been highly dependent on instrumental observations, with the exception of the five-year-resolution multi-century coral Sr/Ca and U/Ca SST reconstructions from the two point locations in the central GBR 23 , 24 , an update at one of these locations 25 , seasonal resolution ‘floating’ (in time) chronologies from the GBR in the Holocene 33 , 34 and point SST estimates further back in time 35 . Thus, the context of recent warming trends in the Coral Sea and GBR and their relation to natural variability on decadal to centennial timescales is largely unknown without reconstructions such as the one we developed here.

Our coral proxy network is located mostly beyond the GBR, in the Coral Sea, and some series are located outside the Coral Sea region (Fig. 2d ). The selection of the Coral Sea as a study region allowed for a larger sample of contributing coral proxy data than exists for the GBR. However, coral bleaching on the GBR can be influenced by factors other than large-scale SST, including local oceanic and atmospheric dynamics that can modulate the occurrence and severity of thermal bleaching and mortality events 13 . Nonetheless, warming of seasonal SSTs over the larger Coral Sea region is likely to prime the background state and increase the likelihood of smaller spatio-temporal-scale heat anomalies. Furthermore, where we use only the five-year resolution series directly from the GBR to reconstruct GBR SSTAs, we draw similar conclusions about the long-term trajectory of SSTAs as for our full coral network (Fig. 2b and Supplementary Fig. 24 ). Furthermore, short modern coral series from within the GBR, analysed in this study, document a multi-decadal warming signal that is coherent with instrumental data (Supplementary Figs. 29 and 30 ). Nonetheless, additional high-resolution, multi-century, temperature-sensitive coral geochemical series from within the GBR would help unravel the local and remote ocean–atmosphere contributions to past bleaching events and reduce uncertainties.

The focus on the larger Coral Sea study region also takes advantage of the global modelling efforts of CMIP6. The large number of ensemble members available for CMIP6 means that greater climate model diversity, and therefore greater certainty in our attribution analysis, is possible compared with most single model analyses. There is also a methodological benefit in having high replication of the same experiments run with multiple climate models. However, coarse-resolution global-scale models do not accurately simulate smaller-scale processes, such as inshore currents and mesoscale eddies in the Coral Sea or the Gulf of Carpentaria, which probably affect local surface temperatures and variations in nutrient upwelling in the GBR 36 , 37 . Upwelling on the GBR is linked to the strength of the East Australian Current 16 , the southward branch of the South Pacific subtropical gyre. The CMIP-scale models we use do capture these gyre dynamics. The models show that the East Australian Current is expected to increase in strength as the climate continues to warm through this century 38 , and this may lead to more nutrient inputs that can exacerbate coral sensitivity to rising heat stress 39 , 40 . As well as focusing our model analysis on the larger Coral Sea region, we use a three-month time step. In doing so, we minimize the impact of model spatio-temporal resolution on our inferences about the role of anthropogenic greenhouse-gas emissions on the SST conditions that give rise to GBR mass bleaching.

Remaining uncertainties

We present analyses and interpretations that are as robust as possible given currently available data and methods. However, several sources of remaining uncertainty mean that future reconstructions of past Coral Sea and GBR SSTs could differ from those presented here. Although bias corrections are applied to observational SST datasets such as ERSST and HadISST, these datasets probably retain biases, especially for the period during and before 1945 (ref. 41 ), and these may not be fully accounted for in the uncertainty estimates 42 . Because our reconstructions are calibrated directly to these datasets, future observational-bias corrections are likely to improve proxy-based reconstructions.

Reconstructions of SST that use coral δ 18 O records may be susceptible to the influence of changes in the coral δ 18 O–SST relationship on time periods longer than the instrumental training period, along with non-SST changes in the δ 18 O of seawater, which can covary with salinity. As such, new coral records of temperature-sensitive trace-element ratios such as Sr/Ca, Li/Mg or U/Ca may prove influential in future efforts to distinguish between changes in past temperature and hydroclimate. Owing to the limited availability of multi-century coral data from within the GBR itself, the reconstructed low-frequency variability of GBR SSTs in recent centuries is likely to change as more temperature proxy data become available. It is also likely that new sub-annual resolution records would aid in removing potential signal damping or bias from our use of some annual-resolution records to reconstruct seasonal SSTAs.

Ecological consequences

With global warming of 0.8–1.1 °C above pre-industrial levels 19 there has been a marked increase in mass coral bleaching globally 43 . Even limiting global warming to the Paris Agreement’s ambitious 1.5 °C level would be likely to lead to the loss of 70–90% of corals that are on reefs today 44 . If all current international mitigation commitments are implemented, global mean surface temperature is still estimated to increase in the coming decades, with estimates varying between 1.9 °C (ref. 45 ) and 3.2 °C (ref. 46 ) above pre-industrial levels by the end of this century. Global warming above 2 °C would have disastrous consequences for coral ecosystems 19 , 44 and the hundreds of millions of people who currently depend on them.

Coral reefs of the future, if they can persist, are likely to have a different community structure to those in the recent past, probably one with much less diversity in coral species 4 . This is because mass bleaching events have a differential impact on different coral species. For example, fast-growing branching and tabulate corals are affected more than slower-growing massive species because they have different thermal tolerance 4 . The simplification of reef structures will have adverse impacts on the many thousands of species that rely on the complex three-dimensional structure of reefs 4 . Therefore, even with an ambitious long-term international mitigation goal, the ecological function 4 of the GBR is likely to deteriorate further 5 before it stabilizes.

Coral adaptation and acclimatization may be the only realistic prospect for the conservation of some parts of the GBR this century. However, although adaptation opportunities may be plausible to some extent 47 , they are no panacea because evolutionary changes to fundamental variables such as temperature take decades, if not centuries, to occur, especially in long-lived species such as reef-building corals 48 . There is currently no clear evidence of the real-time evolution of thermally tolerant corals 48 . Most rapid changes depend on a history of exposure to key genetic types and extremes, and there are limitations to genetic adaptation that prevent species-level adaptation to environments outside of their ecological and evolutionary history 19 . Model projections also indicate that rates of coral adaptation are too slow to keep pace with global warming 49 . In a rapidly warming world, the temperature conditions that give rise to mass coral bleaching events are likely to soon become commonplace. So, although we may see some resilience of coral to future marine heat events through acclimatization, thermal refugia are likely to be overwhelmed 50 . Global warming of more than 1.5 °C above pre-industrial levels will probably be catastrophic for coral reefs 44 .

Our new multi-century reconstruction illustrates the exceptional nature of ocean surface warming in the Coral Sea today and the resulting existential risk for the reef-building corals that are the backbone of the GBR. The reconstruction shows that SSTs were relatively cool and stable for hundreds of years, and that recent January–March ocean surface heat in the Coral Sea is unprecedented in at least the past 400 years. The coral colonies and reefs that have lived through the past several centuries, and that yielded the valuable Sr/Ca and δ 18 O data on which our reconstruction is based, are themselves under serious threat. Our analysis of climate-model simulations confirms that human influence is the driver of recent January–March Coral Sea surface warming. Together, the evidence presented in our study indicates that the GBR is in danger. Given this, it is conceivable that UNESCO may in the future reconsider its determination that the iconic GBR is not in danger. In the absence of rapid, coordinated and ambitious global action to combat climate change, we will likely be witness to the demise of one of Earth’s great natural wonders.

Instrumental observations

The Coral Sea and GBR area-averaged monthly SSTAs relative to 1961–90 for January–March are obtained from version 5 of the Extended Reconstructed Sea Surface Temperature dataset (ERSSTv5) 27 . We compare our results using ERSSTv5 with those generated using the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST1.1) 28 . We use only post-1900 instrumental SST observations here. Although gridded datasets have some coverage before 1900, ship-derived temperature data in the region for that period are too sparse to be reliable for calibrating our reconstruction (Supplementary Information section  1.2 ). The regional mean for the GBR is computed using the seven grid-cell locations used by the Australian Bureau of Meteorology (Supplementary Information section  1.1 ). We define the Coral Sea region as the ocean areas inside 4° S–26° S, 142° E–174° E.

Coral-derived temperature proxy data

We use a network of 22 published and publicly available sub-annual and annual resolution temperature-sensitive coral geochemical series (proxies; Fig. 2d , Supplementary Tables 1 and 2 , and Supplementary Fig. 5a–v ) from the western tropical Pacific in our source data region (4° N–27° S, 134° E–184° E) that cover at least the period from 1900 to 1995. Of these 22 series, 16 are δ 18 O, which are in per mil (‰) notation relative to Vienna PeeDee Belemnite (VPDB) 51 ; the remaining six are Sr/Ca series. The coral data are used as predictors in the reconstruction of January–March mean SSTAs in the Coral Sea region. We apply the inverse Rosenblatt transformation 52 , 53 to the coral data to ensure that our reconstruction predictors are normally distributed. Sub-annually resolved series are converted to the annual time step by averaging across the November–April window. This maximizes the detection of the summer peak values, allowing for some inaccuracy in sub-annual dating and the timing of coral skeleton deposition 54 , 55 . A small fraction (less than 0.8%) of missing data is infilled using the regularized expectation maximization (RegEM) algorithm 56 (Supplementary Information section  2.3 ), after which the proxy series are standardized such that each has a mean of zero and a standard deviation of one over their common 1900–1995 period.

Reconstruction method

To produce our Coral Sea reconstruction, we use nested principal component regression 57 (PCR), in which the principal components of the network of 22 coral proxies are used as regressors against the target-region January–March SSTA relative to the 1961–90 average. We perform the reconstructions separately for each nest of proxies, where a nest is a set of proxies that cover the same time period. The longest nest dates back to 1618, when at least two series are available. The nests allow for the use of all coral proxies over the full time period of their coverage. The 96-year portion of the instrumental period (1900–1995) that overlaps with the reconstruction period is used for calibration and evaluation (or equivalently, verification) against observations. We reconstruct regional SSTAs from the principal components of the coral network of δ 18 O and Sr/Ca data, rather than their local SST calibrations, to minimize the number of computational steps and to aid in representing the full reconstruction uncertainty.

Principal component analysis (PCA) is used to reduce the dimensionality of the proxy matrix, as follows. Let P ( t , r ) denote the palaeoclimate-data matrix during the time period t  = 1,..., n at an annual time step for proxy series r  = 1,..., p . PCA is undertaken on this matrix during the calibration period, P cal . We obtain the principal component coefficients matrix P coeff ( r , e ) for principal components e  = 1,..., n PC and principal component scores P score ( t , e ), which are representations of the input matrix P cal in the principal component space. P score is truncated to include n PC,use principal components to form \({P}_{{\rm{score}}}^{{\prime} }\) such that the variance of the proxy network explained by the n PC,use principal components is greater than \({\sigma }_{{\rm{expl}}}^{2}\) (which we set to 95%). Reconstruction tests in which \({\sigma }_{{\rm{expl}}}^{2}\) is varied from 70% to 95% show that our results are not strongly sensitive to this choice, and tests based on lag-one autoregressive noise for \({\sigma }_{{\rm{expl}}}^{2}\) from 50% to 99% further support this choice (Supplementary Information section  3.2 ). These principal components are used as predictors against which the Coral Sea January–March instrumental SSTAs are regressed. We regress the standardized SSTA target data during the calibration period, I cal , against the retained principal components of the predictor data, \({P}_{{\rm{score}}}^{{\prime} }\) :

Thus, we obtain n PC,use estimates of the regression coefficients γ e with gaussian error term ε t  ~  N (0, \({\sigma }_{N}^{2}\) ). The principal components are extended back into the pre-instrumental period by multiplying the entire proxy matrix P ( t , p ) with the truncated principal component coefficient matrix \({P}_{{\rm{coeff}}}^{{\prime} }\) ( t , e ) to obtain \({Q}_{{\rm{coeff}}}^{{\prime} }\) :

The reconstruction proceeds with the fitted regression coefficients γ e and extended coefficient matrix \({Q}_{{\rm{coeff}}}^{{\prime} }\) to obtain a reconstruction time series R m ( t ) for a given nest of proxy series

The standardized reconstruction R m ( t ) is then calibrated to the instrumental data such that the standard deviation and mean of the reconstruction and target during the calibration interval are equal. As well as obtaining reconstructions for each nest of available proxies, we compute stitched reconstructions S c ( t ) for each calibration period c , which include at each time step the reconstructed data for the proxy nest with maximum coefficient of efficiency 58 , 59 (Supplementary Information section  3.1 ). This procedure is performed for contiguous calibration intervals between 60 and 80 years duration between 1900 and 1995, with interval width and location increments of two years, reserving the remaining data in the overlapping period for independent evaluation, and for all proxy nests. The reconstruction error is modelled with a lag-one autoregressive process fitted to the residuals. We evaluate the capacity of our reconstruction method to achieve spurious skill from overfitting by performing a test in which we replace the coral data with synthetic noise (Supplementary Information section  3.2i ). We find that reconstructions based on synthetic noise achieve extremely low or zero skill and as more noise principal components are included in the regression, the evaluation metrics indicate declining skill. Our reconstruction and evaluation methods therefore guard against the potential for spurious skill.

Pseudo-proxy reconstructions

Our reconstruction method is further evaluated by using a pseudo-proxy modelling approach based on the Community Earth System Model (CESM) Last Millennium Experiment (LME) 60 , for which there are 13 full-forcing ensemble members covering the period 850–2005. We use the pseudo-proxy reconstructions to evaluate our reconstruction method and coral network in a fully coupled climate-model environment. We form pseudo-proxies by extracting from each LME ensemble member the SST and sea surface salinity (SSS) from the 1.5° × 1.5° grid cell located nearest to our coral data. We then apply proxy system models in the form of linear regression models, basing δ 18 O on both SST and SSS, and Sr/Ca on SST only (Supplementary Information section  3.3 ). We set the spatial and temporal availability of the pseudo-coral network to match that of the coral network. We then apply our PCR reconstruction and evaluation procedure to the pseudo-proxy network, taking advantage of the availability of the modelled Coral Sea SSTA data across the multi-century period of 1618–2005, which allows for the evaluation of the pseudo-proxy reconstruction over this entire time period. We first test our method using a ‘perfect proxy’ approach (with no proxy measurement error) before superimposing synthetic noise on the pseudo-proxy time series, evaluating our methodology at two separate levels of measurement error, quantified by signal-to-noise ratios of 1.0 and 4.0. The evaluation metrics for these tests indicate that our coral network and reconstruction method obtain skilful reconstructions of Coral Sea SSTAs in the climate-model environment (Supplementary Figs. 17b , 18 , 20b , 21 , 22b and 23 ).

Comparison with independent coral datasets

We use two multi-century five-year-resolution coral series from the central GBR 23 , 24 (Fig. 2b and Supplementary Fig. 24 ) and a network of sub-annual and annual resolution modern coral series (dated from 1900 onwards but not covering the full 1900–1995 period) from 44 sites in the GBR (Supplementary Information section  4.2 ) for independent evaluation of coral-derived evidence for warming in the region. We estimate five-year GBR SSTAs (Fig. 2b ) by aligning the post-1900 mean and variance of the proxy and instrumental (ERSSTv5) data.

Reconstruction sensitivity to non-SST influences

Of the 22 available coral series, 16 are records of δ 18 O, a widely used measure of the ratio of the stable isotopes 18 O and 16 O. In the tropical Pacific Ocean, δ 18 O is significantly correlated with SST 61 , 62 , 63 , 64 . Coral δ 18 O is also sensitive to the δ 18 O of seawater 65 , which can reflect advection of different water masses and/or changes in freshwater input, such as from riverine sources or precipitation, which in turn co-vary with SSS. Thus, it is generally considered that the main non-SST contributions to coral δ 18 O are processes that co-vary with SSS 62 , 66 . Our methodology minimizes the influence of non-temperature impacts on the reconstruction by exploiting the contrast in spatial heterogeneity between SST and SSS in January–March (Supplementary Information section  5.1 ). SSS is spatially inhomogeneous in the tropical Pacific 66 , 67 , leading to low coherence in SSS signals across our coral network. By contrast, the strong and coherent SST signal across our coral network locations and the Coral Sea region leads to principal components that are strongly representative of SST variations. This produces a skilful reconstruction of SST, as determined by evaluation against independent observations, and low correlations with SSS across the Coral Sea region (Supplementary Fig. 31 ).

Although the likelihood of non-SST influences on our SST reconstruction is low, we nonetheless test the sensitivity of our reconstruction and its associated interpretations to the possibility of these influences on the coral data. The tests compute the correlations between our best-estimate SSTA reconstruction (highest coefficient of efficiency) and observations of SSS, along with a series of additional reconstructions based on subsets of our coral network. The correlations between our highest coefficient of efficiency January–March Coral Sea SSTA reconstruction and January–March SSS are mapped for the Coral Sea and its neighbouring domain using three instrumental SSS datasets (Supplementary Fig. 31 ). Correlations are not statistically significant over most of the domain. Noting differing spatial correlation patterns between the instrumental SSS datasets 68 , which also cover different time periods (Supplementary Information section  5.1 ), we undertake six sensitivity tests using subsets of the coral network (Supplementary Information section  5.2 ). We use the following combinations of coral series: (1) the full network of 22 δ 18 O and Sr/Ca series (Figs. 2a and 3 ); (2) a subset of the six available Sr/Ca series (Supplementary Figs. 32 – 33 ), to test how the reconstruction is influenced by the inclusion of coral δ 18 O records; (3) a fixed nest subset of the five longest coral series, extending back to at least 1700 (Supplementary Figs. 34 – 35 ), to test for the potential influence of combining series of differing lengths (from our splicing of portions of the best reconstructions from each nest); (4) a subset of the ten coral series that are most strongly correlated with the target (Supplementary Figs. 36 and 37 ), to test how our reconstruction is influenced by the inclusion of coral series that are less strongly correlated with our target; (5) a subset of coral series that excludes the six records that are reported to potentially include biological mediation or non-climatic effects, or have low correlation with the target (Supplementary Figs. 38 and 39 ), to test their influence on the reconstruction; and (6) a network perturbation test comprising 22 separate subsets of proxies, in which proxy records are added incrementally in order of highest to lowest correlation with the target, starting with a single coral series and increasing the number of included proxies to all 22 series in our network (Supplementary Information section  5.2.5 ), to systematically quantify the influence of gradually including more coral datasets on our reconstruction and its interpretations.

The evaluation metrics (Fig. 2c and Supplementary Figs. 32b , 34b , 36b and 38b ) indicate a skilful reconstruction back to 1618 for the reconstructions based on the Full, Sr/Ca only, Long, Best-10 and OmitBioMed networks. These reconstructions explain 82.7%, 80.6%, 77.6%, 79.8% and 80.4% (R-squared values) of the variance in January–March SSTAs, respectively, in the independent evaluation periods (using ERSSTv5b). All coral subsets in the network perturbation test produce skilful reconstructions (Supplementary Fig. 40 ). The highest-skill reconstructions for all subsets in the network perturbation test align with our key interpretations (Supplementary Figs. 41 and 42 ). Together, our sensitivity tests show that the coral network, observational data and reconstruction methodology are a sound basis for reconstructing Coral Sea January–March SSTAs in past centuries and contextualizing recent high-SST events ( Supplementary Information ).

Climate-model attribution ensembles and experiments

The multi-model attribution analysis used here is based on simulations from CMIP6. We analyse simulations from the historical experiment (including natural and anthropogenic influences for 1850–2014) and the historical-natural experiment (natural-only forcings for 1850–2014). We select climate models for which monthly surface temperature is available in at least three historical and historical-natural simulations (Supplementary Table 5 ). All model simulations are interpolated to a common regular 1.5° × 1.5° latitude–longitude grid. January–March SSTAs relative to 1961–90 are calculated for each simulation. The full historical all-forcings ensemble is composed of 14 models with 268 simulations for 1850–2014. The natural-only ensemble is composed of the same 14 models with 95 individual simulations. A subset of climate models in the CMIP6 ensemble are considered by the science community to be ‘too hot’, simulating warming in response to increased atmospheric carbon dioxide concentrations that is larger than that supported by independent evidence 31 . We omit these models from our analysis by including only models with a transient climate response in the ‘likely’ range 31 of 1.4–2.2 °C. Our results are not strongly sensitive to this selection (Supplementary Information section  6.3 ). The ten remaining models yield a total of 25,410 years from 154 historical ensemble members and 7,095 years from 43 historical-natural ensemble members. We weight the models equally in our analysis using bootstrap sampling. We report linear trends based on simple linear regression models fitted with ordinary least squares. The statistical significance of linear trends is assessed using the Spearman’s rank correlation test 69 .

Time of emergence of the anthropogenic impact

We assess the anthropogenic influence on SSTAs in the Coral Sea region by starting with the assumption that any anthropogenic influence on SSTAs in the Coral Sea is indistinguishable from natural variability at the commencement of the model experiments. We measure the impact of anthropogenic influence on the climate in the region using a signal-to-noise approach 32 , 70 . We calculate the anthropogenic ‘signal’ as the mean of the difference between the smoothed (using a 41-year Lowess filter) modelled historical Coral Sea SSTA and the mean smoothed modelled historical-natural SSTA. Our ‘noise’ is the standard deviation of the difference between the modelled historical SSTA and its smoothed time series (Supplementary Information section  6 ).

Methods additionally rely on Supplementary Information and refs. 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 , 82 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 99 , 100 , 101 , 102 , 103 , 104 .

Data availability

The ERSSTv5 instrumental SST data are available from the US National Oceanic and Atmospheric Administration at https://psl.noaa.gov/data/gridded/data.noaa.ersst.v5.html . The HadISST1.1 data are available from the UK Met Office at https://www.metoffice.gov.uk/hadobs/hadisst/ . The original coral palaeoclimate data are available at the links provided in Supplementary Table 2 . Land areas for maps are obtained from the Mapping Toolbox v.23.2 in Matlab v.2023b and the Global Self-consistent, Hierarchical, High-resolution Geography (GSHHS) Database at https://www.soest.hawaii.edu/pwessel/gshhg/ through the m_map toolbox by R. Pawlowicz, available at https://www.eoas.ubc.ca/%7Erich/map.html . Prepared data from the coral geochemical series, reconstructions and climate models that support the findings of this study are available at: https://doi.org/10.24433/CO.4883292.v1 .

Code availability

The code that supports the findings of this study is available and can be run at : https://doi.org/10.24433/CO.4883292.v1 .

Australian Institute of Marine Science. Long-Term Monitoring Program. https://www.aims.gov.au/research-topics/monitoring-and-discovery/monitoring-great-barrier-reef/long-term-monitoring-program (2024).

Hughes, T. P. et al. Global warming and recurrent mass bleaching of corals. Nature 543 , 373–377 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Hoegh-Guldberg, O. Climate change, coral bleaching and the future of the world’s coral reefs. Mar. Freshw. Res. 50 , 839–866 (1999).

Google Scholar  

Hughes, T. P. et al. Global warming transforms coral reef assemblages. Nature 556 , 492–496 (2018).

Great Barrier Reef Marine Park Authority. Great Barrier Reef Outlook Report 2019 (Great Barrier Reef Marine Park Authority, 2019).

Hughes, T. P. et al. Coral reefs in the Anthropocene. Nature 546 , 82–90 (2017).

Davis, K. L., Colefax, A. P., Tucker, J. P., Kelaher, B. P. & Santos, I. R. Global coral reef ecosystems exhibit declining calcification and increasing primary productivity. Commun. Earth Environ. 2 , 105 (2021).

Article   ADS   Google Scholar  

Westcott, D. A. et al. Relative efficacy of three approaches to mitigate Crown-of-Thorns Starfish outbreaks on Australia’s Great Barrier Reef. Sci. Rep. 10 , 12594 (2020).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Mellin, C. et al. Spatial resilience of the Great Barrier Reef under cumulative disturbance impacts. Glob. Chang. Biol. 25 , 2431–2445 (2019).

Article   ADS   PubMed   Google Scholar  

Jackson, J. B. C. et al. Historical overfishing and the recent collapse of coastal ecosystems. Science 293 , 629–637 (2001).

Article   CAS   PubMed   Google Scholar  

Hoegh-Guldberg, O. & Smith, G. J. The effect of sudden changes in temperature, light and salinity on the population density and export of zooxanthellae from the reef corals Stylophora pistillata Esper and Seriatopora hystrix Dana. J. Exp. Mar. Biol. Ecol. 129 , 279–303 (1989).

Article   Google Scholar  

DeCarlo, T. M. et al. Acclimatization of massive reef-building corals to consecutive heatwaves. Proc. Biol. Sci. 286 , 20190235 (2019).

PubMed   PubMed Central   Google Scholar  

McGowan, H. & Theobald, A. ENSO weather and coral bleaching on the Great Barrier Reef, Australia. Geophys. Res. Lett. 44 , 10,601–10,607 (2017).

Zhao, W., Huang, Y., Siems, S. & Manton, M. The role of clouds in coral bleaching events over the Great Barrier Reef. Geophys. Res. Lett. 48 , e2021GL093936 (2021).

Oxley, W. G., Emslie, M., Muir, P. & Thompson, A. Marine Surveys Undertaken in the Lihou Reef National Nature Reserve (Australian Institute of Marine Science, 2004).

DeCarlo, T. M. & Harrison, H. B. An enigmatic decoupling between heat stress and coral bleaching on the Great Barrier Reef. PeerJ 7 , e7473 (2019).

Article   PubMed   PubMed Central   Google Scholar  

UNESCO World Heritage Committee. Extended 44th Session of the World Heritage Committee, Fuzhou (China) 16–31 July 2021 . Draft decision 44 COM 7B.90. https://whc.unesco.org/document/188005 (UNESCO, 2021).

UNESCO World Heritage Committee. Extended 45th Session of the World Heritage Committee, Riyadh (Saudi Arabia) 10–25 September 2023 . Decision 45 COM 7B.13. https://whc.unesco.org/document/199654 (UNESCO, 2023).

IPCC. Summary for Policymakers. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (eds Masson-Delmotte, V. et al.) (Cambridge Univ. Press, 2021).

Kamenos, N. A. & Hennige, S. J. Reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 5 , 283 (2018).

Hoegh-Guldberg, O. et al. Commentary: reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 6 , 86 (2019).

DeCarlo, T. M. Commentary: reconstructing four centuries of temperature-induced coral bleaching on the Great Barrier Reef. Front. Mar. Sci. 7 , 30 (2020).

Hendy, E. J. et al. Abrupt decrease in tropical Pacific sea surface salinity at end of Little Ice Age. Science 295 , 1511–1514 (2002).

Calvo, E. et al. Interdecadal climate variability in the Coral Sea since 1708 A.D. Palaeogeogr. Palaeoclimatol. Palaeoecol. 248 , 190–201 (2007).

Zinke, J. et al. North Flinders Reef (Coral Sea, Australia) Porites sp. corals as a candidate global boundary stratotype section and point for the Anthropocene series. Anthropocene Rev. 10 , 201–224 (2023).

Spady, B. L. et al. Global Coral Bleaching Database (NCEI Accession 0228498) (NOAA National Centers for Environmental Information, 2022); https://www.ncei.noaa.gov/archive/accession/0228498 .

Huang, B. et al. Extended reconstructed sea surface temperature, Version 5 (ERSSTv5): Upgrades, validations, and intercomparisons. J. Clim. 30 , 8179–8205 (2017).

Rayner, N. A. et al. Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res. Atmos. 108 , 4407 (2003).

Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9 , 1937–1958 (2016).

Gillett, N. P. et al. The Detection and Attribution Model Intercomparison Project (DAMIP v1.0) contribution to CMIP6. Geosci. Model Dev. 9 , 3685–3697 (2016).

Hausfather, Z., Marvel, K., Schmidt, G. A., Nielsen-Gammon, J. W. & Zelinka, M. Climate simulations: recognize the ‘hot model’ problem. Nature 605 , 26–29 (2022).

Hawkins, E. et al. Observed emergence of the climate change signal: from the familiar to the unknown. Geophys. Res. Lett. 47 , e2019GL086259 (2020).

Gagan, M. K. et al. Temperature and surface-ocean water balance of the mid-Holocene tropical western Pacific. Science 279 , 1014–1018 (1998).

Arzey, A. K. et al. Coral skeletal proxy records database for the Great Barrier Reef, Australia. Preprint at Earth Syst. Sci. Data https://doi.org/10.5194/essd-2024-159 (2024).

Brenner, L. D. et al. Coral record of Younger Dryas Chronozone warmth on the Great Barrier Reef. Paleoceanogr. Paleoclimatol. 35 , e2020PA003962 (2020).

Furnas, M. J. & Mitchell, A. W. Nutrient inputs into the central Great Barrier Reef (Australia) from subsurface intrusions of Coral Sea waters: a two-dimensional displacement model. Cont. Shelf Res. 16 , 1127–1148 (1996).

Wolanski, E., Andutta, F., Deleersnijder, E., Li, Y. & Thomas, C. J. The Gulf of Carpentaria heated Torres Strait and the Northern Great Barrier Reef during the 2016 mass coral bleaching event. Estuar. Coast. Shelf Sci. 194 , 172–181 (2017).

Oliver, E. C. J. & Holbrook, N. J. Extending our understanding of South Pacific gyre ‘spin-up’: modeling the East Australian Current in a future climate. J. Geophys. Res. Oceans 119 , 2788–2805 (2014).

DeCarlo, T. M. et al. Nutrient-supplying ocean currents modulate coral bleaching susceptibility. Sci. Adv. 6 , eabc5493 (2020).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Wiedenmann, J. et al. Nutrient enrichment can increase the susceptibility of reef corals to bleaching. Nat. Clim. Chang. 3 , 160–164 (2013).

Article   ADS   CAS   Google Scholar  

Chan, D. & Huybers, P. Correcting observational biases in sea surface temperature observations removes anomalous warmth during World War II. J. Clim. 34 , 4585–4602 (2021).

Chan, D., Kent, E. C., Berry, D. I. & Huybers, P. Correcting datasets leads to more homogeneous early-twentieth-century sea surface warming. Nature 571 , 393–397 (2019).

Hughes, T. P. et al. Spatial and temporal patterns of mass bleaching of corals in the Anthropocene. Science 359 , 80–83 (2018).

Hoegh-Guldberg, O. et al. Chapter 3: Impacts of 1.5°C global warming on natural and human systems. In Global Warming of 1.5°C (eds Masson-Delmotte, V. et al.) (IPCC, 2018).

Meinshausen, M. et al. Realization of Paris Agreement pledges may limit warming just below 2 °C. Nature 604 , 304–309 (2022).

Matthews, H. D. & Wynes, S. Current global efforts are insufficient to limit warming to 1.5 °C. Science 376 , 1404–1409 (2022).

Coles, S. L. et al. Evidence of acclimatization or adaptation in Hawaiian corals to higher ocean temperatures. PeerJ 6 , e5347 (2018).

Hughes, T. P., Baird, A. H., Morrison, T. H. & Torda, G. Principles for coral reef restoration in the anthropocene. One Earth 6 , 656–665 (2023).

Logan, C. A., Dunne, J. P., Ryan, J. S., Baskett, M. L. & Donner, S. D. Quantifying global potential for coral evolutionary response to climate change. Nat. Clim. Chang. 11 , 537–542 (2021).

Dixon, A. M., Forster, P. M., Heron, S. F., Stoner, A. M. K. & Beger, M. Future loss of local-scale thermal refugia in coral reef ecosystems. PLOS Clim. 1 , e0000004 (2022).

Coplen, T. B. Discontinuance of SMOW and PDB. Nature 375 , 285 (1995).

van Albada, S. J. & Robinson, P. A. Transformation of arbitrary distributions to the normal distribution with application to EEG test-retest reliability. J. Neurosci. Methods 161 , 205–211 (2007).

Article   PubMed   Google Scholar  

Emile-Geay, J. & Tingley, M. Inferring climate variability from nonlinear proxies: application to palaeo-ENSO studies. Clim. Past 12 , 31–50 (2016).

Barnes, D. J., Taylor, R. B. & Lough, J. M. On the inclusion of trace materials into massive coral skeletons. Part II: distortions in skeletal records of annual climate cycles due to growth processes. J. Exp. Mar. Biol. Ecol. 194 , 251–275 (1995).

Article   CAS   Google Scholar  

Gagan, M. K., Dunbar, G. B. & Suzuki, A. The effect of skeletal mass accumulation in Porites on coral Sr/Ca and δ 18 O paleothermometry. Paleoceanogr. Paleoclimatol. 27 , PA1203 (2012).

ADS   Google Scholar  

Schneider, T. Analysis of incomplete climate data: estimation of mean values and covariance matrices and imputation of missing values. J. Clim. 14 , 853–871 (2001).

PAGES 2k Consortium. Consistent multidecadal variability in global temperature reconstructions and simulations over the Common Era. Nat. Geosci. 12 , 643–649 (2019).

Cook, E. R., Briffa, K. R. & Jones, P. D. Spatial regression methods in dendroclimatology: a review and comparison of two techniques. Int. J. Climatol. 14 , 379–402 (1994).

Nash, J. E. & Sutcliffe, J. V. River flow forecasting through conceptual models part I − a discussion of principles. J. Hydrol. 10 , 282–290 (1970).

Otto-Bliesner, B. L. et al. Climate variability and change since 850 CE: an ensemble approach with the Community Earth System Model. Bull. Am. Meteorol. Soc. 97 , 735–754 (2016).

Evans, M. N., Kaplan, A. & Cane, M. A. Optimal sites for coral-based reconstruction of global sea surface temperature. Paleoceanogr. Paleoclimatol. 13 , 502–516 (1998).

Russon, T., Tudhope, A. W., Hegerl, G. C., Collins, M. & Tindall, J. Inter-annual tropical Pacific climate variability in an isotope-enabled CGCM: Implications for interpreting coral stable oxygen isotope records of ENSO. Clim. Past 9 , 1543–1557 (2013).

PAGES Hydro2k Consortium. Comparing proxy and model estimates of hydroclimate variability and change over the Common Era. Clim. Past 13 , 1851–1900 (2017).

Freund, M. B. et al. Higher frequency of Central Pacific El Niño events in recent decades relative to past centuries. Nat. Geosci. 12 , 450–455 (2019).

Gagan, M. K. et al. New views of tropical paleoclimates from corals. Quat. Sci. Rev. 19 , 45–64 (2000).

Thompson, D. M., Ault, T. R., Evans, M. N., Cole, J. E. & Emile-Geay, J. Comparison of observed and simulated tropical climate trends using a forward model of coral δ 18 O. Geophys. Res. Lett. 38 , L14706 (2011).

LeGrande, A. N. & Schmidt, G. A. Global gridded data set of the oxygen isotopic composition in seawater. Geophys. Res. Lett. 33 , L12604 (2006).

Reed, E. V., Thompson, D. M. & Anchukaitis, K. J. Coral-based sea surface salinity reconstructions and the role of observational uncertainties in inferred variability and trends. Paleoceanogr. Paleoclimatol. 37 , e2021PA004371 (2022).

Khaliq, M. N., Ouarda, T. B. M. J., Gachon, P., Sushama, L. & St-Hilaire, A. Identification of hydrological trends in the presence of serial and cross correlations: A review of selected methods and their application to annual flow regimes of Canadian rivers. J. Hydrol. 368 , 117–130 (2009).

Mahlstein, I., Hegerl, G. & Solomon, S. Emerging local warming signals in observational data. Geophys. Res. Lett. 39 , L21711 (2012).

Freeman, E. et al. ICOADS Release 3.0: a major update to the historical marine climate record. Int. J. Climatol. 37 , 2211–2232 (2017).

Huang, B. et al. Uncertainty estimates for sea surface temperature and land surface air temperature in NOAAGlobalTemp version 5. J. Clim. 33 , 1351–1379 (2020).

Druffel, E. R. M. & Griffin, S. Variability of surface ocean radiocarbon and stable isotopes in the southwestern Pacific. J. Geophys. Res. 104 , 23607–23613 (1999).

DeLong, K. L., Quinn, T. M., Taylor, F. W., Lin, K. & Shen, C.-C. Sea surface temperature variability in the southwest tropical Pacific since AD 1649. Nat. Clim. Change 2 , 799–804 (2012).

Quinn, T. et al. A multicentury stable isotope record from a New Caledonia coral: Interannual and decadal SST variability in the southwest Pacific since 1657. Paleoceanography 13 , 412–426 (1998).

Quinn, T. M., Crowley, T. J. & Taylor, F. W. New stable isotope results from a 173-year coral from Espiritu Santo, Vanuatu. Geophys. Res. Lett. 23 , 3413–3416 (1996).

Alibert, C. & Kinsley, L. A 170-year Sr/Ca and Ba/Ca coral record from the western Pacific warm pool: 1. What can we learn from an unusual coral record? J. Geophys. Res. Oceans 113 , C04008 (2008).

Tudhope, A. W. et al. Variability in the El Niño-Southern Oscillation through a glacial-interglacial cycle. Science 291 , 1511–1517 (2001).

Urban, F. E., Cole, J. E. & Overpeck, J. T. Influence of mean climate change on climate variability from a 155-year tropical Pacific coral record. Nature 407 , 989–993 (2000).

Guilderson, T. P. & Schrag, D. P. Reliability of coral isotope records from the western Pacific warm pool: A comparison using age-optimized records. Paleoceanography 14 , 457–464 (1999).

Quinn, T. M., Taylor, F. W. & Crowley, T. J. Coral-based climate variability in the Western Pacific Warm Pool since 1867. J. Geophys. Res. 111 , C11006 (2006).

Gorman, M. K. et al. A coral-based reconstruction of sea surface salinity at Sabine Bank, Vanuatu from 1842 to 2007 CE. Paleoceanography 27 , PA3226 (2012).

Bagnato, S., Linsley, B. K., Howe, S. S., Wellington, G. M. & Salinger, J. Coral oxygen isotope records of interdecadal climate variations in the South Pacific Convergence Zone region. Geochem. Geophys. Geosyst. 6 , Q06001 (2005).

Linsley, B. K. et al. Tracking the extent of the South Pacific Convergence Zone since the early 1600s. Geochem. Geophys. Geosyst. 7 , Q05003 (2006).

Cole, J. E., Fairbanks, R. G. & Shen, G. T. Recent variability in the Southern Oscillation: Isotopic results from a Tarawa Atoll coral. Science 260 , 1790–1793 (1993).

Dassié, E. P. et al. A Fiji multi-coral δ 18 O composite approach to obtaining a more accurate reconstruction of the last two-centuries of the ocean-climate variability in the South Pacific Convergence Zone region. Paleoceanography 29 , 1196–1213 (2014).

Carton, J. A., Chepurin, G. A. & Chen, L. SODA3: A new ocean climate reanalysis. J. Clim. 31 , 6967–6983 (2018).

Zuo, H., Balmaseda, M. A., Tietsche, S., Mogensen, K. & Mayer, M. The ECMWF operational ensemble reanalysis-analysis system for ocean and sea ice: A description of the system and assessment. Ocean Sci. 15 , 779–808 (2019).

Cheng, L. et al. Improved estimates of changes in upper ocean salinity and the hydrological cycle. J. Clim. 33 , 10357–10381 (2020).

Thompson, D. M. et al. Identifying hydro‐sensitive coral δ18O records for improved high‐resolution temperature and salinity reconstructions. Geophys. Res. Lett. 49 , e2021GL096153 (2022).

Wu, Y., Fallon, S. J., Cantin, N. E. & Lough, J. M. Assessing multiproxy approaches (Sr/Ca, U/Ca, Li/Mg, and B/Mg) to reconstruct sea surface temperature from coral skeletons throughout the Great Barrier Reef. Sci. Total Environ. 786 , 147393 (2021).

Sadler, J., Webb, G. E., Leonard, N. D., Nothdurft, L. D. & Clark, T. R. Reef core insights into mid-Holocene water temperatures of the southern Great Barrier Reef. Paleoceanography 31 , 1395–1408 (2016).

Roche, R. C. et al. Mid-Holocene sea surface conditions and riverine influence on the inshore Great Barrier Reef. Holocene 24 , 885–897 (2014).

Reed, E. V., Cole, J. E., Lough, J. M., Thompson, D. & Cantin, N. E. Linking climate variability and growth in coral skeletal records from the Great Barrier Reef. Coral Reefs 38 , 29–43 (2019).

Razak, T. B. et al. Use of skeletal Sr/Ca ratios to determine growth patterns in a branching coral Isopora palifera. Mar. Biol. 164 , 96 (2017).

Marshall, J. F. Decadal-scale, High Resolution Records of Sea Surface Temperature in the Eastern Indian and South Western Pacific Oceans from Proxy Records of the Strontium/calcium Ratio of Massive Porites Corals PhD thesis, Australian National Univ. (2000).

Marshall, J. F. & McCulloch, M. T. An assessment of the Sr/Ca ratio in shallow water hermatypic corals as a proxy for sea surface temperature. Geochim. Cosmochim. Acta 66 , 3263–3280 (2002).

Gagan, M. K. et al. Coral oxygen isotope evidence for recent groundwater fluxes to the Australian Great Barrier Reef. Geophys. Res. Lett. 29 , 43-1–43-4 (2002).

D’Olivo, J. P., Sinclair, D. J., Rankenburg, K. & McCulloch, M. T. A universal multi-trace element calibration for reconstructing sea surface temperatures from long-lived Porites corals: Removing ‘vital-effects’. Geochim. Cosmochim. Acta 239 , 109–135 (2018).

Fallon, S. J., McCulloch, M. T. & Alibert, C. Examining water temperature proxies in Porites corals from the Great Barrier Reef: a cross-shelf comparison. Coral Reefs 22 , 389–404 (2003).

Brenner, L. D., Linsley, B. K. & Potts, D. C. A modern Sr/Ca-δ 18 O-sea surface temperature calibration for Isopora corals on the Great Barrier Reef. Paleoceanography 32 , 182–194 (2017).

Alibert, C. et al. Source of trace element variability in Great Barrier Reef corals affected by the Burdekin flood plumes. Geochim. Cosmochim. Acta 67 , 231–246 (2003).

Murty, S. A. et al. Spatial and temporal robustness of Sr/Ca-SST calibrations in Red Sea corals: Evidence for influence of mean annual temperature on calibration slopes. Paleoceanogr. Paleoclimatol. 33 , 443–456 (2018).

Sayani, H. R., Cobb, K. M., DeLong, K., Hitt, N. T. & Druffel, E. R. M. Intercolony δ 18 O and Sr/Ca variability among Porites spp. corals at Palmyra Atoll: Toward more robust coral-based estimates of climate. Geochem. Geophys. Geosyst. 20 , 5270–5284 (2019).

Otto, F. E. L. Geert Jan van Oldenborgh 1961–2021. Nat. Clim. Chang. 11 , 1017 (2021).

Download references

Acknowledgements

We acknowledge the originators of the coral data cited in Supplementary Tables 1 and 2 ; S. E. Perkins-Kirkpatrick and the deceased G. J. van Oldenborgh 105 for contributions to an earlier version of this manuscript; E. P. Dassié and J. Zinke for discussions and data; R. Neukom for advice on an earlier version of the reconstruction code; and B. Trewin and K. Braganza for advice about the Bureau of Meteorology GBR SST time series. B.J.H. and H.V.M. acknowledge support from an Australian Research Council (ARC) SRIEAS grant, Securing Antarctica’s Environmental Future (SR200100005), and ARC Discovery Project DP200100206. A.D.K. acknowledges support from an ARC DECRA (DE180100638) and the Australian government’s National Environmental Science Program. B.J.H. and A.D.K. acknowledge an affiliation with the ARC Centre of Excellence for Climate Extremes (CE170100023). H.V.M. acknowledges support from an ARC Future Fellowship (FT140100286). A.K.A. acknowledges support from an Australian government research training program scholarship and an AINSE postgraduate research award. Funding was provided to B.K.L. by the Vetlesen Foundation through a gift to the Lamont-Doherty Earth Observatory. Grants to B.K.L. enabled the generation of coral oxygen isotope and Sr/Ca data from Fiji that were used in our reconstruction (US National Science Foundation OCE-0318296 and ATM-9901649 and US National Oceanic and Atmospheric Administration NA96GP0406). We acknowledge the support of the NCI facility in Australia and the World Climate Research Programme’s working group on coupled modelling, which is responsible for CMIP. We thank the climate-modelling groups for producing and making available their model output. For CMIP, the US Department of Energy’s Program for Climate Model Diagnosis and Intercomparison provided coordinating support and led the development of software infrastructure in partnership with the Global Organisation for Earth System Science Portals.

Author information

Authors and affiliations.

Environmental Futures, School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong, New South Wales, Australia

Benjamin J. Henley, Helen V. McGregor & Ariella K. Arzey

Securing Antarctica’s Environmental Future, University of Wollongong, Wollongong, New South Wales, Australia

School of Agriculture, Food and Ecosystem Sciences, University of Melbourne, Parkville, Victoria, Australia

Benjamin J. Henley

School of Geography, Earth and Atmospheric Sciences, University of Melbourne, Parkville, Victoria, Australia

Andrew D. King & David J. Karoly

ARC Centre of Excellence for Climate Extremes, University of Melbourne, Parkville, Victoria, Australia

Andrew D. King

School of the Environment, The University of Queensland, Brisbane, Queensland, Australia

Ove Hoegh-Guldberg

Australian Institute of Marine Science, Townsville, Queensland, Australia

Janice M. Lough

ARC Centre of Excellence for Coral Reef Studies and School of Earth Sciences, University of Western Australia, Crawley, Western Australia, Australia

Thomas M. DeCarlo

Department of Earth and Environmental Sciences, Tulane University, New Orleans, LA, USA

Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY, USA

Braddock K. Linsley

You can also search for this author in PubMed   Google Scholar

Contributions

B.J.H., H.V.M. and A.D.K. conceived the study and developed the methodology. B.J.H. did most of the analysis. A.K.A. contributed analysis of modern coral data (Supplementary Information section  4.2 ). T.M.D. contributed analysis of instrumental data coverage (Supplementary Information section  1.2 ). B.K.L. contributed sub-annual coral data. B.J.H. and H.V.M. led the preparation of the manuscript, with contributions from A.D.K., O.H.-G., A.K.A., D.J.K., J.M.L., T.M.D. and B.K.L. Generative artificial intelligence was not used in any aspect of this study or manuscript.

Corresponding author

Correspondence to Benjamin J. Henley .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature thanks Simon Michel, Miriam Pfeiffer, Claudia Tebaldi and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information ., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Henley, B.J., McGregor, H.V., King, A.D. et al. Highest ocean heat in four centuries places Great Barrier Reef in danger. Nature 632 , 320–326 (2024). https://doi.org/10.1038/s41586-024-07672-x

Download citation

Received : 02 November 2022

Accepted : 04 June 2024

Published : 07 August 2024

Issue Date : 08 August 2024

DOI : https://doi.org/10.1038/s41586-024-07672-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Great barrier reef’s temperature soars to 400-year high.

• Jeff Tollefson

Nature (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.