Tracking South Pacific Coral Reef Bleaching
by Satellite and Field Observations

T. J. Goreau, R. L. Hayes, and A. E. Strong

Global Coral Reef Alliance, 37 Pleasant Street, Cambridge MA 02139 USA

Howard University, Washington, DC, USA 20059

NOAA and US Naval Academy, Annapolis, MD, USA 21402

Proceedings of the 8th International Coral Reef Symposium 2:1491-1494. 1997

NOTE:  This paper shows how satellite data showing anomalously hot sea surface temperatures can be used in real time to identify sites of coral bleaching for field studies of the ecological impacts of bleaching and its relationship to environmental variables.


South Pacific waters with anomalously high surface temperature were tracked by satellite to identify potential sites for field study of coral reef bleaching. Areas with warm season monthly anomalies exceeding +0.9 degrees C were verified to have coral bleaching by local observers, while other areas were not affected. Comparison of 15 environmental variables, measured at 19 field sites across the area affected in 1994, shows that bleaching correlates significantly only with anomalously high temperature. Live coral cover was inversely correlated with many human population density-dependent stresses, but these were not correlated to bleaching. Observations in the Indian and Atlantic Oceans also show that coral reefs worldwide are acclimated close to their upper temperature limits and probably unable to adapt rapidly to a +1 degree C anomalous warming during the warm season.


Mass coral bleaching over vast tracts of coastal ocean, distinguished from local bleaching confined to limited areas, was first observed in the 1980s (Jokiel and Coles 1990; Williams and Bunkley-Williams, 1990; Glynn 1993; Goreau and Hayes, 1994). Detailed analysis of satellite-borne radiometer measurements of sea surface temperature (SST), calibrated against in-situ measurements (Goreau et al., 1993; Goreau and Hayes, 1995a), showed that mass bleaching events took place whenever mean monthly temperatures exceeded a local threshold value, which was higher at the warmest sites (Goreau et al., 1993). Despite differences in the absolute temperatures preceding bleaching, local SST at each site corresponded to an anomalous warming of at least +1.0 degrees Celsius above the long term average in the warmest month at each of seven Caribbean sites: Puerto Rico, Jamaica, Cayman Islands, Cozumel, the Florida Keys, Bahamas, and Bermuda (Goreau et al., 1993). Examination of global SST anomalies since the start of the satellite record in 1983 showed that every major mass bleaching event known from 1983 through 1991 was preceded by a +1.0 degree positive anomaly in the warmest month (Goreau and Haves, 1994). 

If coral bleaching is triggered by high temperatures, it implies that coral reefs are under threat from currently existing climate anomalies, and unable to withstand further global warming without severe ecosystem disruption. Stabilization of the earth's atmosphere-ocean-biosphere-climate system, including critical climatically-sensitive, biodiverse, highly productive, and economically important ecosystems such as coral reefs is the goal of the Framework Convention on Climate Change (1992). Detailed monitoring of climatically sensitive ecosystems is essential to determine whether the Convention is capable of meeting its goals, or if more stringent limitations on emissions of greenhouse gases are required, but such monitoring was omitted in the final treaty draft. 

The purposes of the current study were 1) to determine whether SST anomalies could be used in near real time to estimate the distribution of bleaching episodes, and 2) to compare field observations of coral bleaching, SST anomalies, and other local environmental factors which might also contribute to bleaching.  


Sea surface temperatures from NOAA's global satellite-borne AVHRR radiometer database (Strong et al., 1996; Gleeson and Strong, 1994; Montgomery and Strong, 1994) were tracked on time scales ranging from instantaneous values to monthly averages. These were used to identify specific reef sites worldwide which were subjected to anomalies both above and below the estimated threshold values in the warmest month. In 1994 anomalies with magnitudes and local seasonal timing thought sufficient to trigger bleaching were observed in the South Pacific, South Atlantic, and Western Indian Ocean. While direct field work was conducted only in the core of the South Pacific area affected, contact was made with coral reef researchers and divers inside and outside of affected areas. Satellite-derived SSTs and SST anomalies were tracked at 33 locations worldwide, consisting of 22 Pacific, 6 Indian and 5 Atlantic sites (Table 1). 

We surveyed coral reefs at 19 sites in the South Pacific bleaching area, including fringing, lagoonal, and outer slope habitats, on both high islands and atolls, and at sites near to, as well as remote from, human populations. All common corals and algae seen were identified at least to genus level (Veron, 1986, Magruder and Hunt, 1979). At each site 15 local environmental variables were measured or estimated (wave exposure, temperature, salinity, dissolved oxygen, pH, live coral cover, percentage of bleached coral colonies, fleshy algae cover, calcareous algae cover, local human population density, terrestrial mud, Acanthaster abundance, the number of years since the last cyclone, number of previous mass bleaching events, and pollution). Interactions among all variables at all sites were analyzed by non-parametric statistics, and their significance determined according to Spearman rank order correlation coefficients. Line transects were not used due to the small area covered by such methods and the high variability shown by such measurements. Instead as much bottom as possible was covered by swimming belt transects along and across major local ecotone gradients, and by using photographic documentation with both close-up and wide-angle lenses (Goreau and Hayes, l995b).  

Table 1. Geographic site coordinates to the nearest half degree. Sites visited are designated by an asterisk (*). 

LOCATION                                                                                LATITUDE                    LONGITUDE


Mangareva, Gambier Islands                                           23.5S                            135WMururoa, Tuamotu Islands                                                  22S                              139WRangiroa, Tuamotu Islands *                                               15S                              148W

Tahiti-Moorea, Society Islands *                                      18S                              149W

Tubuai, Austral Islands                                                   23S                              149W

Rurutu, Astral Islands                                                     22S                              151W

Rarotonga, Cook Islands *                                                          22S                              161W

Aitutaki, Cook Islands *                                                  18S                              161W

Manihiki, Cook Islands                                                   11S                              162W

Niue                                                                                          19S                              170W

Tutuila, American Samoa*                                                          14.5S                            171W

Funafuti, Tuvalu                                                              8.5S                             179E

Tokelau                                                                         9S                                172W

Uvea, Wallis and Futuna                                                 13.5S                            176.5E

Vavau, Tonga                                                                 19S                              174W

Lau, Fiji                                                                         18S                              179W

Canton, Kiribati                                                              1S                                172.5W

Christmas, Kiribati                                                                     3N                                156W

Tarawa, Kiribati                                                              2.5N                             172E

Majuro, Marshall Islands                                                 7N                                171E

Okinawa, Ryukyu Islands                                                           26N                              128E

Maui, Hawaii                                                                              21N                              156W



Malindi, Kenya                                                               3S                                40E                  Zanzibar, Tanzania                                                                     6S                                40E

Mayotte, Comoros                                                                     12.5S                            43E

Mauritius                                                                                   20.5S                            58E

Mahe, Seychelles                                                                      5S                                55.5E

Male, Maldives                                                               4N                                73E



Ihla do fernando de Noronha, Brazil                                  3S                                33WRecife, Pernambuco, Brazil                                      8S                                35W

Ilhas dos Abrolhos, Bahia, Brazil                                     18S                              38W

Arraial do Cabo, Rio de Janiero, Brazil                             23S                              41W

Ilha do Sao Sebastiao, Sao Paulo, Brazil                         24S                              46W


Fig. 1. SST anomaly maps for the eastern and central South Pacific Ocean from January (A) through June (E) of 1994. Each field covers 20N to 40S latitude and 60W to 180W longitude. Land masses are in black; >+1.0C anomalies are in dark gray; +0.1-0.9C anomalies are in medium gray; and negative to 0C anomalies are in light gray/white. January 1995 is represented by the upper left field (A); February 1995, by the left middle field (B); March 1995, by the lower left field (C); April 1995, by the upper right field (D); May 1995, by the right middle field (E); and June 1995, by the lower right field (E). Sites visited for field observations in this survey were: (1) Rangiroa, Tuamotu Islands, (2) Tahiti-Moorea, Society Islands, (3) Rarotonga, Cook Islands, (4) Aitutaki, Cook Islands, and (5) Tutuila, American Samoa.  


The data suggest that positive warm season sea surface temperature anomalies are the major factor that triggers bleaching. Reefs within the +1 C anomaly zone were all affected by mass bleaching, regardless of the presence or absence of other local stresses, and bleaching was not reported outside the anomalously hot area. Figure 1 shows the distribution of monthly thermal anomalies across the Pacific from January through June of 1994. Reefs where bleaching did not take place had warmest month anomalies less than +0.9 degrees Celsius, or had anomalies above this value only in cooler seasons. Bleaching correlated much more closely with thermal anomalies (Figure 2) than with absolute temperatures (Figure 3). Bleaching was positively correlated only with temperature, and with no other factor (Figure 4). In sharp contrast, live coral cover was negatively correlated with a wide variety of variables (Figure 5). Variables negatively affecting live coral cover were highly positively correlated with each other, and were all proportional to human population density dependent stresses. Although a large variety of variables had a negative effect on coral abundance, none of these had any effect on coral bleaching. While other factors do not appear to play a role in triggering bleaching, the presence of additional stresses such as algae overgrowth of corals or high terrestrial sedimentation appeared to increase mortality and retard recovery (Goreau and Hayes, 1995b). 

At most sites where satellite-derived monthly mean SSTs have been calibrated against in-situ measurements, they have been found to be accurate to about + 0.1 to 0.2 degrees Celsius, but tend to under-estimate in-situ values under hot and calm conditions (Goreau and Hayes 1994; Goreau and Hayes, 1995a). The anomaly threshold during the warm season appears to be similar in all oceans. However a few cases of bleaching were reported where measured monthly satellite anomalies were between +0.6 and +0.8 degrees Celsius. All of those sites were near strong SST spatial gradients due to cool offshore currents and proximity to coastal mountain ranges which induce high levels of cloudiness and rainfall along the continental Brazilian coastline which could significantly bias satellite data, especially when averaged over large scales. In-situ data showed that actual ocean temperatures in those areas were higher than estimated by satellite and were in fact above the anomaly threshold (Migotto, 1995). In shallow waters around coral reefs the actual in-situ anomaly would be expected to differ from that measured for offshore surface waters by satellite radiometer due to local water and air movements, but the average offset should be relatively constant.  

Fig. 2:    Coral reef bleaching reports and maximum SST anomalies. 

Fig. 3:   Coral reef bleaching reports and maximum sea surface temperatures. 

Figure 4: Significant positive correlations between field variables based on Spearman rank coefficients. The denser the arrow connecting variables, the more statistically significant the relationship: P <0.001 is represented by the thickest arrows; P <0.01, by the thinner arrows; and P <0.05 by the thinnest arrows. 

Figure 5: Significant negative correlations between field variables based on Spearman rank coefficients. The denser the arrow connecting variables, the more statistically significant the relationship: P <0.001 is represented by the thickest arrows; P <0.01, by the thinner arrows; and P <0.05 by the thinnest arrows.  


Our results indicate that significant information about timing, spatial distribution, and intensity of mass bleaching can be obtained from the distribution and magnitude of thermal anomalies, rather than from the absolute temperatures themselves. While satellite data is available averaged on a variety of time scales, we have found that monthly average data is sufficient to identify major mass bleaching episodes. Extremely intense short duration events, which are not recorded in monthly average data, might also trigger bleaching. Instantaneous SST data is subject to instrumental noise, fine scale spatial variability, and local cloud contamination, and needs averaging to produce statistically robust data. Furthermore the bulk SST in open ocean waters cannot vary rapidly on short time scales because of the high heat capacity of water. Local rather than mass bleaching can also occur in back reef and inshore environments due to local warming caused by restricted lagoonal circulation or by freshwater runoff (Goreau, 1964), and may not be identified by satellite data. Local bleaching from high temperatures or freshwater flooding has been known since 1918, but mass bleaching appears to be exclusively triggered by high temperatures and was unknown before the 1980s, when global temperatures rose markedly above that of previous decades (IPCC, 1996). 

If warming is not held within the capacity of corals to adapt, over 100 countries stand to lose much or all of their major source of marine biodiversity, fisheries, shore protection, white sand, and tourism revenues. There is an urgent need to continue monitoring SST anomalies and coral bleaching throughout the tropical ocean. Only then can we determine, in a more statistically significant fashion, if the hot period since the early 1980s is due to global warming or results from decadal cycles. However, if action is delayed until absolute statistical certainty is achieved, it may be too late.  


We gratefully acknowledge support from the United States Department of State, through the International Coral Reef Initiative, for the field work reported in this paper, and NOAA for support of research leading to preparation of thermal anomaly maps. This work would not have been possible without the generous field help and detailed information provided by interviews with local divers, fishermen, and environmental agents in each island visited. We especially thank all of those individuals who assisted in field work, who are unfortunately too many to mention by name here. We also particularly thank Jennifer Clarke, Robert Trench, Peter Glynn, Ernest Williams, and John Sapper for constructive comments and practical discussions during the preparation of this manuscript.  


Gleeson MW, Strong AE (1994) Applying MCSST to coral reef bleaching. Proc COSPAR, Hamburg, Germany

Glynn PW (1993) Coral reef bleaching in the 1980s and possible connections with global warming. Trends Ecol Evol 6: 175-179 

Goreau TE (1964) Mass expulsion of zooxanthellae from Jamaican reef communities after Hurricane Elora. Science 145: 383-386

 Goreau TJ, Hayes RL, Slarke JW, Basta DJ, Robertson CN (1993) Elevated sea surface temperatures correlate with Caribbean coral reef bleaching. In: Geyer R (ed) A Global Warming Forum: Scientific, Economic, and Legal Overview. pp 225-255, CRC Press, Boca Raton, FL 

Goreau TJ, Hayes RL (1994) Coral bleaching and ocean "hot spots". Ambio 23: 176-180 

Goreau TJ, Hayes RL (1995a) Monitoring and calibrating sea surface temperature anomalies using satellite and in-situ data to study the effects of weather extremes and climate change on coral reefs. Proc Conf Remote Sensing and Environmental Monitoring for the Sustainable Development of the Americas, San Juan. PR 

Goreau TJ, Hayes RL (199Sb) Coral reef bleaching in the South Central Pacific during 1994. pp 1-201, Global Coral Reef Alliance, Cambridge, MA USA 

Intergovernmental Panel on Climate Change(1996) Climate Change 1995: The science of climate change, summary for policymakers. World Meteorological Organization and United Nations Environment Programme, London, England 

Jokiel P, Coles S (1990) Response of Hawaiian and other Indo-Pacific corals to elevated temperatures. Coral Reefs 8: 155-162 

Magruder WH, Hunt JW (1979) Seaweeds of Hawaii: a photographic identification guide, Oriental Publishing, Honolulu 

Migotto A (1995) Anthozoan bleaching on the southern coast of Brazil, summer 1994. Proc Int Coelenterate Symp, Groningen, Netherlands 

Montgomery RS, Strong AE (1994) Coral bleaching threatens oceans. EOS 75: 145-147 

Strong AK, Barrientos CS, Sapper J (1996) Improved satellite techniques for monitoring coral reef bleaching. Proc 8th Int Coral Reef Symp, Panama City, Panama 

United Nations (1992) Framework Convention on Climate Change. United Nations 

Veron C (1986) Corals of Australia and the IndoPacific. U Hawaii Press, Honolulu 

Williams EH, Bunkley-Williams L (1990) The world-wide coral bleaching cycle and related sources of coral mortality. Atoll Res Bull 335: 1-71