Bleaching and environmental data for global coral reef sites from 1998-2017 (NCEI Accession 0291401)

This dataset contains meteorological and physical data collected in the Andaman Sea or Burma Sea, Arabian Sea, Bali Sea, Banda Sea, Bay of Bengal, Bismarck Sea, Caribbean Sea, Celebes Sea (Sulawesi Sea and Mindanao Sea), Coral Sea, East China Sea (Tung Hai), Flores Sea, Gulf of Aden, Gulf of Aqaba, Gulf of Boni, Gulf of Guinea, Gulf of Oman, Gulf of Suez, Gulf of Thailand, Gulf of Tomini, Indian Ocean, Java Sea, Laccadive Sea, Makassar Strait, Malacca Straits, Molucca Sea, Mozambique Channel, North Atlantic Ocean, North Pacific Ocean, Persian Gulf (Gulf of Iran), Philippine Sea, Red Sea, Savu Sea, Solomon Sea, South Atlantic Ocean, South China Sea (Nan Hai), South Pacific Ocean, and Sulu Sea from 1998-08-23 to 2017-12-20. These data include depth, sea surface temperature, water temperature, and wind_speed. These data were collected by Robert van Woesik of Florida Institute of Technology and Deron Burkepile of University of California-Santa Barbara as part of the "Identifying coral reef 'bright spots' from the global 2015-2017 thermal-stress event (Coral Reef Brightspots)" project. The Biological and Chemical Oceanography Data Management Office (BCO-DMO) submitted these data to NCEI on 2019-12-09.

The following is the text of the dataset description provided by BCO-DMO:

Bleaching and environmental data for global coral reef sites

Dataset Description:
Acquisition Description:
The coral bleaching data were composed of the Reef Check dataset ( reefcheck.org ), collected by a mixture of professional scientists (56%) and trained and certified citizen-scientists (44%) using a standardized transect protocol. The dataset includes counts of the number of coral colonies showing bleaching (i.e., the percent of reef corals that were recorded as bleached), which was classified as site-wide bleaching. We also examined the prevalence of coral bleaching per coral ecoregion (as defined by Veron et al. 2015). We used the global Coral Reef Temperature Anomaly Database (CoRTAD Version 6) from the National Oceanic and Atmospheric Administration ( www.nodc.noaa.gov/sog/cortad/ ) to predict coral bleaching prevalence and intensity across reefs worldwide. All CoRTAD variables were weekly data provided on a grid cell basis, of ~4km resolution, from 1982 to 2017(Sully 2019).

A Pair-wise Pearson's correlation of coefficients was used to determine which covariates were highly collinear. We used generalized linear mixed models, within a Bayesian framework, to examine the influence of the covariates on coral bleaching. Some sites were repeatedly surveyed and therefore site was treated as a random effect. Covariates were modeled with flat normal priors. The Bayesian model was implemented in R34 and run through the rjags package that calls JAGS35, with 3 chains, a burn-in of 4000, and 5000 iterations (Sully 2019).