Australian Coral Bleaching Multifactor Dataset
Description
Australian Coral Bleaching Multifactor Dataset is an analysis-ready tabular dataset of 2,985 quality-controlled coral bleaching observations across Australian reef regions and adjacent seas, spanning 1998–2017. Each record represents a unique 0.05° grid cell and survey month, with geolocation (LAT, LON), time (Year, Month), bleaching severity, and a suite of multi-factor environmental predictors assembled for statistical modelling and benchmarking. Bleaching outcomes Bleaching observations and severity information in this release were sourced from the updated global mass coral bleaching database (Version 2.0) compiled by Virgen-Urcelay & Donner (2023, v2). The original event-level records were subset to the Australian region and then harmonized to the analysis format used here (unique 0.05° grid cell × survey month). Bleaching is provided as an ordinal severity class (Seve2, 0–3) together with a continuous proxy (Seve) mapped to (0,1) for downstream statistical modelling. Thermal-stress and climate variables Using the NOAA Coral Reef Watch (CRW) 5-km CoralTemp v3.1 SST product, we derived multiple thermal-exposure metrics aligned to each survey record, including Degree Heating Weeks (DHW), maximum HotSpot (maxhs), Heating Increase Rate (HIR; daily rate of increase in DHW during the exposure window), and sea-surface temperature anomaly metrics (SSTA, SSTA_SD, SSTA_freq, SSTA_freq_SD). Large-scale climate variability is represented by the Oceanic Niño Index (ONI), computed by NOAA CPC from ERSSTv5 as the 3-month running mean SST anomaly in the Niño-3.4 region (5°N–5°S, 170°–120°W). Optical environment To characterize underwater light regime and water clarity, the dataset includes monthly Level-3 MODIS OceanColor composites: photosynthetically available radiation (PAR) and the diffuse attenuation coefficient at 490 nm (Kd₄₉₀). Cyclone exposure Cyclone exposure indicators (cyclone, cyclone_100, cyclone_200) were derived from the Australian Bureau of Meteorology (BoM) best-track database using distance and time windows around each survey record. Following Lugo-Fernández and Gravois (2010), a 200 km radius was used to indicate cyclone impact. Bathymetry and proximity metrics Depth was taken from event records where available; missing values were supplemented using AusBathyTopo (Geoscience Australia, 2017). Distance to land (DIST_L) was derived from the NASA OBPG Distance to the Nearest Coastline global grid (based on GMT intermediate coastline) and resampled to 0.05°. Given evidence that mangrove-associated environments can mitigate bleaching, distance to mangroves (DIST_M) was computed as distance to the nearest polygon from Global Mangrove Watch v3 (Bunting et al., 2022). Regional context Regional labels are provided via ECOREGION codes and the original site/region field (Site). Near-surface wind is included as an additional environmental covariate.
Files
Institutions
- Ocean University of ChinaShandong, Qingdao
- The University of SydneyNew South Wales, Sydney