This dataset explores a new approach to predict coral bleaching events. It uses a temperature anomaly map to create a spatially
dynamic temperature threshold for the calculation of degree heating weeks (DHW) instead of using a static constant. The dynamic
threshold was used to classifies map areas with low, medium or high risk of coral bleaching for years 2016 and 2017.
Understanding that the combination of several variables could provide better explanatory value than each individual variable
by itself, we used a classification tree prediction model (Breiman et al, 1984) to select the relevant variables and determine
the threshold values for each of them as the best prediction solution for the bleaching category. Using the data from 2016
and 2017 aerial bleaching surveys at specific reefs, we derived the corresponding anomaly values and paired them with
the estimated bleaching response. The classification tree algorithm will select the values of the variables that produce
the most efficient partition of the data into the bleaching categories. The algorithm was trained using a randomly selected
sample of 80% of the survey locations (training set), and the remaining 20% was used for validation of the results (test
set). The accuracy of the classification system was calculated comparing the predicted bleaching category of the test set
and comparing it with the observed bleaching category.
Using a recursive partition approach we were able to create a system that correctly classified more than 66% of the reef
bleaching conditions. The importance of the variables in the classification procedure according to the number of splits attributed
to that variable is DHWmax anomaly > MHW count anomaly > Proportion of the mixed water column > PAR anomaly > Upwelling anomaly
> MHW duration anomaly. Having a DHWmax anomaly of 4.4 °C-week above the expected climatological value and 0.3 °C above the
expected value for the upwelling anomaly are the conditions linked to a severe bleaching in any reef. No or mild bleaching
occurs when DHWmax anomaly was below 4.4 °C-week, and the water column was mostly stratified.
The data is in geoTIFF format.
CRS: EPSG:4326 - WGS 84 - Geographic
eReefs THREDDS catalogue
NOAA Coral Reef Watch Daily 5km Satellite Coral Bleaching Heat Stress Monitoring Products (Version 3.1)
Beaman, R.J. 2017. High-resolution depth model for the Great Barrier Reef - 30 m. Geoscience Australia, Canberra. http://dx.doi.org/10.4225/25/5a207b36022d2
Simpson, J. H., Tett, P. B., Argote-Espinoza, M. L., Edwards, A., Jones, K. J., and Savidge, G. (1982). Mixing and phytoplankton
growth around an island in a stratified sea. Continental Shelf Research 1, 15–31. doi:10.1016/0278-4343(82)90030-9.
Steven AD, Baird ME, Brinkman R, Car NJ, Cox SJ, Herzfeld M, Hodge J, Jones E, King E, Margvelashvili N, Robillot C. eReefs:
an operational information system for managing the Great Barrier Reef. Journal of Operational Oceanography. 2019 Nov 20;12(sup2):S12-28.
Liu, G., Heron, S., Eakin, C., Muller-Karger, F., Vega-Rodriguez, M., Guild, L., et al. (2014). Reef-Scale Thermal Stress
Monitoring of Coral Ecosystems: New 5-km Global Products from NOAA Coral Reef Watch. Remote Sensing 6, 11579–11606. Doi:10.3390/rs61111579.
Liu, G., Strong, A. E., and Skirving, W. (2003). Remote sensing of sea surface temperatures during 2002 Barrier Reef coral
bleaching. Eos, Transactions American Geophysical Union 84, 137–141. Doi:10.1029/2003EO150001.
Breiman L., Friedman J. H., Olshen R. A., and Stone, C. J. (1984) Classification and Regression Trees. Wadsworth.
Benazzouz, A., Mordane, S., Orbi, A., Chagdali, M., Hilmi, K., Atillah, A., et al. (2014). An improved coastal upwelling
index from sea surface temperature using satellite-based approach – The case of the Canary Current upwelling system. Continental
Shelf Research 81, 38–54. Doi:10.1016/j.csr.2014.03.012.
This dataset is filed in the eAtlas enduring data repository at: data\custodian\2018-2021-NESP-TWQ-4\4.2_Oceanographic-drivers-of-bleaching