Coral Bleach Risk Prediction Portal

Description:

Our dataset has been built using daily simulated sea-surface temperature (SST) from 8 Coupled Model Intercomparison Project Phase 6 (CMIP6) models. Data for coral bleaching risk, using standard NOAA bleaching threshold calculation methods, is provided for all 8 models, across 3 Shared Socioeconomic Pathways (SSP). All models have been downscaled to a common 0.5°´0.5° resolution and bias-corrected against a high-quality satellite derived SST dataset. Data for the CMIP6 models can be downloaded freely from the Earth System Grid Federation.

All data files are available through FigShare at https://doi.org/10.25909/25143128.

General Workflow: The first step for exploring this dataset is to select a region of interest, as well as a climate model, SSP and a particular year. To do this, click on the ‘View Map’ tab on the landing page. This will open the Main page (see details below).

The Main page allows the user to select the Year, Scenario, Climate Models, Color scale, and major predefined regions (Areas) for data selection and visualisation. Alternatively, the user can also manually select a region by drawing a polygon around it, or specify geographic coordinates for the region boundaries. Once these options have been selected, the user can click on ‘Submit’ to access the next page (Dashboard).

The Dashboard then displays information and statistics for the selected options, including a map of DHWmax for the selected region/year/scenario/model, a time series with annual DHWmax, annual duration and onset (i.e., start) of Bleaching Alert 1 and 2 conditions (see Dashboard page for details).

The user can then export the map as a .png file, the time series as a .csv file, or the entire dataset as a .nc file.

The Main Page: The Main page allows the user to select the following options through the drop-down lists at the very top of the page:

• Year: any year between 1986 and 2100. • Scenarios: ◦ SSP 2-4.5 “Middle of the road” ◦ SSP 3-7.0 “Regional rivalries” ◦ SSP 5-8.5 “Fossil-fueled development” • Climate models: Either one of the eight CMIP6 models, or the 5-model ensemble (ens5), or the 8-model ensemble (ens8). See Mellin et al. (in review) for details. • Diverging colormaps: One of four common colour scales used for climate data. • Areas: Predefined areas for region selection including ◦ IPCC: IPCC AR6 reference regions ◦ LME: Large Marine Ecosystems ◦ MEOW: Marine Ecoregions of the World ◦ GLOBAL: Global extent. • Corals (on/off): this will display the coral reef mask based on the NOAA Coral Reef Watch dataset.

The user can then select one of the predefined areas by clicking on it, which will highlight it in yellow. Alternatively, a user-defined region can be manually drawn using the polygon icon, or selected by using the ‘Insert coords’ tab.

Once an area has been selected, the ‘Submit’ tab will send this selection to the Dashboard page, where the data will be displayed for this particular area/climate model/scenario (note that it is possible to change the climate model or scenario from the Dashboard page too).

The “Clear” tab will erase any selection.

The “.PNG” tab in the top right corner will download a .png image of the currently displayed map. The “.NC” tab will download a raster stack of all DHW metrics for the 1986-2100 time period, for the climate model and scenario selected.

The Dashboard:

The dashboard contains the following: • Same year, scenarios, climate models and colour scale drop-down lists as on the main page. Changing these options will refresh the dashboard and update the metrics based on this new selection. • A map of the selected area, showing the DHWmax in the year selected and according to the model/scenario selected. The colour scale is indicated above the map. • Geographical extent for the selected area. The user can select a given pixel directly on the map by clicking on the icon underneath the +/- zoom function. The DHWmax for that particular point will then appear in the “Selected point” tab. • Minimum, Maximum, Standard Deviation and Mean DHW in the year selected and according to the model/scenario selected. These values are calculated over the geographic extent selected (or for a single pixel if selected as per above). • Onset DHW>4 and Onset DHW>8: Day of the year when Bleaching Alert 1 (DHW > 4 degrees C-week) and Bleaching Alert 2 start (DHW > 8 degrees C-week), relative to the 1st of January. • Duration DHW>4 and Duration DHW>8: Annual number of days in Bleaching Alert 1 (DHW > 4 degrees C-week) and Bleaching Alert 2 (DHW > 8 degrees C-week). • DHW increment (degrees C-week/year): Mean annual increase in DHWmax for the region selected between 1986-2100. This value corresponds to the slope of the regression line on the plot to the right (DHW yearly maximum over time) • Time series of DHW projections: by default, this is set to display the DHWmax over time (with slope of the regression line corresponding to the DHW increment score to the left). The time series can be adjusted to display the annual onset of bleaching or annual duration of bleaching by clicking on the corresponding tabs, directly above the DHW increment score.

The data can then be downloaded as: • .png file (image of the map being displayed), • .csv file (data displayed in the time series), • .nc file (netcdf file containing all yearly projections for the climate model and scenario selected, in addition to the Maximum Monthly Mean [MMM] sea surface temperature).

Administrative features: This is a beta version of the data portal that will be made publicly accessible upon publication of the references below. Meanwhile, please contact camille.mellin@adelaide.edu.au for any questions or comments.

References: Mellin C, Brown SC, Heron SF, Fordham DA (in review). CoralBleachRisk – Global projections of coral bleaching risk in the 21st century.

Mellin C, Brown SC, Cantin N, Klein-Salas E, Mouillot D, Heron SF, Fordham DA (in review). Cumulative risk of future bleaching for the world’s coral reefs.

All data files are available through FigShare at https://doi.org/10.25909/25143128.