GENERATING BIOLOGICALLY RELEVANT ENVIRONMENTAL DATA FROM REMOTE SENSING IMAGERIES AND OCEANOGRAPHIC MODELS TO SUPPORT SPATIAL PRIORITIZATION OF MARINE BIODIVERSITY CONSERVATION AND MANAGEMENT IN INDONESIA

Long term Earth observation data stored in Google Earth Engine (GEE) can be ingested and derived to biologically relevant environmental variables that can used as the predictors of a species niche. The aim of this research was to create a script using GEE to generate biologically meaningful environmental variables from various Earth observation data and models in Indonesia. Elevation and bathymetry raster data from GEBCO were land masked and benthic terrain modelling were done in order to get the aspect, depth, curvature, and slope. HYCOM and MODIS AQUA dataset were filtered using spatial (Indonesia and surrounding region) and temporal filter (from 2002–2017), and reduced to biologically meaningful variables, the maximum, minimum, and mean. Water speed vector (northward and eastward) data were also converted in to scalar unit. In order to fill data gaps, kriging was done using Bayesian slope. Results shows the water depth in Indonesia ranges from 0 – 6827 m, with slope ranging from 0 – 34.33°, aspect from 0 – 359.99°, and curvature from 0 – 0.94. Variables representing water energy, mean sea surface elevation ranges from 0 – 0.85 m, and mean scalar water velocity 0 – 4 m/s. Mean surface salinity ranges from 20.09 – 35.32‰. Variables representing water quality includes mean of particulate organic carbon which ranges from 25.31 – 953.47‰ and mean of clorophyll-A concentration from 0.05 – 13.63‰. These data can be used as the input for species distribution models or spatially explicit decision support systems such as Marxan for spatial planning and zonation in Marine and Coastal Zone Management Plan.