Create a distance to coast map and use it in your Random Forest model
Step 1: get your data from openstreet map
Step 2: Clip your data for your specific area of interst
Step 3: Update the data to GEE.
Step 4: Use the code below to create the distance to coast layer
// Load Study Area var mekongBuffer = ee.FeatureCollection('ft:1LEGeqwlBCAlN61ie5ol24NdUDqB1MgpFR_sJNWQJ'); var mekongRegion = mekongBuffer.geometry(); // Buffered region var coastline = ee.FeatureCollection("users/servirmekong/shapefiles/coastLines").filterBounds(mekongRegion); coastline = coastline.map(function(feat){ return feat.set("class",1); } ); // Make an image out of the population attribute and display it. var coast =coastline.reduceToImage({properties: ['class'], reducer: ee.Reducer.first() }); var kernel = ee.Kernel.euclidean(500000,"meters"); var distance = coast.distance(kernel,false).clip(mekongRegion) Map.addLayer(distance,{min:0,max:500000},'distance');
http://www.distancetocoast.com is a quick way to find in the US.
LikeLike