CoastalDEM: A global coastal digital elevation model improved from SRTM using a neural network

Positive vertical bias in elevation data derived from NASA's Shuttle Radar Topography Mission (SRTM) is known to cause substantial underestimation of coastal flood risks and exposure. Previous attempts to correct SRTM elevations have used regression to predict vertical error from a small number of auxiliary data products, but these efforts have been focused on reducing error introduced solely by vegetative land cover. Here, we employ a multilayer perceptron artificial neural network to perform a 23-dimensional vertical error regression analyses, where in addition to vegetation cover indices, we use variables including neighborhood elevation values, population density, land slope, and local SRTM deviations from ICESat altitude observations. Using lidar data as ground truth, we train the neural network on samples of US data from 1-20 m of elevation according to SRTM, and assess outputs with extensive testing sets in the US and Australia. Our adjustment system reduces mean vertical bias in the coastal US from 3.67 m to less than 0.01 m, and in Australia from 2.49 m to 0.11 m. RMSE is cut by roughly one-half at both locations, from 5.36 m to 2.39 m in the US, and from 4.15 m to 2.46 in Australia. Using ICESat data as a reference, we estimate that global bias falls from 1.88 m to −0.29 m, and RMSE from 4.28 m and 3.08 m. The methods presented here are flexible and effective, and can be effectively applied to land cover of all types, including dense urban development. The resulting enhanced global coastal DEM (CoastalDEM) promises to greatly improve the accuracy of sea level rise and coastal flood analyses worldwide.