Revisiting Topographic Horizons in the Era of Big Data and Parallel Computing

Widely used to calculate illumination geometry for estimates of solar and emitted longwave radiation, and for correcting remotely sensed data for topographic effects, digital elevation models (DEMs) are now extensive globally at 10–30-m spatial resolution and locally at spatial resolutions down to a few centimeters. Globally, regionally, or locally, elevation datasets have many grid points. Many software packages calculate gradients over every grid cell or point, but in the mountains, shading by nearby terrain must also be assessed. Terrain may obscure a slope that would otherwise face the Sun. Four decades ago, a fast method to calculate topographic horizons at every point in an elevation grid required computations related only linearly to the size of the grid, but grids now have so many points that parallel computing still provides an advantage. Exploiting parallelism over terrain grids can use alternative strategies: among columns of a rotated grid, or simultaneously at multiple rotation angles, or on different tiles of a grid. On a multi-processor machine, the improvement in computing time approaches 2/3 the number of processors deployed.