A cloud-based pre-processor for physically-based hydrologic models

Physically-based hydrologic models are a powerful tool to study hydrologic systems whose application requires creating a computational mesh and typically an essential number of input parameters. This work describes the WMF-GHOST module, a pre-processor for the physically-based hydrological model GHOST. The module uses the Watershed Modeling Framework (WMF) to define a set of functions that rapidly create all the input files needed to run the GHOST model. WMF-GHOST creates Voronoi polygons that describe the watershed terrain and its connectivity. Moreover, the module extracts information on the land use and soil type at each polygon. WMF-GHOST was developed as a python class that allows the user to define the watershed of interest and obtain all GHOST's input parameters. WMF-GHOST has several functions that allow the user to modify the mesh polygons, retrieve information from Google Earth Engine, and define regions with different levels of refinement. So far, we have tested WMF-GHOST in the Google Colab service with successful results, allowing us to implement GHOST in several watersheds relatively quickly. The WMF-GHOST is of significant relevance since it is an open tool that helps implement a complex hydrological model and could be adapted and use by the community for their needs.