Integrating Multi-Datasets in Google Earth Engine for Advanced Hydrological Modeling Using the Soil Conservation Service Curve Number Method

Abstract

This research investigates the potential of employing cloud computing and open-data sources for hydrological modeling. In Google Earth Engine (GEE) soil conservation service Curve Number was employed for runoff estimation. The SCS CN model is very effective for surface runoff estimation and is used quite often in the simulation of rainfall and runoff estimation. Google Earth Engine contains a vast range of datasets that can be easily retrieved and visualized. In this research work Soil conservation service curve number implemented in Google Earth Engine using rainfall data and antecedent moisture condition (AMC). Firstly, USDA soil texture data is used to derive Hydrological soil groups than MODIS land use land cover (LULC) is overlayed with hydrological soil groups. Three rainfall datasets, TRMM, GPM and CHIRPS from 2005 to 2015 are thoroughly analyzed. To ensure the accuracy and reliability of our model we first estimated the runoff against each rainfall and then performed comparative analysis between the rainfall runoff datasets. The 10-year datasets display a clear seasonal trend as well as significant variations observed in average rainfall and runoff values in pre monsoon and post monsoon seasons. In the year 2015 the highest average rainfall of 1412 mm was observed, which resulted in an average runoff of 215 mm. On the other hand, the lowest average rainfall of 672 mm was observed in the year 2009 which resulted in a mean runoff of 78 mm. For validation of our model, observed meteorological data from the Climate Forecast System Reanalysis (CFSR), Water and Power Development Authority (WAPDA), and Pakistan Meteorological Department (PMD) were utilized. In 2008, 2009, and 2010, CHIRPS consistently proves better accuracies in comparison to GPM and TRMM, with accuracies of 90%, 79%, and 86% respectively. Furthermore, the sensitivity analysis conducted on the parameters of the SCS CN model reveals the impact of initial abstraction and Curve Number values on the estimation of runoff. In conclusion, this research work offers significant contributions to the understanding of hydrological processes in regions primarily influenced by monsoons and presents useful suggestions for the implementation of sustainable practices in water resource management.