Groundwater recharge and water table levels modelling using remotely sensed data and cloud-computing

Abstract

Hydrological modeling is still a challenge for better management of water resources since most of the established models are based on point data. The advent, improvement and popularization of remote sensing has brought new perspectives to modelers, allowing access to reliable and representative data over vast areas. However, several tools are still under-explored and actually used in the water resources planning and decision-making process, especially groundwater, which is a hidden resource. The objective of this work was to contribute to groundwater dynamics comprehension assessing the suitability of using remote sensing data in the water-budget equation for estimating groundwater recharge (GWR) and its impact at water table depths (WTD) in a representative Guarani Aquifer System (GAS) outcrop area. The GAS is the largest transboundary groundwater reservoir in South America, yet recharge in the GAS outcrop zones is one of the least known hydrological variables. The remotely sensed WTD model was adapted from the Water Table Fluctuation (WTF) method. We used Google Earth Engine to extract time series of precipitation, evapotranspiration, and surface runoff from the Famine Early Warning Systems Network Land Data Assimilation System (FLDAS) dataset for the Angatuba Ecological Station (EEcA), Sao Paulo State, Brazil, over 2014–2017 period. GWR and WTD were modeled in eight groundwater monitoring wells. Bias analysis of precipitation data from FLDAS were perfomed using rain gauge data (2000–2018). Two GWR scenarios (S1 and S2) were assessed as well as the impact of the specific yield values ( $${S}_{\mathrm{y}}$$ ) in the model outputs. In C1 GWR ranged from 10.8% to 19.69% of rain gauge, although in C2 GWR ranged from 0.9 to 13%. The WTD model showed RMSE values ranging from 0.36 to 1.12 m, showing better results in the shallow wells than the deeper ones. These results are useful for future studies on assessing groundwater recharge in the GAS outcrop zones. This remotely sensed approach can be reproduced in regions where data are scarce or nonexistent.