Assessing recharge using remotely sensed data in the Guarani Aquifer System outcrop zone

Thursday, 18 December 2014
Murilo Cesar Lucas, USP University of Sao Paulo, São Paulo, Brazil, Paulo Tarso Sanches Oliveira, Escola Engenharia Sao Carlos, Tucson, AZ, United States, Davi Diniz Melo, USP University of Sao Paulo, São Carlos, Brazil and Edson Wendland, USP University of Sao Paulo, Sao Carlos, Brazil
Groundwater recharge is an essential hydrology component for sustainable water withdrawal from an aquifer. The Guarani Aquifer System (GAS) is the largest (~1.2 million km2) transboundary groundwater reservoir in South America, supplying freshwater to four countries: Brazil, Argentina, Paraguay and Uruguay. However, recharge in the GAS outcrop zones is one of the least known hydrological variables, in part because studies from hydrological data are scarce or nonexistent. We assess recharge using the water-budget as the difference of precipitation (P) and evapotranspiration (ET). Data is derived from remotely sensed estimates of P (TRMM 3B42 V7) and ET (MOD16) in the Onça Creek watershed over the 2004­–12 period. This is an upland-flat watershed (slope steepness < 1%) dominated by sand soils and representative of the GAS outcrop zones. We compared the remote sensing approach against Water Table Fluctuation (WTF) method and another water-budget using ground-based measurements. Uncertainty propagation analysis were also performed. On monthly basis, TRMM P exhibited a great agreement with ground-based P data (R2 = 0.86 and RMSE = 41 mm). Historical (2004–12) mean(±sd) satellite-based recharge (Rsat) was 537(±224) mm y-1, while ground-based recharge using water-budget (Rgr) and WTF (Rwtf) method was 469 mm y-1 and 311(±150) mm y-1, respectively. We found that ~440 mm y-1 is a reasonable historical mean (between Rsat, Rgr and Rwtf) recharge for the study area over 2004–2012 period. The latter mean recharge estimate is about 29% of the mean historical P (1,514 mm y-1). Our results provide the first insight about an intercomparison of water budget from remote sensing and measured data to estimate recharge in the GAS outcrop zone. These results should be useful for future studies on assessing recharge in the GAS outcrop zones. Since accurate and precise recharge estimation still is a gap, our recharge satellite-based is considered acceptable for the Onça Creek watershed.