Remote sensing estimates of supplementary water consumption by arid ecosystems of central Argentina

Abstract

Besides precipitation inputs, evapotranspiration of irrigated and natural oases, i.e. riparian and phreatophytic ecosystems, of rain-shadow deserts is strongly influenced by lateral water inputs supplied by mountain rivers and foothill-recharged aquifers. A better understanding of these supplies and their imprint on the water consumption of those inflow dependent ecosystems (IDEs) across arid regions is critical to manage agricultural outputs and ecosystem conservation, and the hydrological trade-offs that emerge among them. Actual operative satellite and physically-based algorithms able to map evapotranspiration (ET) rates at regional scales still fail when they are applied in ungauged regions because of their high parameterization and meteorological data requirements. We introduce an ecological and satellite-based approach to explore the impacts of external water supplies on arid ecosystems, focusing on the Central Monte desert and its water supplies from the Andean Cordillera, in Argentina. Mean annual precipitation (MAP) and the Enhanced Vegetation Index (EVI) from MODIS imagery, used as a surrogate of ET, were the input variables of our empirical model. Two related biophysical indexes were generated for the whole territory of interest based on a MAP–EVI regional function calibrated for the region: the EVI Anomaly (i.e. deviation from a reference with similar MAP) and the ET Anomaly (i.e. additional water consumption besides MAP). These indexes allowed us to identify IDEs and to quantify the impact of remote lateral inflows as well as local constrains on the water balance of rangelands, and irrigated and natural oases. The performance of this satellite-based approach was evaluated through comparisons with independent ET estimates based on plot (known crop coefficients) and basin (measured water budgets) scale approaches. Relative errors in the 2–18% range at plot and basin scale are in agreement with those uncertainties reported by other satellite and physically-based approaches. Our approach provides a simple yet robust diagnostic tool to characterize water balance in arid regions, aimed to improve the identification of inflow dependent ecosystems and their management under the demanding pressures of land use and climate change.