Remote sensing and open top chamber field measurements of seasonal patterns of evapotranspiration in an extreme cold Patagonian desert

CRISTIANO PM; MADANES N; BUCCI SJ; GOLDSTEIN G; PEREYRA DA; SCHOLZ FG

HYDROLOGICAL PROCESSES

JOHN WILEY & SONS LTD

Lugar: LOndres; Año: 2016

0885-6087

Abstract Accurate estimates of seasonal evapotranspiration (ET) at different temporal and spatial scales are essential for understanding the biological and environmental determinants of ecosystem water balance in arid regions and the patterns of water utilization by the vegetation. For this purpose, remote sensing ET estimates of a Patagonian desert in Southern Argentina were verified with field measurements of soil evaporation and plant transpiration using an open top chamber. Root distribution and seasonal variation in soil volumetric water content was also analyzed. There was a high correlation between remote sensing and field measurements of ecosystem water fluxes.A substantial amount of the annual ET occurred in spring and early summer (73.4 mm)using winter rain stored in the soil profile and resulting in water content depletion of theupper soil layers. A smaller amount of annual ET was derived from few rainfall eventsoccurring during the mid or late summer (41.4 mm). According to remote sensing, the 92.9 % of the mean annual precipitation returns to the atmosphere by transpiration or evaporation from the bare soil and by canopy interception. Only 7.1 % infiltrates to soil layers deeper than 200 cm contributing to the water table recharge. Fourier time series analysis, cross-correlation methods and multiple linear regression models were used to analyze 11 years of remote sensing data to assess determinants of water fluxes. A linearmodel predicts well the variables that drive complex ecosystem processes such as ET.Leaf area index and air temperature were not linearly correlated to ET because of the multiple interaction among variables resulting in time lags with ET variations and thusthese two variables were not included in the linear model. Soil water content, the fraction of photosynthetic active radiation and precipitation explained 86% of the ET monthly variations. The high volumetric water content and the small seasonal variations at 200 cm depth were probably the result of little water uptake from deeper soil horizons by roots with low hydraulic conductivity.