Uncertain modeling on spatio-temporal analysis of water table depths in a watershed

Abstract

The aim of this work was to evaluate water table depths in a watershed during the dry season as well as modeling space-time variability of water table depths through multivariate geostatistics techniques. The study area was the high portion of Jardim's river watershed, Distrito Federal, Brazil. Water regime of that region is ruled by seasonality, resulting in wet (October until April) and dry (May until September) periods well-defined that control the behavior of the water table. On space-time index phenomena, geostatistics provide solutions through linear co-regionalization model, allowing quantifying and mapping groundwater reserves. The lost water volume was quantified, as well as the associated estimation risk, through uncertain measures, creating critical scenarios of the explored aquifer volume during the dry season. During the dry season, there was a decrease of water in that aquifer. The quantile map 10% indicated favorable areas for water use, where the phreatic levels oscillated less. The quantile map 90% indicated areas with great oscillations of the phreatic levels that contribute more to the drainage and recharge of the aquifer.