Predictive Risk Mapping of Water Table Depths in a Brazilian Cerrado Area

Information about hydrological systems and water table fluctuations is important for water management and is needed to assess choices for long-term water management policy. Preferably, uncertainty about the future water table depths is quantified to enable risk assessment. The aim of this study is to map water table depths and quantify the associated uncertainty, in order to indicate extreme scenarios of the hydrological system and areas with potential risks of future water table depths in a Brazilian Cerrado area. To this purpose, a specific type of transfer function-noise (TFN) model, the Predefined Impulse Response Function In Continuous Time (PIRFICT) model is applied. Being the most important driving forces of water table fluctuation, precipitation and potential evapotranspiration are incorporated as exogenous variables into the model. Series of semi-monthly observed water table depths of 37 months length are available from 37 wells in the Jardim river watershed. In the Jardim area almost all natural Cerrado vegetation has been replaced by agricultural crops, some of which are intensively irrigated. Considering that climate conditions of the past 30 years can represent the future climate conditions, statistical characteristics of the future water table dynamics are estimated by stochastic simulation from the calibrated time series models and using 30-year time series on precipitation and evapotranspiration as input. The results are interpolated spatially using Universal Kriging, utilizing a Digital Elevation Model as ancillary information. The resulting probability maps indicate areas with potential risks of water shortage and shallow water table depths for an arbitrary selected date.