干旱区生态环境知识资源中心

The BAGROV method has been modified in the last decade to improve the estimation of the long-term averaged actual evapotranspiration and groundwater recharge. Among others, GLUGLA et al. (1999) made the method applicable to sites with groundwater influence and impervious surfaces in urban areas and applied it to the whole region of Berlin. GLUGLA et al. (2003) validated and improved the method with long-term lysimeter measurements and thus provided the basis for the regionalization of evapotranspiration and groundwater recharge in the context of the revision of the Hydrological Atlas of Germany (JANKIEWICZ 2005). With the benefit of meanwhile improved possibilities to consider the hydraulic properties of the soils, a new methodological extension is presented here that allows to make easy and reliable estimates of the contribution of soils to groundwater recharge from the soil texture class, the depth of the groundwater table, and the monthly mean values of precipitation and potential evapotranspiration for a wide range of soils under selected Central-European climate conditions. In this paper, the numerical simulation model SWAP (KROES et al. 1999, KROES & VAN DAM 2003) was used to improve the treatment of soil hydrologic processes, yielding accurate hydropedotransfer functions that allow the prediction of actual evapotranspiration E-a, and hence groundwater recharge, for all soil texture classes, depth of the groundwater table, monthly averages of precipitation P. and potential evapotranspiration E-p. The acceptable and comprehensive database needed for that purpose was created by the physical-based numerical simulation model SWAP for 135 different site conditions and 30 simulation years each. Site conditions were specified by three climate regions, four soil types (14 soil texture classes), and three groundwater-depth conditions. The procedure was performed here only with the example of grassland. Based on simulated values of actual evapotranspiration, a hydropedotransfer function was developed to consider site-specific conditions by employing the parameter b in BAGROV’s differential equation d E-a/d P = 1 -(E-a/E-p)(b) (E-a, E-p actual and potential evapotranspiration, P = precipitation). Under humid conditions (P - E-p < 0), the BAGROV method predicts long-term averages of actual evapotranspiration and groundwater recharge with a standard error of 15 mm per year (R = 0.96). Simulation results gained by SWAP revealed that under semi-arid climatic conditions (P - E-p < 0), simulated actual evapotranspiration exceeds precipitation. Thus, lowland aquifers (recharged by inflow from hilly regions) might be depleted. Since the former BAGROV equation was not valid for these conditions, an alternative statistics-based transfer function was developed to predict groundwater recharge including groundwater depletion with a standard error of 26 mm only (R = 0.975). The software is available using the link www.small-scale-hydrology.de.