Weather radar precipitation as functional input for the operative calculation of potential infiltration water volume: the pilot case of the Lombardy Region

Since its application in the meteorological field, radar has shown great potential for hydrologic application. Although indirect measures, radar precipitation values have the advantage to be spatially continuous, avoiding errors due to the interpolation of rain-gauge data (often sparse and poorly located). Despite this, radar has not yet achieved the expected success as input for hydrologic models and water balance estimations. In this context, this study aims to assess whether the use of this dataset is costeffective for estimating the volume of Potential GroundWater Recharge (PGWR), at the annual scale, in the Lombardy Region. This application included the use of Swiss radar data (1 x 1 km2 resolution) and datasets from 137 weather stations at the annual scale for the 2011-2020 period. At the Italian station locations, radar underestimates the amount of precipitation of about 21%. For this reason, the radar data were corrected by adding the interpolated error calculated at the rain-gauge locations using deterministic interpolation techniques: Thin Plate Spline (RAD+TPS) and Inverse Distance Weighting (RAD+IDW). Using the simplest water balance equation to maximize the operativity of the proposed approach, the PGWR volume was calculated multiplying effective rainfall by potential infiltration coefficients based on the definition of hydrogeological complexes. Finally, the outcomes were evaluated by comparison with the BIG-BANG project (ISPRA). Although the two methods apply different calculation procedures, the results are comparable and of the same magnitude (on average the precipitation values are around 28.2 Gm3/y while PGWR values are around 6.7 Gm3/y). Comparing the precipitation results, a strong correspondence was found considering the mountain part of the Region (alpine and pre-alpine belt) with differences of only -4% and +3% for RAD+IDW and RAD+TPS, respectively. In the Po Plain, differences reached up to -9% (RAD+IDW) and 29% (RAD+TPS). The volume of PGWR led to similar results with a variation of about -11% (RAD+IDW) and +19% (RAD+TPS) compared to the BIG-BANG project. The comparison with the national project showed that the use of corrected radar precipitation data as input in a simplified balance formula is a useful solution to efficiently obtain satisfactory results at yearly level. To increase its usefulness for groundwater management (e.g., drought events), a finer temporal resolution (monthly or daily) should be tested.