From the national pilots in TACTIC we have locally sound estimates of long-term groundwater recharge but lack data to fill the gaps in between. At GEUS we are working on mapping these gaps, this will unify the local model estimates on groundwater recharge into one harmonized Pan-European groundwater recharge map.

We approach this issue by using distributed data sets of precipitation and remote sensing-based observed evapotranspiration. This provides us with an estimate of effective precipitation at a 1-km resolution covering all of Europe.

With a machine learning algorithm, we are then training the effective precipitation data set against local estimates of recharge coefficients in order to identify the correlation between effective precipitation and recharge, explained by distributed co-variates such as surface-near geology, vegetation cover, topography, climate, BFI, etc.

These efforts will provide a fully distributed European map of long-term potential groundwater recharge for the period 1980-2010. This provide not only useful information on the gaps themselves but also the possibility of comparing the map with the local estimates from the national pilots.