Groundwater stable isotope maps for Germany: recommendations and perspectives working with heterogeneous data

Abstract Background Stable water isotopes $$\delta ^{18}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mi>δ</mml:mi> <mml:mn>18</mml:mn> </mml:msup> </mml:math> O and $$\delta ^2$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mi>δ</mml:mi> <mml:mn>2</mml:mn> </mml:msup> </mml:math> H in the water molecule are ideal tracers of the hydrological cycle to understand mixing processes between different aquifers, groundwater-surface water interactions or groundwater recharge. To date, no centralized access to standardized groundwater isotope data is available in Europe. This study addresses this gap in Germany by assembling existing historical data, completing the spatial coverage with new samplings and presenting the gained data in the form of spatially interpolated maps, so-called isoscapes. Results Collected data from more than 8 000 groundwater sampling stations cover the entire German territory and span the last 50 years. We successfully implemented co-kriging interpolation algorithms taking into account elevation and temperature. This produced several isoscapes for 15 years periods to accurately represent typical young groundwater residence times. Here we present two representative $$\delta ^{18}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mi>δ</mml:mi> <mml:mn>18</mml:mn> </mml:msup> </mml:math> O isoscapes for the periods 1965-1980 and 2015-present. These examples allow assessments of this commonly used interpolation method. The emerging trends correspond to similar observations in precipitation. Despite the large discrepancies in data quality and density, the absolute error of the isoscapes remained under 3.1 $$\permil$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>‰</mml:mi> </mml:math> for $$\delta ^{18}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mi>δ</mml:mi> <mml:mn>18</mml:mn> </mml:msup> </mml:math> O. Major limitations are the extrapolation to regions with no primary data as well as boundaries between data-rich and data-scarce areas. Conclusions With this work, we provide the basis for further comprehensive large-scale isotope research, as well as reliable background values for local investigations in Germany. We also formulate recommendations to work with heterogeneous and scarce data, to ensure the plausibility of isotope values and to identify unknown deep groundwater samples using additional information such as tritium activities. These strategies could be applied to develop similar initiatives in other European countries and across national borders.