Isotopic and statistical methods of analyzing groundwater- surface water interaction in an urban area (Munich, Germany)

Patrick Kotyla1, Arno Rein2, Anja Wunderlich2, Kai Zosseder2, Florian Einsiedl2
1 Department of Health and Environment of the City of Munich (RGU)
2 Chair of Hydrogeology, Technische Universität München

P 8.6 in Isotope and tracer methods in hydrogeology



Bank filtration is the infiltration of surface water, mostly from a river system into a groundwater system. It represents an excellent tool to improve the water quality during its passage through the ground below surface. Therefore, it plays a major role in water supply of many countries around the world. River water infiltration into groundwater also influences the seasonal temperature regime in an aquifer leading to higher temperatures in summer and lower temperatures during the winter months. The effectiveness of a shallow geothermal system may therefore highly depend on the travel time and contribution of river water to the shallow geothermal system (Allen et al. 2003). Consequently, it is important to describe qualitatively and quantitatively the effect of river water infiltration to groundwater systems.

In the last two years, attempts were made to find out the hydraulic and thermal effects of the river Isar on the quaternary aquifer system in Munich. For this purpose, a test field with seven observation wells along the river Isar has been installed in the southern part of Munich including data-loggers for continuous measurements of hydraulic head, groundwater temperature and electrical conductivity. Additionally, a sensor with an integrated data logger was installed in the Isar upstream of the test site. Weekly measurements of 18O-content in the river and wells were used to determine the contribution of river water into the local groundwater (Fig. 1). The isotopic data was also linked with a 1D-advection-dispersion model to estimate aquifer parameters such as dispersivity and mean transit times (Małoszewski & Zuber 1982). The results obtained from modeling were compared with the analysis of continuously measured specific electrical conductivity data. An easy approach of identifying a single characteristic transfer time between two time series is by cross-correlation, in which the time shift with the highest correlation coefficient is interpreted as the effective travel time (Vogt et al. 2009). Linked with the linear regression coefficient it results in a travel time distribution of the infiltrated groundwater.

The 1D-dispersion-advection model yielded mean residence times between 3 and 20 days for the wells in various distances to the river. The estimated portion of bank-infiltrated river water for the different observation wells ranged from 32 to 100 %. The isotopic data reveal a good agreement with the electrical conductivity measurements for the observation wells near to the river. However, the observation wells far away from the stream show longer mean residence times with electrical conductivity as tracer.

Fig 1: 18O-contents measured in the Isar River and the observation wells P1 and P8 (recharge area)
Fig 1: 18O-contents measured in the Isar River and the observation wells P1 and P8 (recharge area)



Allen, A., Milenic, D. & Sikora, P. (2003): Shallow gravel aquifers and the urban ʻheat islandʼ effect: a source of low enthalpy geothermal energy. ˗ Geothermics, 32(4-6):569-578.

Małoszewski, P. & Zuber, A. (1982): Determining the turnover time of groundwater systems with the aid of environmental tracers: I. Models and their applicability.  - J. Hydrol., 57 : 207—231. 

Vogt, T., Hoehn, E., Schneider, P. & Cirpka, O.A. (2009): Untersuchung der Flusswasserinfiltration in voralpinen Schottern mittels Zeitreihenanalyse. – Grundwasser 14(3), 179-194.