What determines hydrological systems’ resilience to climate change?

The increasing frequency of extreme climatic events challenges both science and water resources management. Flood risk assessment on the one hand, and drought risk assessment on the other hand are usually considered the tasks of different sub-disciplines. However, recent studies suggest that the two might be more closely related than widely assumed. There is some evidence that the information content of stream discharge in regard to groundwater systems is widely underrated and that groundwater dynamics is key for long-term flood risk assessment. Here we bring together two different lines of evidence in order to gain better understanding of how the interplay between groundwater and streams determines regional hydrological systems resilience to climate change.

On the one hand, findings from a joint analysis of 292 time series of stream discharge and groundwater head from a 36,000 km² region covering a 43 years period are reported. Spatial variability, that is, different behaviour at different sites, could largely be traced back to spatially varying input reflecting regional climatological patterns, and to different degrees of damping and low-pass filtering of the hydrological input signal in the subsurface. Stream discharge and groundwater head dynamics differed in regard to the latter, but not without remarkable overlap. Both for stream discharge and groundwater head the degree of low-pass filtering was very closely related to long-term trends, similar as in other studies (Lischeid et al. 2021). Beyond that there was no clear distinction between surface and subsurface hydrological dynamics.

The second study aimed at determining the key drivers of extreme floods based on 73,350 synthetic hydrographs from a comprehensive modelling study, comprising 163 catchments with 450 model realizations each. On the one hand, tail heaviness of flood distribution was assessed by the shape factor of the extreme value distribution (Macdonald et al. 2024). On the other hand, a newly developed Cumulative Periodogram Convectivity (CPC) index was tested which is based on the degree of low-pass filtering of hydrological time series. Both indices were closely related to the extreme value behaviour of precipitation, to the upper subsurface storage and, to a lesser degree, to catchment area. However, these relationships were less close for the shape factor which suffered from the problem of fitting an extreme value distribution to bent flood frequency curves. In contrast, the CPC index was much more robust.

To conclude, low-pass filtering of hydrological signals in the subsurface proved to be the unifying element for stream discharge and groundwater head as well as for flood and drought hazard characteristics. Contrary to usual expectations, time series of deep groundwater head rather than of shallow groundwater or stream discharge turned out to be the most efficient early warning indicators for the effects of climate change in regard to extreme events. Thus common analysis of runoff and groundwater hydrographs is strongly recommended for science and water resources management.

 

References:

Lischeid et al. (2021), Journal of Hydrology 596, 126096, DOI: 10.1016/j.jhydrol.2021.126096

Macdonald et al. (2024), Hydrol. Earth Syst. Sci., 28, 833–850, https://doi.org/10.5194/hess-28-833-2024