Attributing the spatial patterns of hydrological change to the effects of climate and land use change by distributed modelling

This study tends to attribute the spatial patterns of hydrologic alteration in mid-latitude montane basins to the key driving forces being the climate and land use change. Physically-based distributed modeling system MIKE SHE was used for the analysis of changing spatiotemporal patterns of extreme runoff processes in montane catchments by using time series of hydrometeorological  observations, and spatially distributed MODIS data for evapotranspiration (ET) and leaf area index (LAI) as key resources for the model setup.

Czech Republic is surrounded by mountain ranges from all sides but the southeastern border, which is drained towards the Danube. Due to this concentric orientation of topography each mountain range has different aspect and exposition due to atmospheric processes. Does the basin hydrological reaction on changing the environment depend on the aspect or is there an overall trend present in all basins? In order to answer such a question, it is necessary to understand the main drivers of changes and to quantify the effect of each separately. 

8 headwater basins were analyzed of average size of 73 km2 where significant trends of hydrologic processes were detected from long-termed time series (1952 to 2018). Some of the trends are common for all basins such as seasonal shift of snowmelt period but other trends are rather site specific such as frequency of peak flows. Previous studies show that the hydrologic reaction on climate signal is the most dominant driver of the hydrologic alterations however there are other drivers such as forest disturbances that can mislead the interpretation of trend behavior. 

The aim of the study was to separate the effects of those main drivers by a detailed distributed physically-based modeling system MIKE SHE. Input data originated from official and publicly available sources in order to design a methodology that could be reproduced in other basins of comparable properties. Models are bent together thus results of similar spatial-temporal quality were obtained for further analysis. Stational data but also remote sensed data in the grid format were gathered in a comprehensive database. 

Two groups of scenarios were applied. First group was focused on climate signals (namely trends in precipitation, mean daily temperature and potential evapotranspiration) and the second group included land use changes such as bark beetle outbreak. Effect of both groups was quantified and compared with baseline simulation across all basins.

The model proved the long-term shifts in runoff seasonality, driven by the air temperature rise, and apparent across the mountain ranges. The seasonal runoff changes are marked by the shift of spring snowmelt toward an earlier season and a decline in spring flows. The second aspect of the changing seasonality is an earlier and prolonged period of summer low flows.

The results proved the dominancy of climate change as a main factor of runoff alteration, acting in large scale patterns, despite the local variations in physiography and land use.