Hydrological change in mountain basins: understanding observed changes and robust modeling
Final Report Abstract
Mountain areas play an important role for the water supply of the adjacent lowlands, yet they are particularly vulnerable to future climate change. Insights on the behavior of hydrologic systems under climate variations may be gained from investigating past observed changes in hydrologic fluxes and the relations to their drivers. However, only a few studies exist that examined changes in catchment evapotranspiration. While several studies have shown that hydrological models do not perform well when applied to periods with climate conditions that differ from those during model calibration, the causes of these model problems are not well understood. The aim of this project was to improve our understanding of the behavior of hydrologic systems in mountain regions under climate change, based on observed climatic variations in the recent past. The first objective was to analyse changes in catchment evapotranspiration and their drivers over the last 40 years. The second objective was to advance our understanding of the causes of poor model performance in a changing climate. Austria was selected as study region due to the strong climate signal over the last decades and the excellent database. Analyses based on the water balance show that catchment evapotranspiration significantly increased in 60 % of the 156 catchments (p ≤ 0.05) with an average increase of 4.9 ± 2.3 % decade^−1 (± standard deviation) over 1977–2014. Pan evaporation based on 22 stations has, on average, increased by 6.0 ± 1.0 % decade^−1. Reference evapotranspiration over the 156 catchments estimated by the Penman- Monteith equation has increased by 2.8 ± 0.7 % decade^−1. Estimates of a modified reference evapotranspiration accounting for changes in stomata resistance based on changes in a satellite derived vegetation index indicate that increases in vegetation activity led to a similar increase in reference evapotranspiration as changes in the climate parameters. Increases in evapotranspiration were positively correlated with increases in precipitation, with a sensitivity of 0.22 ± 0.05 mm y^−2 increase in evapotranspiration to 1 mm y^−2 increase in precipitation. A synthesis of the analyses suggests that 43 ± 15 % of the observed increase in catchment evapotranspiration may be attributed to increased atmospheric demand and available energy, 34 ± 14 % to increased vegetation activity, and 24 ± 5 % to increases in precipitation. In order to analyze the causes of poor hydrological model performance in a transient climate, I revisited a study that demonstrated the inability of a conceptual semi-distributed HBV-type model to correctly simulate the observed discharge response to increases in precipitation and air temperature for catchments in Austria. Using the original model, trends between simulated and observed discharge differed by 95 ± 50 mm y^−1 per 35 years. Hypotheses for the possible causes of the mismatch between the observed and simulated changes in discharge were evaluated using simulations with modifications of the model. The results suggest that the model problems in a transient climate were largely caused by neglecting changes in vegetation activity and problems in the precipitation data, explaining 36 ± 9 mm y^−1 per 35 years and 39 ± 26 mm y^−1 per 35 years, respectively, of the difference between simulated and observed discharge trends. Extending the calibration period from 5 to 25 years, including annually aggregated discharge data or snow data in the objective function, or estimating reference evapotranspiration with the Penman-Monteith instead of the Blaney-Criddle approach had little influence on the simulated discharge trends. The results of this project have important implications for hydrological modelling in a transient climate. Hydrological models used in climate impact studies need to be thoroughly tested and improved for their ability to simulate long-term variations in evapotranspiration, and such models should account for the effects of possible climate-induced vegetation changes on evapotranspiration.
Publications
- (2017): Änderungen der Evapotranspiration in alpinen Gebieten über die letzten 40 Jahre (Changes in evapotranspiration in alpine catchments over the last 40 years), Tri-national Workshop - Hydrologic processes in the high mountains, 15.- 17. November 2017, Obergurgl, Austria
Duethmann, D., J. Parajka, G. Blöschl
- (2018): Factors controlling alterations in the performance of a runoff model in changing climate conditions. J. Hydrol. Hydromech., 66, 1-12
Sleziak, P., Szolgay, J., Hlavčová, K., Duethmann, D., Parajka, J., Danko, M.
(See online at https://doi.org/10.2478/johh-2018-0031) - (2018): Marked increases in catchment evapotranspiration in an Alpine region and their possible drivers. Geophysical Research Abstracts, Vol. 20, EGU2018- 13787
Duethmann, D., G. Blöschl
- (2018): Why has catchment evaporation increased in the past 40 years? A data-based study in Austria. Hydrol. Earth Syst. Sci., 22, 5143-5158
Duethmann, D., Blöschl, G.
(See online at https://doi.org/10.5194/hess-22-5143-2018) - (2019): Zunahmen der Gebietsverdunstung und deren Ursachen – neues Wissen aus einer datenbasierten Studie (Increases in catchment evapotranspiration and their causes – new insights from a data-based study). Tag der Hydrologie, 28./29. März 2019, Karlsruhe, Deutschland
Duethmann, D., G. Blöschl
- (2020): Investigating the reasons for poor model performance in a changing climate. EGU General Assembly 2020, EGU2020-487
Duethmann, D., G. Blöschl, J. Parajka
(See online at https://doi.org/10.5194/egusphere-egu2020-487) - (2020): Why does a conceptual hydrological model fail to predict discharge changes in response to climate change? Hydrol. Earth Syst. Sci., 24, 3493-3511
Duethmann, D., Blöschl, G., Parajka, J.
(See online at https://doi.org/10.5194/hess-24-3493-2020)
