Are we using the right fuel to drive hydrological models? A climate impact study in the Upper Blue Nile
| Autor: | Stefan Liersch, Julia Tecklenburg, Henning Rust, Andreas Dobler, Madlen Fischer, Tim Kruschke, Hagen Koch, Fred Hattermann |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
lcsh:GE1-350
010504 meteorology & atmospheric sciences hydrological models lcsh:T lcsh:Geography. Anthropology. Recreation 0207 environmental engineering 02 engineering and technology lcsh:Technology 01 natural sciences lcsh:TD1-1066 6. Clean water lcsh:G 13. Climate action lcsh:Environmental technology. Sanitary engineering 020701 environmental engineering lcsh:Environmental sciences 0105 earth and related environmental sciences |
| Zdroj: | Hydrology and Earth System Sciences, 22 . pp. 2163-2185. Hydrology and Earth System Sciences, Vol 22, Pp 2163-2185 (2018) |
| Popis: | Climate simulations are the fuel to drive hydrological models that are used to assess the impacts of climate change and variability on hydrological parameters, such as river discharges, soil moisture, and evapotranspiration. Unlike with cars, where we know which fuel the engine requires, we never know in advance what unexpected side-effects might be caused by the fuel we feed our models with. Sometimes we increase the fuel's octane number (bias-correction) to achieve better performance and find out that the model behaves differently but not always as was expected or desired. This study investigates the impacts of projected climate change on the hydrology of the Upper Blue Nile catchment using two model ensembles consisting of five global CMIP5 Earth System Models and ten Regional Climate Models (CORDEX Africa). WATCH forcing data were used to calibrate an eco-hydrological model and to bias-correct both model ensembles using slightly differing approaches. On the one hand it was found that the bias-correction methods considerably improved the performance of average rainfall characteristics in the reference period (1970–1999) in most of the cases. This also holds true for non-extreme discharge conditions between Q20 and Q80. On the other hand, bias-corrected simulations tend to overemphasise magnitudes of projected change signals and extremes. A general weakness of both uncorrected and bias-corrected simulations is the rather poor representation of high and low flows and their extremes, which were often deteriorated by bias-correction. This inaccuracy is a crucial deficiency for regional impact studies dealing with water management issues and it is therefore important to analyse model performance and characteristics, the effect of bias-correction, and eventually to exclude some climate models from the ensemble. However, the multi-model means of all ensembles project increasing average annual discharges in the Upper Blue Nile catchment and a shift in seasonal patterns, with decreasing discharges in June and July and increasing discharges from August to November. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|