Stochastic downscaling precipitation temperature and wind fields in high spatial and temporal resolution for hydrodynamical and hydrological modeling

• Applicant: Professor Dr.-Ing. András Bárdossy
• Subject Area: Hydrogeology, Hydrology, Limnology, Urban Water Management, Water Chemistry, Integrated Water Resources Management
• Term: from 2008 to 2013
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 101148628

• Final Report Year: 2012

Final Report Abstract

Large lakes like Lake Constance will react to climate change. The issue of assessing the reaction is twofold: (1) specifics about meteorological conditions in the future are uncertain and (2) lakes are complex systems and the processes and dynamics within them are not fully understood. Therefore, simply taking the output of climate models, converting it to fit the local context (downscaling) and use that as input for lake models, is too short-sighted for a reasonable estimation of possible impact. For the example of downscaling wind, breakpoint analysis of reanalysis data has shown inhomogeneities which have to be dealt with. The homogenisation of wind speeds by using a seasonal quantile-quantile transformation yielded reasonable results. While wind with high resolution is important for some aspects of hydrodynamics, it is not necessarily the most important input when dealing with long term climate change. We came to the conclusion that for lakes especially, a change of meteorological conditions has to encompass more than one variable. In an example we showed that lake reactions are more drastic when a change in air temperature is accompanied with a statistically equal change of the other meteorological variables. This so called balanced weather is produced by a multivariate weather generator that was developed within the project. It generates synthetic time series that exhibit properties set by the user. These can be changed mean temperature, increased variability, a combination of changed mean temperature and variability or to follow specific temperature time series. Although changes are defined in terms of temperature, all variables are changed according to how the dependencies between temperature and the other variables were in the past. A tool like this helps to explore the lake’s reactions on different but plausible input and by doing this, to gain a deeper process understanding. This can then form the background on which to interpret consequences of higher carbon emissions. The advancements in guiding the weather generation process made it possible to include detailed information from climate model output.

Publications

• Empirical Downscaling of Windfields for Hydrodynamic Modeling of Lakes, EGU General Assembly 2010, volume 12, (2010), p.11593
  Schlabing, D., and Bárdossy, A.
  
• A Vector-Autoregressive “co-shiftable” Weathergenerator for Hydrodynamic Modeling of Lakes, EGU General Assembly 2011, volume 13, (2011), p. 11632
  Schlabing, D., Eder M. M., Frassl M. A., Rinke, K. and Bárdossy, A.
  
• Climate sensibility of a large lake - a scenario study using a 3D hydrodynamic model and a statistical weather generator, EGU General Assembly 2011, volume 13, (2011), p.10796
  Eder M. M., Schlabing, D., Frassl M. A., Rinke, K. and Bárdossy, A.
  
• Advances in estimating the climate sensibility of a large lake using scenario simulations, EGU General Assembly 2012, volume 14, (2012), p.12840
  Eder M. M., Schlabing, D., Frassl M. A., Rinke, K. and Bárdossy, A.
  
• Simulating the effect of meteorological variability on a lake ecosystem, EGU General Assembly 2012, volume 14, (2012), p.13352
  Frassl, M. A.; Schlabing, D.; Eder, M. M.; Rothhaupt, K.-O.; Rinke, K.
  
• Stochastic Downscaling for Hydrodynamic and Ecological Modeling of Lakes, EGU General Assembly 2012, volume 14, (2012), p.12222
  Schlabing, D., Eder M. M., Frassl M. A., Rinke, K. and Bárdossy, A.