Large ensemble climate and hydrology simulations for studies of extreme events

04 maart 2019

Extreme hydrological events such as droughts and floods can have devastating effects on people and nature. Researchers from KNMI and Utrecht University have worked together to create a unique dataset of combined climate and hydrological model simulations that can be used to investigate these impactful events. The first results were recently published in Geophysical Research Letters.

To protect societies or ecosystems against the effects of extreme hydrological events it is important to know what kind of events to expect. Because extreme events are rare by definition, long time series of data are required for such investigation. In absence of such time series, statistical methods are often used to extrapolate information on observed extreme events to infer information about unobserved events. However, there is a risk of misrepresenting complex or compound events, especially when multiple physical processes can cause extreme events.

New simulations with KNMI’s global climate model are comprised of 2000 years of simulated weather conditions. These data were used to simulate 2000 years of global hydrology, which makes it possible to investigate very extreme events without relying on statistical models. These large ensemble hydrological model simulations with a physical model are unique and make it possible to start new lines of research.

In the published study, a technical comparison of the statistical approach to the new large ensemble approach is made. It is shown that the traditional statistical methods fail to describe extreme events in some river basins. An empirical method based on the new large ensemble dataset provides more accurate estimates in these cases (Figure 1). Therefore, projections of the expected impacts of climate change on extreme hydrological events can be made with more precision (Figure 2). In the future, the dataset will be used to investigate the physical processes leading to extreme droughts and floods in more detail.

K van der Wiel, N Wanders, FM Selten, MFP Bierkens (2019): Added value of large ensemble simulations for assessing extreme river discharge in a 2 °C warmer world. Geophysical Research Letters. In press.

A figure showing the return period of discharge in the Amazon River
Figure 1: Analysis of extreme high discharge in the Amazon River. Left: GEV-fits based on 100 years of data. Middle: GEV-fits based on 2000 years of data. Right: Empirical distribution estimate based on sampling of 2000 years of data. Figure adapted from
A map showing significant changes in river discharge between present-day and a 2°C-warming scenario
Figure 2: Estimated change in the 1-in-100 year drought based on river discharge. Percentage change between a 2°C-warming scenario and present-day conditions, only statistically significant changes are shown. Figure adapted from Van der Wiel et al. (2019)

Read more: