WorkBlock 4 - Extremes: Frequency, Severity and Scale

Co-ordinators: Henny van Lanen (Wageningen University) and Lena M. Tallaksen (University of Oslo), Version 0.1, October 2007

The main objective of Workblock 4 is to advance our knowledge of the impact of global change on hydrological extremes, including spatial and temporal patterns of droughts and large-scale floods.


The 2003 drought in Europe; approximately spatial extent (left) and streamflow drought in River Gardon (Pont du Gard), France, August 2003 (right; Photo by H.A.J. van Lanen).

Work Block 4 (WB4) Extremes will based on an assessment of global datasets and climate model outputs at different scales, analyse our ability to model 20th century hydrological extremes, i.e. droughts and large-scale floods, and subsequently predict the characteristics of future, 21st century extremes. The latter includes the potential impact on hydrological extremes due to global change, i.e. climate change and anthropogenic influences like abstraction and land use change. Time series of historical extremes will be analysed and possible causes for observed trends and changes assessed using both stochastically and physically based modeling tools. WB4 aims at a better process understanding, identification of physical indicators at different spatial scales (river basin to globe), and the development of quantitative analyzing tools that enables possible changes in frequency, severity and scale of hydrological extremes and their related uncertainty to be assessed.

Extended summary of Work Block 4: Extremes: Frequency, Severity and Scale
Hydrological extremes will be assessed at three scales: the global (50 km), the regional (10 km) and the river basin scale (1 km) (grid size is given in brackets). At the global and regional scales, the input will come from WB1 and WB3, which will deliver gridded daily time series at the global and regional scales based on Land Surface Hydrological Models (LSHM) and off-line global hydrological models (GHM). The ability of these models to simulate 20th Century large-scale hydrological extreme events will be assessed by comparison with observed extremes (both temporal and spatial aspects will be evaluated). The knowledge gained will be used to improve the simulation of hydrological extremes events within LSHMs, including the prediction of future 21st Century droughts and large-scale floods. Data for these analyses will stem from re-analysis datasets, observed daily time series and simulated time series of gridded runoff, river flow (through routing), groundwater and soil moisture derived from the LSHMs.

At the river basin scale more detailed modeling studies will be performed using river basin hydrological models (RBHM), which are distributed, physically based models. Four case study basins in Norway, Czech Republic, Slovakia and Spain have been identified based on their relevance for certain sectors (e.g. water supply for irrigation or hydropower). At this more detailed scale, hydrological processes causing extreme events and their time-space variability will be explored. A principal aim of the river basin studies is to evaluate the regional-scale model output from the LSHMs in terms of ability to simulate extreme events.

The main outcome of WB4 Extremes is improved knowledge on the frequency, severity and scale of i) historical extremes, including possible causes for observed trends/changes in the extremes, and ii) future extremes, including an assessment of the sensitivity in the extremes to climate change and anthropogenic influences (detection and attribution). Relevant results, particularly related to drought, will be provided as input to WB6, which addresses the vulnerability of global water resources.

Droughts and large-scale floods are partly treated separately due to their different nature in terms of generating processes and characteristic features, like the time and spatial scales over which they operate. For droughts, important aspects include the spatial extent of the event (climate control), the patchiness/variability within the affected area (catchment control), the dynamics of the event (growth and decay), and possible recurrent patterns in space. For large-scale floods it is important to consider the climate and catchment processes influencing the development of hydrological response over large areas (e.g. large-scale rainfall patterns and snowmelt processes). For both floods and droughts the link to large scale climate drivers will be investigated as well as the occurrence of multiple (synchronous) events around the world.

The work block is structured in three work packages:

  1. WP4.1 Detection and attribution of extremes in the 20th century will detect and analyse historical droughts and large-scale floods in the 20th century using both observed and simulated time series from re-analysis data.
  2. WP4.2 Indices and tools for extremes and uncertainty is devoted to the development of indices and tools to detect hydrological extremes and to assess the uncertainty in the predicted extremes.
  3. WP4.3 Likely frequency, severity and scale of future extremes will provide an analysis of drought and large-scale floods based on the global and regional scenarios for the 21st century (WB3) as compared with the simulations for the 20th century.

Work Performed and Results achieved during Year 2 (February 2008 – 31 January 2009)
The focus of the second year of Work Block 4 has been on finalisation of river basin datasets, including a comprehensive metadata catalogue for the WATCH river basins that are included in WB4. These datasets are used to study processes generating drought, including the propagation from meteorological drought into hydrological drought and spatial and temporal patterns of drought at the river basin scale.

Methodologies for analysing the space-time development of drought and large-scale floods at the regional scale have been advanced. About 500 time series of streamflow data from small undisturbed river basins from the European Water Achieve have been collected and updated to the year 2004/05. Methodologies to link hydrological extremes and weather types have been developed. A first overview of physical indicators and a methodology for change detection in time series of gridded hydrometeorological data have been published (WATCH Technical Reports). An approach for the propagation of uncertainty in the chain climate-hydrology-water resources has been elaborated. The Twinned Ensemble Experiment (TEE) has been developed to quantify the impact of external drivers of climate change on the expected frequency of occurrence of extreme weather events, in particular floods.

Achievements have included: