WorkBlock 1 - The global water cycle of the 20th century

Co-ordinator: Graham Weedon, UK Meteorological Office
Deputy co-ordinator: Nicola Gedney, UK Meteorological Office

The main objective of Workblock 1 is to consolidate gridded data sets, improve the hydrological representation of hydrology in hydrological models and investigate the 20th century global water cycle using a combination of models and data.

This work block will determine the global water cycle throughout the 20th century and quantify the uncertainty in the resulting datasets due to a number of sources. It will also detect and attribute changes during the 20th century due to human intervention in the water cycle. This workblock aims to produce global forcing datasets in a common format, these will be used to create a new global forcing dataset at 50 km spatial resolution and sub-daily temporal resolution.

To capture the main physical processes within the global water cycle, land surface models will be improved to represent more hydrological processes. Using these improved models we will:

  • improve spatial and temporal resolutions of available datasets
  • address processes included in the existing numerous Land Surface Hydrological Models or River Basin Hydrological Models
  • try to quantify Uncertainty using ensemble approaches.

The outputs from this work block will be used to help with the understanding of extremes in the hydrological cycle and will attribute these to the respective causes. Attribution of changes in hydrological variables is a multi-stage process. In WB1, we will use a combination of regional and global land-surface models to separate out the contribution of meteorological forcing to observed changes in hydrological variables over the 20th century. In conjunction with the new internally-consistent precipitation datasets developed by the WATCH consortium, diagnosed changes in meteorological forcing will then be subjected to a formal detection and attribution analysis using global coupled climate models run under the ENSEMBLES IP in order to separate out natural from anthropogenic contributions and quantifying the propagation of uncertainties through the hydrological cycle.

The work is divided into 4 workpackages:

  • WP1.1 will concentrate on gridded observed hydrological dataset for the 20th century. In addition validation datasets will also be collected and put into a common data format
  • WP1.2 will develop current land surface models to include hydrological processes such as groundwater, irrigation, dams and river routing. This will ensure land surface models are more consistent with hydrological models
  • WP1.3 will use an ensemble of model runs to produce a 20th century dataset of the global water cycle. The uncertainties from various sources in this dataset will be addressed, including differences in the land surface models, the temporal and spatial scales of the forcing datasets. Additionally the uncertainties from using climate model output for forcing the land surface models will be assessed and used to reduce the uncertainties in the prediction of the global water cycle for the 21st century
  • WP1.4 will investigate the sensitivity of the global water cycle of the 20th century to the impact of human intervention. It will also attempt to detect and attribute changes in the water cycle to specific human activities.

Work Performed and Results achieved during Year 2 (February 2008 – 31 January 2009)
During the second year of WATCH, Work Block 1 has focused on developing validation and forcing datasets in a common format (i.e. netCDF) to enable the global hydrology and land surface models to be run in a consistent way. In conjunction with WB6 the initial forcing data was used in initial runs of land surface and hydrological models.

Work has incorporated additional physical processes and human interactions to improve land surface models and hydrological models and new verification data, such as new satellite-based global maps of land surface temperature and global inundation have been created. Schemes have been developed for incorporating and parameterising: i) groundwater ii) crops and irrigation iii) and dams iv) river routing within various land surface- and hydrological-models.

The water cycle in the 20th century has been partly investigated in terms of precipitation intensities in Europe, cloud parameterisation and water storage and transport in relation to inundation and flooding and a methodology for assessing means and extremes has been developed.