linked in logo | IS-ENES | Contact

Exploring climate model data

FEWS/NHI hydrological framework

Use of climate model data in FEWS/NHI hydrological framework

Goal: In this use case, climate model data is selected and downloaded from the portal.

Delivery of climate model data from different runs to the FEWS/NHI system. Delft-FEWS, originally Flood Early Warning System, is a operational forecasting system to manage data and models in a real time environment. Delft FEWS retrieves and prepares hydrological and meteorological data for the models. FEWS is able to handle many protocols and data formats, like FTP, Grib and netCDF. Internally an XML model is used for hydrographical data and netCDF for binary data.

One of the models running operationally under FEWS is the National Hydrological Instrument (NHI). The NHI is a model which provides insight into the actual and forecasted states of the surface, ground and soil water in the Netherlands to support decision making during periods of droughts. NHI is driven by measured and forecasted precipitation and evaporation (ECMWF-DET and -EPS). The tool also gives insight in the actual and forecasted water demands. To predict future hydrology in the Netherlands, NHI can be used in combination with climate model data. The NHI model can make use of daily gridded data.

read more on FEWS/NHI on the Deltares site

Data needs:

  • All data on surface level, (2 meter ground level)
  • Parameters: precipitation, potential evaporation, relative humidity, radiation, pressure, temperature, minimum temperature
  • Period: Current time resolution of 6 hours
  • Projection: Data will be delivered in a regular grid according to EPSG:4326 (WGS84 latitude longitude).
  • Different climate scenarios and multiple ensembles.

Additional data needs:

  • Resolution: Current resolution of 100 km is already suitable, but a higher spatial resolution in the order of 10 km is desired on the long term.
  • Space and time resolutions should be in balance
  • Area and location: cutout for the Netherlands is preferred, but not required
  • Data format: netCDF4
  • Data delivery: FTP and OGC services
  • Consistent data format: Within subsequent data deliveries, the data must comply to the same data format, units and metadata.

How to obtain the data

To select the CMIP5 data, matching the criteria as mentioned in data needs, the user can go to Facets use in DataSearch. and then:

  1. To retrieve the files, click on 'Basket' in the main menu (right top of the page)
  2. In the Basket, click on 'get' to retrieve the file

When the user is not logged on to the portal, the user will be prompted to provide its OpenID to the ESG node containing the data. When the user wants to obtain the other variables, the user only has to adjust the Variable dropbox and select files again. Note that not all variables requested are available in 6 hour time resolution. Here the user has to adjust the time frequency as well. To see how this works in practice, see the YouTube movie In a first iteration of preparing this use case, the ESSENCE dataset was pre-processed and provided through OPenDAP. This dataset can be found and previewed in the Catalogs on the EPICIC site (open the 'essence' map to see the individual model runs) or directly access the data using the EPICIC THREDDS catalog.

Sources of Uncertainty

In the transformation from the original Gaussian Reduced Grid to Regular (lat lon) grid, Ferret is used. This transformation may introduce interpolation errors, especially at land-sea transitions. There are no regular grids matching exactly the Gaussian Reduced Grid, i.e. interpolation is unavoidable. This should be well documented and provided to the user of the data. Deltares is aware of this and knows how to deal with it. To study the sensitivity of the FEWS models for climate model data several scenarios will be used as input. Read more on uncertainties.

 

The IS-ENES project has received funding from the European Union's Seventh Framework Programme for research, technological development and demonstration.

Disclaimer