by Hermann Asensio
Numerical weather prediction models need a number of geographical localized datasets like the orographic hight or plant cover, the so called external parameters provide these datasets for the numerical models.
This document provides an overview to the current required external parameters for the numerical weather prediction models COSMO and GME and the used raw datasets. Further alternatives to the currently used datasets and procedures are discussed.
A brief description of the current used datesets to determin the external parameters for the numerical weather prediction models COSMO and GME is given here.
The GLOBE dataset from the National Geophysical Data Center contains globally the orographical hight of the land surface in a resolution of 30 arcseconds. The dataset is available at DEM
Differences from the GLOBE dataset to the GTOPO30 dataset can be particularly found in the orographic hight of Greenland. The GLOBE dataset is used to determine following external parameters:
The GLC200 database (Global Landcover 2000 Database) is provided by the Joint Research Center of the European commission. Except for Antarctica data for plant characteristics are given for the whole earth with a resolution of 1 km. The dataset is based on NDVI measurements with the SPOT4 satellite. Links: publications and GLC200
The GLC2000 database is used to determine following external parameters:
The following table shows the association of the GLC2000 Dataset to the external parameters of GME.
| No. | Description | z0 (m) | max. FP (%) | max LAI (-) | root depth (m) | min RS (m/s) | surface emissivity (-) |
|---|---|---|---|---|---|---|---|
| 1 | Tree Cover, broadleaved, evergreen | 1.0 | 80 | 5 | 1 | 175 | 0.996 |
| 2 | Tree Cover, broadleaved, deciduous, closed | 1.0 | 90 | 6 | 1 | 240 | 0.990 |
| 3 | Tree Cover, broadleaved, deciduous, open | 0.15 | 80 | 4 | 2 | 240 | 0.993 |
| 4 | Tree Cover, needle-leaved, evergreen | 1.0 | 80 | 5 | 0.6 | 500 | 0.996 |
| 5 | Tree Cover, needle-leaved, deciduous | 1.0 | 90 | 5 | 0.6 | 500 | 0.990 |
| 6 | Tree Cover, mixed leaf type | 1.0 | 90 | 5 | 0.8 | 350 | 0.993 |
| 7 | Tree Cover, regularly flooded, fresh water | 1.0 | 80 | 5 | 1 | 350 | 0.996 |
| 8 | Tree Cover, regularly flooded, saline water | 1.0 | 80 | 5 | 1 | 350 | 0.996 |
| 9 | Mosaic: Tree cover / Other natural vegetation | 0.20 | 80 | 2.5 | 1 | 300 | 0.985 |
| 10 | Tree Cover, burnt | 0.05 | 50 | 0.6 | 0.3 | 300 | 0.950 |
| 11 | Shrub Cover, closed-open, evergreen | 0.20 | 80 | 3 | 1 | 225 | 0.985 |
| 12 | Shrub Cover, closed-open, deciduous | 0.15 | 80 | 1.5 | 2 | 225 | 0.993 |
| 14 | Sparse Herbaceous or sparse Shrub Cover | 0.05 | 50 | 0.6 | 0.3 | 110 | 0.950 |
| 15 | Regularly flooded Shrub and/or Herbaceous Cover | 0.05 | 80 | 2 | 0.4 | 110 | 0.992 |
| 16 | Cultivated and managed areas | 0.07 | 90 | 3.3 | 1 | 180 | 0.990 |
| 17 | Mosaic: Cropland / Tree Cover / Other natural vegetation | 0.25 | 80 | 3 | 1 | 200 | 0.990 |
| 18 | Mosaic: Cropland / Shrub or Grass Cover | 0.07 | 90 | 3.5 | 1 | 150 | 0.990 |
| 19 | Bare Areas | 0.05 | 5 | 0.6 | 0.3 | 150 | 0.950 |
| 20 | Water Bodies | 0.0002 | 0 | 0 | 0 | 150 | 0.991 |
| 21 | Snow and Ice | 0.01 | 0 | 0 | 0 | 150 | 0.9999 |
| 22 | Artificial surfaces and associated areas | 1.0 | 20 | 1 | 0.6 | 150 | 0.960 |
The soiltype is determinend of the Digital Soil Map of the World (DSMW) from the FAO. The resolution of the dataset is 5 arcminutes. The FAO database is used to determine following external parameters:
For the GME model the Normalized Differential Vegetation Index (NDVI) from a NDVI climatology of the NASA from the SEAWiFS sensor is used. The data are stored in following fields:
As lower boundary condition in the soil model (TERRA) the climatology of the 2m temperature is given as external parameter for the GME model. In the COSMO model the temperature of the lowest soil level is given as boundary data from the driving (global) model.
The Climate Research Unit of the University of East Anglia (CRU) supplies a Climatology of the 2m temperature above land from observations for the normal baseline period from 1961 to 1990 in a resultion of 0.5 degrees.
For data gaps (over the oceans for some islands) ERA40 data are used for the
2m temperature.
From reanalysis of the European Center for Midrange Weather Forecasts (ECMWF ERA40)
are provided on a gaussian grid with a grid spacing of 1.125 degrees.
a number of further databases exist, which are suitable as raw data for external parameters. Such data sources are listed here briefly, making no claim to be complete.
NOAA provided a global relief model of the earth, which contains next to the land surface hight also ocean depth in a resolution of 1 arc minute. See Global Relief
Two versions are available, one with the hight of the ice sheet and one with the height of the bedrock underneath the ice sheets (Antarctic and Greenland).
The U.S. Geological Service offers the freely available global dataset for orographic height GTOPO30 with a resolution of 30 arc seconds, see GTOPO30
In older model configurations of COSMO the GTOPO30 height data have been used, but due to problems in Greenland with the GTOPO30 dataset the GLOBE dataset is used for the current configurations COSMO-EU and COSMO-DE.
From the GTOPO30 dataset the U.S. Geological Service derived the dataset Hydro1k for hydrological applications with corrected hight information, slope, aspect etc. under HYDRO1k
the derived quantities like slope are better generated directly for the target grid from the raw data.
The dataset of the Shuttle Radar Topography Mission (STRM) is free available with a resolution of 3 arcseconds under EarthExplorer (STRM DTED Level 1). Higher resolutions are partly free available (for the USA) otherwise subject to license conditions (STRM DTED Level 2). The data cover more than 80 % of the land mass between 60 degrees north and 56 degrees south. Especially in the mountain regions are data gaps.
The data gaps could be filled with GLOBE, GTOPO30 or further (national) datasets, in a consistent way this has been done for the ETOPO1 dataset (1 arcminute resolution).
The National Snow and Ice Data Center (Boulder) offers height information for Antarctica in 1 km resolution in the dataset "Radarsat Antarctic Mapping Project Digital Elevation Model Version 2". See Radarsat These data have also been used in the ETOPO1 dataset.
For the sea model Flake data with sea depth are required. D. Mironov has a list with such data.
CORINE Land Cover 2000 is a dataset of the EU with a resolution of 250 m. GME, COSMO-EU and COSMO-DE use the GLC2000 Data.
The Global Land Cover Characteristics (GLCC) dataset have been compiled from AVHRR-Data and are provided by the U.S. Geological Service for free.
A somewhat older global dataset (1983) called "Metthews vegetation" is provided by NASA under Land datasets
A number of derived products about land use and vegetation like Leaf Area Index or albedo (surface reflectance) from the MODIS Instrument (Moderate Resolution Imaging Spectroradiometer) on the satellites Terra and Aqua is available, detailed information under MODIS
In order to validate the MODIS data direct measurements of the LAI have been compiled to one dataset, see Leaf Area Index.
GLCC, CORINE and GLC2000 are based on AVHRR, Spot4 and Landsat data. Whether the MODIS product are suitable for the generation of external parameters is subject to further research; eventually more current AVHRR data from NOAA and MetOP satellites are better appropriate. Also the LSA SAF (land surface analysis satellite applications facility) could provide useful data.
The database ECOCLIMAP (MeteoFrance) can be used to generate external parameters, especially the aggregation of external parameters for different ecosystem within one gridelement is supported. See EcoCliMap
A rather new (2008) global database is the "Harmonized World Soil Database" with a resolution of 30 arcseconds. This database also contoins Datasets of SOTER, ESD, WISE and the FAO DSMW. See HWSD
The use of this database to generate external parameters for numerical weather prediction models should be considered.
The project SOTER (Soil and Terrain Resources) compiles soil data in a "Global Soil and Terrain Database" (World-SOTER), but the coverage is not global at present. See SOTER
The European Soil Database (ESD) from JRC contains several database, including the SGDBE (Soil Geographical Database of Eurasia), see ESDB and ESDB-2
Some datasets with soil data can be found at ISRIC WISE (World Inventory of Soil Emission Potentials): ISRIC
The climate classification according to Köppen is used in ECOCLIMAP for the partition of the ecosystems. See Climate maps
| HadCRUT3 | The HadCRUT3 is a data update from January 2006 by the CRU (Climate Research Unit, University of East Anglia). With observations of the 2m temperature above land and data from the Hadley-Center for the sea surface temperature (SST) a global dataset with a resolution of 5 degrees has been compiled for the base period 1961-1990. Data available at Temperature |
| CRU CL 1.0 | With a resolution of 0.5 degrees (30 arc minutes) the CRU dataset CRU CL 2.0 provides 2m temperature over land for the base period 1661-1990 based on observations (no SST). Data available at High resolution gridded datasets and observational climatologies |
| CRU CL 2.0 | With a resolution of 10 arc minutes the CRU dataset CRU CL 2.0 provides 2m temperature over land for the base period 1661-1990 based on observations (no SST). This dataset should replace the "old" CRU climatology. Data available at gridded datasets |
A model-derived global distribution of Aerosol Optical Thickness can be found at GACP datasets
As Vector Map Level 0 the National Imagery and Mapping Agency's (NIMA) provides a global dataset with vector data with different thematic topics like transport, political boundary, hydrography and shorelines in a public domain license.
The University of Hawaii maintains a "GSHHS - A Global Self-consistent, Hierarchical, High-resolution Shoreline Database". See GSHHS. This databes is available as a supplement to the GMT (Generic mapping tool).
| Parameter | GRIB name | Unit | used raw dataset |
|---|---|---|---|
| geopotential | FIS | m2s-2 | Globe |
| land cover | FR_LAND | 1 | GLC2000 (Globe, DSMW) |
| standard deviation of subgrid scale orogr. height | SSO_STDH | m | Globe |
| anisotropy of topography | SSO_GAMMA | 1 | Globe |
| angle betw. principal axis of orogr. and global E | SSO_THETA | 1 | Globe |
| mean slope of subgrid scale orography | SSO_SIGMA | 1 | Globe |
| surface roughness | Z0 | m | GLC2000, Globe |
| soil texture | SOILTYP | 1 | DSMW |
| surface emissivity (long wave) | EMISS_RAD | 1 | GLC2000 |
| root depth | ROOTDP | m | GLC2000 |
| vegetation (plant cover) | PLCOV | % | GLC2000 |
| ground fraction covered by plants (vegetation p.) | PLCOV_MX | 1 | GLC2000 |
| ground fraction covered by evergreen forest | FOR_E | 1 | GLC2000 |
| ground fraction covered by deciduous forest | FOR_D | 1 | GLC2000 |
| leaf area index (vegetation period) | LAI_MX | 1 | GLC2000 |
| plant resistance | plant res | s2m-1 | GLC2000 |
| annual max. of norm. differential vegetation index | NDVI_MAX | 1 | SEAWiFS |
| (monthly) normalized differential vegetation index | NDVI | 1 | SEAWiFS |
| (monthly) proportion of act.value/max. norm.diff.veg.index | NDVIRATIO | 1 | SEAWiFS |
| temperature 2 m above ground (climatological mean) | T_2M_CL | K | CRU, ERA 40 |
| Parameter | GRIB name | Unit | used raw dataset |
|---|---|---|---|
| geometrical height | HSURF | m | Globe |
| geopotential | FIS | m2s-2 | Globe |
| land cover | FR_LAND | 1 | GLC2000 (Globe, DSMW) |
| standard deviation of subgrid scale orogr. height | SSO_STDH | m | Globe |
| anisotropy of topography | SSO_GAMMA | 1 | Globe |
| angle betw. principal axis of orogr. and global E | SSO_THETA | 1 | Globe |
| mean slope of subgrid scale orography | SSO_SIGMA | 1 | Globe |
| surface roughness | Z0 | m | GLC2000, Globe |
| soil texture | SOILTYP | 1 | DSMW |
| surface emissivity (long wave) | EMISS_RAD | 1 | GLC2000 |
| root depth | ROOTDP | m | GLC2000 |
| vegetation (plant cover) | PLCOV | % | GLC2000 |
| ground fraction covered by plants (time of rest) | PLCOV_MM | 1 | GLC2000 |
| ground fraction covered by plants (vegetation p.) | PLCOV_MX | 1 | GLC2000 |
| ground fraction covered by evergreen forest | FOR_E | 1 | GLC2000 |
| ground fraction covered by deciduous forest | FOR_D | 1 | GLC2000 |
| leaf area index (time of rest) | LAI_MN | 1 | GLC2000 |
| leaf area index (vegetation period) | LAI_MX | 1 | GLC2000 |