62
Chapter 8
The initialization of soil moisture fields because soil wetness can have a
strong impact on the forecast of low level atmospheric quantities
(Mahfouf 1991, Bouttier et al. 1993).
Satellite remote sensing data in the visible and infrared spectral domains
2.
can bring a significant input into operational algorithms. Two examples of
exploitation of satellite data in Numerical Weather Prediction models can be
summarized as follows:
1. The use of AVHRR/NOAA data for the specification of vegetation
properties in relation to the ISBA scheme (Noilhan and Planton, 1989) in
the ARPEGE French system, and
2. The use of METEOSAT visible and infrared data in the SEBAL
algorithm developed to retrieve the soil wetness in the KNMI modeling
system (Hurk et al. 1997).
2.
A HIGH RESOLUTION VEGETATION MAP FOR
OPERATIONAL PURPOSES
The implementation of a land surface scheme in operational numerical
weather prediction model is a long and difficult task. Bazile and Giard
(1996) describe the various aspects of the implementation of the ISBA
scheme in the ARPEGE computing system. The low complexity of ISBA
(only one surface energy balance, and two soil layers for heat and water
diffusion) makes it a good candidate for daily forecasts. The scheme has
been widely calibrated at the local scale for a large number of land uses,
using data collected during numerous large field experiments conducted
during the last decade. See Noilhan and Mahfouf (1996) for a review of the
calibration of ISBA, and the chapter by Jochum et al. in this book for a
description of land-surface experiments. In addition, a sequential
assimilation scheme for soil moisture based on optimum interpolation has
been developed (Giard and Bazile 1997, 1999). The implementation of ISBA
requires the specification of surface parameters prescribed from a vegetation
map and a soil map. Up to now, the most widely used data base in the
meteorological community is the classification of Wilson and HendersonSellers (1985) at the global scale, but with a coarse resolution of one by one
degree (latitude and longitude). This resolution is not satisfactory for NWP
models operating with a grid box size of a few km. For this reason, a
methodology for generating vegetation parameter maps over Europe with a
resolution of 1 km has been developed by Champeaux and Le Gléau (1995).
The method takes full advantage of the regular remote sensing information
provided by the AVHRR/NOAA. Fist of all, an annual archive has been
created by Météo-France over Europe since the early 1990s. AVHRR
Chapter 8
The initialization of soil moisture fields because soil wetness can have a
strong impact on the forecast of low level atmospheric quantities
(Mahfouf 1991, Bouttier et al. 1993).
Satellite remote sensing data in the visible and infrared spectral domains
2.
can bring a significant input into operational algorithms. Two examples of
exploitation of satellite data in Numerical Weather Prediction models can be
summarized as follows:
1. The use of AVHRR/NOAA data for the specification of vegetation
properties in relation to the ISBA scheme (Noilhan and Planton, 1989) in
the ARPEGE French system, and
2. The use of METEOSAT visible and infrared data in the SEBAL
algorithm developed to retrieve the soil wetness in the KNMI modeling
system (Hurk et al. 1997).
2.
A HIGH RESOLUTION VEGETATION MAP FOR
OPERATIONAL PURPOSES
The implementation of a land surface scheme in operational numerical
weather prediction model is a long and difficult task. Bazile and Giard
(1996) describe the various aspects of the implementation of the ISBA
scheme in the ARPEGE computing system. The low complexity of ISBA
(only one surface energy balance, and two soil layers for heat and water
diffusion) makes it a good candidate for daily forecasts. The scheme has
been widely calibrated at the local scale for a large number of land uses,
using data collected during numerous large field experiments conducted
during the last decade. See Noilhan and Mahfouf (1996) for a review of the
calibration of ISBA, and the chapter by Jochum et al. in this book for a
description of land-surface experiments. In addition, a sequential
assimilation scheme for soil moisture based on optimum interpolation has
been developed (Giard and Bazile 1997, 1999). The implementation of ISBA
requires the specification of surface parameters prescribed from a vegetation
map and a soil map. Up to now, the most widely used data base in the
meteorological community is the classification of Wilson and HendersonSellers (1985) at the global scale, but with a coarse resolution of one by one
degree (latitude and longitude). This resolution is not satisfactory for NWP
models operating with a grid box size of a few km. For this reason, a
methodology for generating vegetation parameter maps over Europe with a
resolution of 1 km has been developed by Champeaux and Le Gléau (1995).
The method takes full advantage of the regular remote sensing information
provided by the AVHRR/NOAA. Fist of all, an annual archive has been
created by Météo-France over Europe since the early 1990s. AVHRR