18 Computation of Hydrological Data
405
from convective clouds.
Spatial resolution. The resolution of the data in space is correlated with the area
of the river basin under consideration. Geostationary satellite systems have a spatial resolution between 2.5 km by 2.5 km and 8.0 km by 8 km, while the spatial
resolution of available polar orbiting satellite systems lies between 10m by 10m
and 1.0 km by 1.0 km, or even larger. The unfavourable spatial resolution of Meteosat and GOES implies that these systems can be considered only for larger
catchment areas (>2.500 km 2 ).
Spectral resolution. GOES, NOAN A VHRR, Meteosat and other meteorological
satellite systems transmit their images in the visible (0.4 J.Ull- 1.1 J.Ull) and thermal
infrared band (10.5 J.Ull- 12.5 /lm) through the "atmospheric windows". Furthermore, Meteosat transmits images in the water vapour absorption band (5.1 J.Ull -
7.1 J.Ull).
On the basis of brightness characteristics of satellite images taken in the visible
spectral band or temperature characteristics taken in the thermal infrared band it is
possible to distinguish precipitating clouds from those without precipitation (Barrett and Martin, 1981). Non-precipitating high, thin ice cirrus clouds are very
transparent (low brightness) in the visible spectral region, while in the thermal
infrared region they are very bright (very cold). Furthermore, it is well known that
two convective precipitating cloud cells of comparable size yield different precipitation amounts, depending on the mean relative humidity of the cells' environment. A mean relative humidity is defined as the arithmetic mean of the relative
humidity values in different atmospheric layers.
Using the water vapour spectral band of the Meteosat satellite system it is possible to obtain information on the humidity conditions of atmospheric layers situated
above a certain level. For a given temperature profile an increase in humidity leads
to a decrease of the transmittance towards space. For example, under tropical atmospheric conditions the 600 mb-Ievel should be considered as the lower limit of
the layer while the 300 mb-Ievel corresponds to the upper limit of the layer of the
relevant water vapour contribution function (Poc et aI., 1983). An existing relationship between radiance in the water vapour channel and mean relative humidity
can be used for the assessment of the state of the humidity in the near environment
of convective clouds.
2. Data processing
In order to adapt the spectral data obtained from the satellite imagery for hydrological purposes a number of preparatory efforts has to be made:
conversion of the geographical coordinates of the catchment area under consideration into image pixel coordinates;
normalization of the satellite images in the visible spectral band for sun-angle
if the visible information is to be used;
405
from convective clouds.
Spatial resolution. The resolution of the data in space is correlated with the area
of the river basin under consideration. Geostationary satellite systems have a spatial resolution between 2.5 km by 2.5 km and 8.0 km by 8 km, while the spatial
resolution of available polar orbiting satellite systems lies between 10m by 10m
and 1.0 km by 1.0 km, or even larger. The unfavourable spatial resolution of Meteosat and GOES implies that these systems can be considered only for larger
catchment areas (>2.500 km 2 ).
Spectral resolution. GOES, NOAN A VHRR, Meteosat and other meteorological
satellite systems transmit their images in the visible (0.4 J.Ull- 1.1 J.Ull) and thermal
infrared band (10.5 J.Ull- 12.5 /lm) through the "atmospheric windows". Furthermore, Meteosat transmits images in the water vapour absorption band (5.1 J.Ull -
7.1 J.Ull).
On the basis of brightness characteristics of satellite images taken in the visible
spectral band or temperature characteristics taken in the thermal infrared band it is
possible to distinguish precipitating clouds from those without precipitation (Barrett and Martin, 1981). Non-precipitating high, thin ice cirrus clouds are very
transparent (low brightness) in the visible spectral region, while in the thermal
infrared region they are very bright (very cold). Furthermore, it is well known that
two convective precipitating cloud cells of comparable size yield different precipitation amounts, depending on the mean relative humidity of the cells' environment. A mean relative humidity is defined as the arithmetic mean of the relative
humidity values in different atmospheric layers.
Using the water vapour spectral band of the Meteosat satellite system it is possible to obtain information on the humidity conditions of atmospheric layers situated
above a certain level. For a given temperature profile an increase in humidity leads
to a decrease of the transmittance towards space. For example, under tropical atmospheric conditions the 600 mb-Ievel should be considered as the lower limit of
the layer while the 300 mb-Ievel corresponds to the upper limit of the layer of the
relevant water vapour contribution function (Poc et aI., 1983). An existing relationship between radiance in the water vapour channel and mean relative humidity
can be used for the assessment of the state of the humidity in the near environment
of convective clouds.
2. Data processing
In order to adapt the spectral data obtained from the satellite imagery for hydrological purposes a number of preparatory efforts has to be made:
conversion of the geographical coordinates of the catchment area under consideration into image pixel coordinates;
normalization of the satellite images in the visible spectral band for sun-angle
if the visible information is to be used;