Energy and Water Cycles in the Climate System ...
17
cloudiness and are, therefore, not very accurate. Over the oceans, both properties are very
stable within small uncertainty limits. More recent operational analyses of the sea surface
temperatures, making use of the multi-spectral capabilities of the NOAA-AVHRR, reach error
limits of about lK and even less. These are excellent accuracies to monitor anomalies in the
surface temperature, which are related to oceanic circulation phenomena.
Thus greater care has been given to the derivation of downward fluxes of short- and longwave
radiative energy which according to some sensitivity studies should be known within limits of
less than 10 Wm- 2 for monthly averages over areas of about 100 x 100 km 2 •
The downward atmospheric radiation has been shown to be determined mainly by the temperature and humidity (water content!) of the lowest layers of the troposphere (up to about
1.000 to 2.000 m), and on the altitude and temperature of cloud bottoms (e.g.: Schmetz, 1989).
Their relative influence increases with decreasing water content; thus it is very large over polar
regions and almost negligible in the tropics. The more operational methods of the work within
the WCRP use therefore operational analyses of the temperature and moisture fields as "pre_
dictors". Kornblueh (1995, private communication) tries an additional estimate of the cloud
bottom height using a simplified assimilation scheme and obtained in case studies an agreement
within 50 m comparing his data with simultaneous lidar measurements (over the Arctic ice).
Fig. 1.10 shows some of Kornblueh's results.
It is believed, that only active remote sensing with lidar and cloud profiling radar, the latter
measuring at 3 or 8 mm wavelength, will improve the situation. The GEWEX Science Steering
Group has therefore established a working group to support and critically review ongoing
activities in Europe, USA and Japan. An example of such measurements, made from ground
is shown in Fig. 1.11.
The downward solar radiation can in principle be estimated using the cloud albedo as an
indicator for the cloud optical thickness (e.g.: Moser and Raschke, 1983). Various procedures
have been developed in the past years and applied to regional and also global data sets (see
contributions of Stuhlmann and Rossow, this volume). They yield uncertainties - as compared
to the often quite inaccurate ground-based data - of 5 to 10% of monthly averages, i.e. about
10 to 20 Wm- 2 • Another, and possibly simpler approach makes use of the vertical profiles of
solar net radiation in each layer of the atmosphere (Li et aI., 1993), to be estimated with the
cloud information from satellite data. But here also some information on the surface albedo is
required (see comments made above). Such retrievals require steady control by measurements
from stations, which are representative for areas of about 10 to 50 km extent, dependent on the
spatial scale of available satellite data. The BSRN-data (e.g. Whitlock et aI., 1995) might be
able to meet these requirements, although the location of the individual stations has not been
chosen according to the representativeness for larger areas.
1.6 Precipitation
(see also sections by DelGenio, Hallikainen, Joe, Noll, Stewart, Simmer)
Precipitation in form of rain or snow falls out of mostly unstable clouds; thus most precipitation
events should be expected in regions of the earth with convective activity and synoptic-scale
activities. But also orography plays a major role to destabilize the flow approaching it and
causes precipitation up to a certain altitude, where clouds cannot carry large amounts of water.
Various global precipitation climatologies have been established earlier on such informations,
on ground-based measurements in inhabitated areas and a variety of ship reports (e.g.: Hulme,
1995; Jager, 1976; Legates, 1995). Precipitation maps show patterns which are closely related to
the instabilities mentioned above. An example is shown in Fig. 1.12. Such still "conventional"
17
cloudiness and are, therefore, not very accurate. Over the oceans, both properties are very
stable within small uncertainty limits. More recent operational analyses of the sea surface
temperatures, making use of the multi-spectral capabilities of the NOAA-AVHRR, reach error
limits of about lK and even less. These are excellent accuracies to monitor anomalies in the
surface temperature, which are related to oceanic circulation phenomena.
Thus greater care has been given to the derivation of downward fluxes of short- and longwave
radiative energy which according to some sensitivity studies should be known within limits of
less than 10 Wm- 2 for monthly averages over areas of about 100 x 100 km 2 •
The downward atmospheric radiation has been shown to be determined mainly by the temperature and humidity (water content!) of the lowest layers of the troposphere (up to about
1.000 to 2.000 m), and on the altitude and temperature of cloud bottoms (e.g.: Schmetz, 1989).
Their relative influence increases with decreasing water content; thus it is very large over polar
regions and almost negligible in the tropics. The more operational methods of the work within
the WCRP use therefore operational analyses of the temperature and moisture fields as "pre_
dictors". Kornblueh (1995, private communication) tries an additional estimate of the cloud
bottom height using a simplified assimilation scheme and obtained in case studies an agreement
within 50 m comparing his data with simultaneous lidar measurements (over the Arctic ice).
Fig. 1.10 shows some of Kornblueh's results.
It is believed, that only active remote sensing with lidar and cloud profiling radar, the latter
measuring at 3 or 8 mm wavelength, will improve the situation. The GEWEX Science Steering
Group has therefore established a working group to support and critically review ongoing
activities in Europe, USA and Japan. An example of such measurements, made from ground
is shown in Fig. 1.11.
The downward solar radiation can in principle be estimated using the cloud albedo as an
indicator for the cloud optical thickness (e.g.: Moser and Raschke, 1983). Various procedures
have been developed in the past years and applied to regional and also global data sets (see
contributions of Stuhlmann and Rossow, this volume). They yield uncertainties - as compared
to the often quite inaccurate ground-based data - of 5 to 10% of monthly averages, i.e. about
10 to 20 Wm- 2 • Another, and possibly simpler approach makes use of the vertical profiles of
solar net radiation in each layer of the atmosphere (Li et aI., 1993), to be estimated with the
cloud information from satellite data. But here also some information on the surface albedo is
required (see comments made above). Such retrievals require steady control by measurements
from stations, which are representative for areas of about 10 to 50 km extent, dependent on the
spatial scale of available satellite data. The BSRN-data (e.g. Whitlock et aI., 1995) might be
able to meet these requirements, although the location of the individual stations has not been
chosen according to the representativeness for larger areas.
1.6 Precipitation
(see also sections by DelGenio, Hallikainen, Joe, Noll, Stewart, Simmer)
Precipitation in form of rain or snow falls out of mostly unstable clouds; thus most precipitation
events should be expected in regions of the earth with convective activity and synoptic-scale
activities. But also orography plays a major role to destabilize the flow approaching it and
causes precipitation up to a certain altitude, where clouds cannot carry large amounts of water.
Various global precipitation climatologies have been established earlier on such informations,
on ground-based measurements in inhabitated areas and a variety of ship reports (e.g.: Hulme,
1995; Jager, 1976; Legates, 1995). Precipitation maps show patterns which are closely related to
the instabilities mentioned above. An example is shown in Fig. 1.12. Such still "conventional"