Seminar Ozean, Eis, Atmosphäre
Seminar Ozean, Eis, Atmosphäre |
|
Termin: 05.02.2002
Referent/in:
T. Hewison, UK Met Office
Title: HOW REALISTIC ARE ATMOSPHERIC ABSORPTION MODELS AT MILLIMETRE WAVELENGTHS?
To fully exploit data from satellite sounding instruments,
such as AMSU-B, accurate radiative transfer models are required
to calculate radiances from a known atmospheric state.
These models give similar results near the water vapour absorption
lines at 22.235 and 183.31 GHz, but differ at intermediate frequencies,
where emission is dominated by the water vapour continuum.
This paper reports the results of field campaigns used to
validate the models of Liebe and Rosenkranz, MPM89, MPM93 and MIT98.
During 1999-2001, the Met Office conducted a series of experiments
using instruments on a C-130 aircraft aimed at validation of
radiative transfer models in clear air.
Flights took place in a range of atmospheric conditions from
the Tropics to the Arctic, and included profiles measuring pressure,
temperature and humidity from 15m to 8km altitudes.
The accuracy of the model validation is limited by uncertainty in
the humidity profile. A number of different sensors were compared to
establish an accurate input profile for the models.
These included aircraft instruments (chilled mirror hygrometer,
Lyman-alpha total water probe and fluorescence water vapour sensors),
dropsondes fitted with RS90 sensors, and various sensors launched
on nearby radiosonde balloons. Generally, the aircraft Lyman-alpha
sensor was found to be most useful for this experiment,
although this required calibration against an absolute reference.
A Microwave Airborne Radiometer Scanning System (MARSS)
measured up- and down-welling radiation in channels similar to those
of AMSU-B, at 89, 157 and 183 GHz. Careful investigation of its
calibration has shown it can be used during aircraft ascents/descents
to measure the absorption over 40 hPa layers from 1000 to 400 hPa.
This produces a dataset with high vertical resolution and ensures a
representative profile is available from simultaneous measurements by
aircraft sensors. Otherwise, inhomogeneities in the atmospheric humidity
fields can dominate the comparison between measured and modelled absorption.
Results will be presented showing the MPM'89 and MIT'98 models produce
similar results. All models perform well near the 183 GHz water vapour
line, but show a consistent negative bias in window channels
at low temperatures. This can be explained by including a
temperature dependence in the foreign broadened part of the water
vapour continuum.