Odin-SMR observations of the tropical upper troposphere
Patrick Eriksson(1), Mattias Ekström(1), Bengt Rydberg(1) and Donal Murtagh(1)
(1) Chalmers University of Technology, Hörsalsvägen 11, SE-41296 Göteborg, Sweden
Water vapour and clouds in the upper troposphere (UT) represent a major uncertainty for prediction of the future climate. This is because these constituents have a crucial impact on the radiative balance, but are today poorly treated in climate models. This in turn is caused by a lack of relevant measurement data. In short, the global observation systems of today fail to provide results that can be used for further improvement of climate models. Main considerations for satellite sounding of water vapour in UT are achievable vertical resolution and cloud penetration capability. These requirements are probable best addressed by microwave limb sounding. The sub-mm radiometer onboard the Odin satellite, Odin-SMR, was designed for strato- and mesospheric research and performs limb sounding at frequencies around 500 GHz since the launch in February 2001. The observations have sensitivity down to about 10 km, but spectra with tangent altitudes inside the troposphere have been ignored due to unknown impact of cloud scattering. This effect can now be rigorously simulated by a new radiative transfer software (ARTS) and first retrievals of tropospheric quantities have been performed.
The possible tropospheric retrieval products of Odin-SMR are water vapour and cloud ice content, above 10 km. A first retrieval scheme has been developed, where only spectra with tangent altitudes below 9km are used. In this way spectra can be inverted individually, uncertainties in sideband filter characteristics can be ignored, and the presence of clouds is comparably easily determined. This retrieval scheme gives the (all weather) mean relative humidity between 10 - 14km, and column of cloud ice above 10 km. The main water vapour retrieval uncertainties are compensation for cloud scattering and spectroscopic information. Reliable validation data for tropospheric UT humidity are very sparse, but a first comparison with MOZAIC data showed that systematic errors are at least below 20%. Retrieved cloud ice columns have been shown to be in general agreement with corresponding ECMWF field, while large deviations were found in comparison to a climate model. The main retrieval issue for cloud ice is the necessary assumption on particle size distribution, a problem shared with, and probably more difficult for, all other existing remote sounding techniques.
In comparison to visible and IR measurements, the impact of clouds is here considerably weaker and this has two advantages: 1. Humidity is measured with a much smaller interference by clouds. 2. Estimates of cloud ice represent bulk properties (not biased towards the cloud top ). The reference data set for UT humidity is also UARS MLS, that operated around 200 GHz where the cloud impact is even smaller. Since 2004 these observations are being continued by EOS MLS. The final quality of the MLS and Odin-SMR UT data should be similar, but with a a much less dense data set for Odin-SMR. However, Odin fills partly the gap between the UARS and AURA missions and gives an independent view of the tropical UT. The MLS and SMR cloud ice retrievals are an important complement to the CloudSat/CALIPSO mission. The latter performs lidar and radar observations that are most sensitive to smallest and largest ice particles, respectively. The passive sub-mm measurements give a lower spatial resolution, but the cloud signal is here more directly related to the total ice mass.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,