Sub-Millimetre Measurements of Upper Tropospheric Humidity
Humidity and clouds in the upper troposphere are important factors for the Earth's radiative balance.
Consequently a weakness in global climate models is that very little is known about the concentration of water vapour or the distribution and properties of clouds in this altitude region.
The balloon sondes that are regularly launched around the globe show very poor agreement in the upper troposphere and available satellite data have problems connected to spatial resolution and the occurrence of clouds.
The frequent cloud cover restricts satellite applications of visible and infrared to down-looking geometries, resulting in low vertical resolution and measurements most sensitive to the middle troposphere.
Observations in the microwave region are less sensitive to cloud scattering and therefore limb sounding can be performed to give improved vertical resolution in the upper troposphere.
The sub-millimetre radiometer onboard the Odin satellite, Odin-SMR, performs limb sounding at frequencies around 500 GHz since its launch in 2001.
The observations have sensitivity down to 10 km, but tangent altitudes inside the troposphere have up to now been ignored due to cloud scattering. This effect can now be simulated by a new radiative transfer software and first retrievals of tropospheric quantities are possible.
The thesis presents the steps taken towards a first retrieval of upper tropospheric humidity and ice cloud properties.A thorough investigation of the Odin-SMR calibration at high brightness temperature has been performed. The results show that the systematic calibration error is low, and that a retrieval of humidity from the tropospheric spectra is achievable.
A method to retrieve upper tropospheric humidity from the lowest tangent altitudes has been realised, including a cloud correction scheme, used for spectra with apparent cloud signature. The products of the retrieval are the all-weather humidity field and ice cloud signal. Together with an accompanying method to derive ice cloud properties, the
retrieved ice cloud signal can be translated into ice water content with additional altitude information.