|Title||Assimilation of MIPAS limb radiances in the ECMWF system. Part I: Experiments with a 1-dimensional observation operator.|
|Publication Type||Technical memorandum|
|Year of Publication||2006|
|Secondary Title||ECMWF Technical Memorandum|
|Authors||Bormann, N, Thépaut, J-N|
Emitted infrared limb radiances from the Michelson Interferometer for Passive Atmospheric Sounding MIPAS) are for the first time assimilated directly in a global Numerical Weather Prediction system, using a fast radiative transfer model that assumes local horizontal homogeneity. The impact of the limb radiance assimilation on the resulting analyses and forecasts is assessed by considering the fit to other assimilated observations and by comparing the resulting analyses with independent data. The results demonstrate the feasibility of direct assimilation of limb radiances and highlight how information on stratospheric temperature, humidity, and ozone can be extracted from the radiances directly within the assimilation system. The assimilation of MIPAS radiances leads to considerable changes in the mean stratospheric analyses, without a significant degradation of the fit to other observations used in the assimilation. The assimilation of MIPAS limb radiances corrects temperature biases with an oscillatory structure above 10 hPa in the analyses, and the assimilation leads to a considerable moistening of the stratosphere, typically by 20-40 %. For ozone, the assimilation causes an increase in the tropical ozone maximum, and a reduction of ozone over the poles. The changes to the humidity and ozone fields are retained in the subsequent 10-day forecast. The changes introduced in the analyses are overall supported by independent retrievals of temperature and humidity, whereas results for ozone are more mixed. Analyses with MIPAS radiances agree better with independent ozone sondes or retrievals over the North Polar region, whereas over the tropics the changes to the mean ozone analyses are not supported by other observations. The limb radiance assimilation shows considerable sensitivity to the bias correction applied to the limb radiances, suggesting that the assimilation would benefit from refinements in the bias correction method.