Analysis of MTG‐IRS observations and general channel selection for numerical weather prediction models

IRS (infra-red Sounder) is an infra-red Fourier transform spectrometer that will be on board the Meteosat Third Generation series of future EUMETSAT geostationary satellites and have a unique 4D look at atmosphere. After its planned launch in 2024, it able to measure radiance emitted by Earth top atmosphere using 1960 channels two spectral bands between 680 – 1210 cm-1 (long-wave infra-red) 1600 2250 (mid-wave sampling 0.6 cm-1. It perform measurements over full disk with high spatial temporal resolution 4 km nadir 30 minutes Europe. Thus, huge amount data from present challenges, particularly transmission, storage assimilation into NWP models. To reduce volume data, various methods are available including sampling, principal component analysis channel selection. The latter technique discussed this paper proposing general selection provide Numerical Weather Prediction objective improve essential variables for such as temperature, humidity, skin temperature ozone. This work has required development large observation database takes account main developments techniques, use observation-error covariance matrices or ozone global models example. study performs specific sensitivity observations proposes first 300 allows us error troposphere 48% 65% humidity 17%