Harmonization of Meteosat First and Second Generation Datasets for Fog and Low Stratus Studies

Operational weather satellites, dating back to 1970s, currently provide the best basis for climatological investigations, such as an analysis of changes in cloud cover. Because clouds are highly dynamic time, temporally high-resolution data from geostationary orbit preferred order take variations diurnal cycles into account. For studies, a consistent dataset space and time is mandatory, but not yet available. Ground-based point measurements various parameters, ceiling, visibility, type often sparsely spread inconsistent, making it difficult derive reliable spatio-temporal information over large areas. The Meteosat program has generally provided suitable Europe since 1977, different spatial, spectral, radiometric resolution instruments individual including early-years calibration uncertainties, makes harmonization necessary finally series applicable any kind study. In this study, machine learning-based approach been employed generate long-term with high extensive coverage by First Generation (MFG) Second (MSG) satellite datasets (1991–2020). A random forest (RF) regressor trained on overlap period (2004–2006), where both generation (MFG MSG) available predict MFG Water Vapour (WV) Infrared (IR) channels brightness temperature (BT) values based MSG channels. aim study synthesize MVIRI SEVIRI series. results indicate good match synthesized original mean absolute error 0.7 K WV model 1.6 IR model, out-of-bag (OOB) R² score 0.98 models. Based models, scenes years 2006 2020. homogeneity generated analyzed. harmonized will be applied continuous fog low stratus (FLS) occurrence scale 30 years.