Evaluation of the distribution of subtropical free tropospherichumidity in AMIP-2 simulations using METEOSAT water vaporchannel data
In the framework of the Atmospheric ModelIntercomparison Project (AMIP) phase 2, we haveestablished a diagnostic of the free tropospheric humidity(FTH) distribution using METEOSAT data over the 1984–1995 period for 14 climate models. The methodology ofevaluation follows a two step ‘‘model-to-satelli...
Published in: | Geophysical Research Letters |
---|---|
Main Authors: | , , |
Other Authors: | , , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
HAL CCSD
2005
|
Subjects: | |
Online Access: | https://hal.science/hal-00159814 https://hal.science/hal-00159814/document https://hal.science/hal-00159814/file/Brogniez_et_al-2005-Geophysical_Research_Letters-2.pdf https://doi.org/10.1029/2005GL024341 |
Summary: | In the framework of the Atmospheric ModelIntercomparison Project (AMIP) phase 2, we haveestablished a diagnostic of the free tropospheric humidity(FTH) distribution using METEOSAT data over the 1984–1995 period for 14 climate models. The methodology ofevaluation follows a two step ‘‘model-to-satellite’’ approach.First the raw METEOSAT ‘‘Water Vapor’’ radiances aresimulated from the model profiles of temperature andhumidity using the RTTOV-7 radiative transfer model.Second, the radiances are converted into FTH using thesame coefficients as in the satellite product offering a directcomparison. The analysis is focused on the dry subtropicalareas observed by METEOSAT: the Eastern Mediterraneanand the tropical South Atlantic Ocean. Most of the modelsreproduce the observed seasonal cycle both in terms ofphasing and magnitude, despite an overall moist bias. A fewmodels are in close agreement with the satellite data. Themagnitude of the satellite estimated inter-annual variability isalso generally captured by models. Again, a small subsetof models shows close agreement with the observations.This comparison suggests general improvements of themodels with respect to the AMIP-1 simulations. |
---|