Axel Hutt;
H. Reich;
R. Faulwetter;
A. Rhodin;
M. Weissmann;
C. Schraff;
R. Potthast
German Meteorological Service
The 8th EnKF Data Assimilation Workshop
Vertical localisation of zero cloud impact SEVIRI infrared data using the KENDA ensemble data assimilation system
Talk:
Montreal18_hutt.pdf
Infrared brightness temperatures (BTs) from geostationary satellites provide detailed information about the cloud distributions. More effective use of this information in modern data assimilation systems has the potential to greatly improve the forecast accuracy for high impact weather events by producing a more accurate initial state in sensitive regions. Specifically water vapor tends to be one of the least accurate variables in initialization datasets due to a lack of in situ observations.
As a first step to attack the all-sky problem, we consider zero-cloud impact observations that exhibit large brightness temperatures neglecting high clouds. Moreover, the application of the Localised Ensemble Transform Kalman Filter (LETKF) requires to localise vertically satellite data. Typically the weighting function of the respective radiation transfer model indicates the corresponding localisation height and radius. We discuss how to choose well the localisation height based on the Jacobians with respect to the temperature (Ht) and the relative humidity (Hq).
In this presentation, we discuss results from ongoing efforts to assimilate zero cloud impact water vapour channel brightness temperatures from the SEVIRI sensor onboard the MSG satellite in the Kilometer Scale Ensemble Data Assimilation (KENDA) system considering the radiative transfer model RTTOV10. The spatial domain of the COSMO model considered in this study covers most of central Europe with 2.8 km resolution, with observations assimilated once per hour. We show in detail how a modified Jacobian Hq improves the vertical localisation and reveal a positive impact on the first guess departure statistics of temperature and relative humidity during a cycled data assimilation experiment in May and June 2016.
Infrared brightness temperatures (BTs) from geostationary satellites provide detailed information about the cloud distributions. More effective use of this information in modern data assimilation systems has the potential to greatly improve the forecast accuracy for high impact weather events by producing a more accurate initial state in sensitive regions. Specifically water vapor tends to be one of the least accurate variables in initialization datasets due to a lack of in situ observations.
As a first step to attack the all-sky problem, we consider zero-cloud impact observations that exhibit large brightness temperatures neglecting high clouds. Moreover, the application of the Localised Ensemble Transform Kalman Filter (LETKF) requires to localise vertically satellite data. Typically the weighting function of the respective radiation transfer model indicates the corresponding localisation height and radius. We discuss how to choose well the localisation height based on the Jacobians with respect to the temperature (Ht) and the relative humidity (Hq).
In this presentation, we discuss results from ongoing efforts to assimilate zero cloud impact water vapour channel brightness temperatures from the SEVIRI sensor onboard the MSG satellite in the Kilometer Scale Ensemble Data Assimilation (KENDA) system considering the radiative transfer model RTTOV10. The spatial domain of the COSMO model considered in this study covers most of central Europe with 2.8 km resolution, with observations assimilated once per hour. We show in detail how a modified Jacobian Hq improves the vertical localisation and reveal a positive impact on the first guess departure statistics of temperature and relative humidity during a cycled data assimilation experiment in May and June 2016.
Thanks to McGill's Atmospheric and Oceanic Sciences department for hosting this web-site.