Priority Project "1DVAR":
Assimilation of Satellite Radiances with 1DVAR and Nudging

Last updated: September 2008

Project leader: Reinhold Hess (DWD)

Description

Radiance data from polar orbiting and geostationary satellites (ATOVS, AIRS/IASI, and SEVIRI) are used for LME (and its derivates) in order to improve forecast quality of LME and also quality of boundary values for subsequent models as LMK. Assimilation of level 1 radiance data is per- formed in two steps: Firstly, profiles (including screen level and skin values) of temperature and humidity are retrieved from the radiances using a 1DVar analysis approach. Secondly, these profiles are assimilated with the nudging scheme of the COSMO similarly to other conventional data.

Future domains of COSMO installations will include large parts of data sparse areas (e. g. north Atlantic for LME) for conventional data. The use of satellite radiances especially over sea has the potential to provide essential information in order to improve analyses and subsequent forecasts on the regional domain.

Project subtasks

Task 1

Prepare technical implementation for MSG data

• Pre-processing of data: gross quality control, formatting into net CDF
• Reading satellite data and auxiliary ASCII input
• Implement 1D-Var core: interpolate LM model first guess
• Writing satellite feedback file including model first guess
• Provide retrievals to nudging scheme as other conventional data

Task 2

Prepare technical implementation for ATOVS using developments for MSG

Task 3

Adapt technical implementation and developments for AIRS/IASI

Task 4

Setup monitoring: run implementation on a day by day basis and generate time series of observed minus background radiances for bias correction, evaluate statistics for selecting viable channels and adjusting observation errors

• at SIM-ARPA for MSG data
• at IMGW for ATOVS/AMSU-A data
• at DWD for AIRS/IASI data

Task 5

Cloud / rain detection and bias correction

• Validate NWC-SAF cloud detection for the use of quantitative assimilation of MSG data, compute and apply bias correction for MSG data
• Implement and validate rain and ice-detection for ATOVS/AMSU-A data,compute and apply bias correction for ATOVS/AMSU-A data

Task 6

Derive and tune background errors over sea and over land; assess use of MSG data over land

• Ensemble B matrix

Task 7

Validate usage of EMCWF climate values above LM model top for radiative transfer computation and compare it to extrapolation with statistical regression

Task 8

Validate retrievals by comparing them to radiosonde data and to model profiles

Task 9

Tune nudging of 1D-Var retrievals, nudging coefficients, etc (with case studies)