1 / 29

DWD Extramural research

Application of an adaptive radiative transfer parameterisation in a mesoscale numerical weather prediction model. DWD Extramural research. Annika Schomburg 1) , Victor Venema 1) , Felix Ament 2) , Clemens Simmer 1) 1) Department of Meteorology, University of Bonn, Germany

judd
Download Presentation

DWD Extramural research

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Application of an adaptive radiative transfer parameterisation in a mesoscale numerical weather prediction model DWD Extramural research Annika Schomburg1) , Victor Venema1), Felix Ament2), Clemens Simmer1) 1) Department of Meteorology, University of Bonn, Germany 2) University of Hamburg

  2. Outline • The adaptive radiative transfer scheme • General idea • Implementation • Results • 3in1 runs • Single runs • Preliminary new result • Outlook

  3. Adaptive parameterizations • Accurate parameterization • Process-based • Computationally expensive • Fast parameterization • Less processes (statistical) • Typically: biased • Adaptive scheme • Combine: accurate and fast parametrization • Accurate one corrects biases of fast one

  4. Adaptive RT: Spatial scheme • Uses spatial correlations • Update every 2.5 minutes one out of 5x5 columns • For other 24 columns: search for similar column in the vicinity (search region 5x5 pixels) • Similarity index to be minimised:

  5. Implemented in COSMO 4.0 • Adaptive scheme • Called every 2.5 minutes • Reference high-resolution • δ-two-stream approximation (Ritter & Geleyn) • Full field computed every 2.5 min • Comparison • Coarse-scheme COSMO-DE (2x2 columns) • Called every 15 minutes

  6. Diurnal cycle error net flux SW LW RMSD Bias

  7. Diurnal cycle error heating rates SW LW RMSD Bias

  8. Error height profile heating rates SW LW RMSD Bias

  9. Scale dependent errors Surface net flux Atmospheric heating rate

  10. Errors net flux for COSMO-EU SW LW

  11. Physical consistency: LWP SW LW

  12. Consistency: diurnal cycle SW LW

  13. Spread single runs Surface pressure Total precipitation 2m-Temperature

  14. Selection column accurate computation • Optimized pattern: as before in this talk • Global difference: largest difference in full field • Local difference: largest difference in 5x5 regions • Spiral pattern: regular pattern, close together Preliminary new results

  15. Conclusions • Adaptive radiative transfer makes computations more accurate (or efficient) • Employs spatial and temporal correlations in atmosphere  in error fields of simplified computations

  16. Outlook • Develop a temporal spatial adaptive scheme • Improve our results for heating rates • Question: what is a good error measure? • Bias & RMSD • Scales (temporal, spatial) • Locations (layers, regions) • Heating rates, fluxes & PAR • Other parameterizations • Surface module (looking for 2 PhD students) • Aerosols, etc.

  17. References Schomburg, A., V. Venema, F. Ament, and C. Simmer. Application of an adaptive radiative transfer scheme in a mesoscale numerical weather prediction model. Quarterly Journal of the Royal Meteorological Society, accepted 2011. Venema, V.K.C., A. Schomburg, F. Ament, and C. Simmer. Two adaptive radiative transfer schemes for numerical weather prediction models. Atmospheric Chemistry and Physics, 7, 5659-5674, doi: 10.5194/acp-7-5659-2007, 2007. Download: http://www2.meteo.uni-bonn.de/venema/articles/

  18. Errors in the solar heating rates (W m-2) in the LM at the surface for 12.30 h UTC. (a) The two-stream calculation of the solar surface flux is the reference field (b) Cloud cover of low clouds (c) Total cloud cover (d) the 1-h persistence assumption, (e) the adaptive perturbation scheme, (f) the adaptive search scheme. The corresponding errors are shown in the same order in the third row.

  19. The Idea: Adaptive parameterisation Grid points where… Recalculate radiation fluxes with exact scheme calculate error- estimator based on a simple radiation scheme for each grid point …Δ‘large‘ …Δ ‘small‘ Apply „perturbation method“ for surface fluxes Perturbation method:

  20. RMSE perturbation methods

  21. Approach solar cloud free infrared cloud free solar cloudy infrared cloudy • Simple radiation scheme: → Multivariate linear regression • Predictands: • longwave: • shortwave: transmissivity: • Distinction of 4 categories, with different sets of predictors:

  22. Simple radiation scheme Predictors: SOLAR

  23. Simple radiation scheme: Predictors INFRARED

  24. Approach Implementation of adaptive scheme into LM • First tests and configuration on PC (small model domain) • After successfull implementation: exemplary cases on LMK-domain on parallel machine at DWD • Horizontal resolution: 2.8 km • Frequency of call to adaptive scheme: 2.5 min

  25. Approach • Problem: Radiation of 3 separate model runs not comparable due to different evolution of cloud field → Development of a model version „3 in 1“: • Calculation of the radiation fluxes • hourly • adaptive • frequently (every 2.5 min) ... in the same model run • Dynamics only influenced by frequent radiation • Test for 3 summer days characterised with much • convection

  26. Correlation lengths for error fields (15:30 UTC) Hourly Adaptive Hourly Adaptive The covariance functions of the errors in the solar (a) and infrared (b) fluxes at the surface.

  27. Instantaneous RMSE 21June 2004 with adaptive scheme smoother error curves

  28. Smoother developing of model variables with time Surface temperature 21 June 2004 Adaptive approach prevents „wavy“ structure of developing of variables with time

More Related