1 / 17

Karl-Göran Karlsson Colin Jones Ulrika Willén Klaus Wyser SMHI, Norrköping Sweden

Use of a high-resolution cloud climate data set for validation of Rossby Centre climate simulations. Karl-Göran Karlsson Colin Jones Ulrika Willén Klaus Wyser SMHI, Norrköping Sweden.

cargan
Télécharger la présentation

Karl-Göran Karlsson Colin Jones Ulrika Willén Klaus Wyser SMHI, Norrköping Sweden

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. Use of a high-resolution cloud climate data set for validation of Rossby Centre climate simulations Karl-Göran Karlsson Colin Jones Ulrika Willén Klaus Wyser SMHI, Norrköping Sweden Presentation at the EUMETSAT Meteorological Satellite Conference, Prague, Czech Republic, 31 May - 4 June 2004

  2. Use of a high-resolution cloud climate data set for validation of Rossby Centre climate simulations Contents: • The cloud datasets from SCANDIA, the Rossby Centre RCA climate model and the ERA-40 project • Results from direct comparisons - area means and mean 2-D distribution of cloudiness • Ambiguities due to contributions from optically very thin clouds • Introduction of a filtering concept based on RTM simulations • Conclusions and future plans

  3. (%) 100 80 60 40 20 0 The SCANDIA cloud climatology from NOAA AVHRR Example for afternoon cloud frequencies in July 1991-2000 Reference: Karlsson, K.-G., 2003:A 10-year cloud climatology over Scandinavia derived from NOAA AVHRR imagery, Int. J. Climatology, 23, 1023-1044. Climatology also available on two CD-ROMs

  4. The Rossby Centre RCA2 climate model • A high resolution (44 km grid, 24 layers) climate simulation model for studies of regional climate change effects (downscaling experiments) • Simulation of present climate using ECMWF analyses as forcing and for lower and lateral boundary conditions • Large-scale condensation scheme by Rasch and Kristjánsson (1998) and Slingo (1987) • Convective parameterisation by Kain and Fritsch (1990) • 17-year climate simulation 1985-2001 • Assuming Maximum-Random cloud overlap Reference: http://www.smhi.se/en/index.html (choose R&D, regional climate modelling)

  5. The ERA-40 cloud data set • ECMWF Re-Analysis data set (second version) • ECMWF T159L60 model version (~130 km horizontal resolution, 40 vertical layers) • Forecasted cloud data set (four six-hour forecasts per day) • Maximum-Random Cloud Overlap Reference: http://www.ecmwf.int/research/era/

  6. Results for the entire Nordic area Time series 1991-2000 of mean monthly total cloud amount (%) for SCANDIA, RCA2 and ERA-40.

  7. Seasonal 2-D distribution for the entire SCANDIA area SCANDIA RCA2 ERA-40 WINTER SPRING SUMMER AUTUMN

  8. We conclude: • Simulated cloud amounts agree quite well with observed amounts concerning the 2-D distribution but amounts are generally higher, especially in winter and over land surfaces. But, is this what we should expect? DEFINITELY NOT!!

  9. Two real cloud situations with two different resulting horizontal cloud coverages (black bar) One and the same model representation of the two real cloud cases Cloud representation in climate models - Finite vertical resolution in models creates ambiguities in interpretation of 2D cloud datasets Conclusion: We should expect modelled cloud amounts to be understimated!!!

  10. Indications of limited cloud detection ability from satellite-retrievals and SYNOP observations Results from the SHEBA year

  11. Conclusion: • These clouds must be filtered out from the RCA data set if they are not detected by SCANDIA! Question: • Are there significant contributions to the RCA-simulated fractional cloud cover from optically very thin clouds? YES!!

  12. Method for filtering optically very thin clouds • Simulate cloudy radiances for AVHRR channels with RTM model • Chosen RTM: Signal Simulator for Cloudy Retrieval (SSCR)- Discrete-Ordinate (DISORT) method described by Nakajima and King (1992) and Nakajima (1995).- Simulation of clouds by inserting them as homogeneous sub-layers with certain specified characteristics (water phase, volume size distribution and particle optical thickness) • Determine the lower detectability limit for SCANDIA • Filter out clouds with optical thicknesses below the SCANDIA detectability limit from the RCA data set

  13. Example of SSCR simulations over land surfaces *) *) Not normalised for sun elevation and scaled with a factor 2.55 Indicating detectability limit close to optical thickness = 1.0

  14. Example of SSCR simulations over ocean surfaces *) Not normalised for sun elevation and scaled with a factor 2.55 Indicating detectability limit close to optical thickness = 1.0

  15. Conclusions - Lessons learnt: • RCA2 cloudiness resembles SCANDIA in the horizontal distribution and in the seasonal cycle. • Significant overestimation of cloudiness over land surfaces is seen for the winter half of the year indicating both dynamical and physical problems • ERA-40 cloudiness show better agreement with SCANDIA but shares the RCA2 problem over land surfaces • RCA2 excessive cloudiness in winter is not accompanied with a positive bias in surface temperatures Complex compensating error structure involving radiation scheme and surface flux parameterisation

  16. Conclusions - Lessons learnt: (cont’d) • However, filtering of optically very thin clouds from model data sets appears necessary for final confirmation of results • The cloud detectability limit for SCANDIA, as indicated by SSCR simulations, appears to lie close to optical thicknesses of 1.0 • Filtering work started based on the filtering of clouds where the average vertically averaged optical depth in the cloud covered fraction of a grid column is below 1.0

  17. Future plans • Filter model data set for optically very thin clouds (thinner than optical thickness 1.0) and repeat the comparison with the SCANDIA climatology • Further filtering attempts exploring the possibility to filter out potentially vertically resolved clouds in RCA - Enables studies of the performance of simulated effects from resolved and non-resolved clouds (e.g., parameterisation of convection) • Development of Climate Monitoring SAF tools to adapt cloud climatological data sets for enabling comparisons with climate models - CM-SAF data simulation tool

More Related