1 / 36

Regional-Scale Climate Simulations with GEM Model

Explore diagnostic results from GEMCLIM simulations – Part 1. Learn about climate model configurations, physics, and comparisons with ERA40 data. Follow-up with insights on North-American and European simulations in Part 2.

cklatt
Download Presentation

Regional-Scale Climate Simulations with GEM Model

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. www.ec.gc.ca GEM day – June 12, 2007 Ouranos, Montréal Diagnostic results from regional-scale climate simulations with the GEM model: Part 1 – GEM Global Bernard Dugas Meteorological Research Division Environment Canada in collaboration with A. Zadra, K. Winger and P. Vaillancourt

  2. The model GEMCLIM is the climate-mode version of the GEM forecast model used operationally by Environment Canada. GEMCLIM provides an estimate of the forecast model’s long term bias as it will generally use configurations that are close that of the operations. It is also being used by our seasonal forecast system. The main difference between GEM and GEMCLIM is probably the latter’s extensive post-processing climate diagnostics. The upcoming CRCM5 will be based on the GEM dynamical core and several projects are already underway at UQÀM to evaluate the model’s behaviour with its native physics. In this context, the CCCma physics is also being ported to the GEM dynamical core by RPN and CCCma personnel. All runs discussed here were conducted with the same GEMCLIM library, version 3.2.1. They all started in January 1978 and ended in March 2004 including an 11 month spinup. The exception is the global uniform 0.5 “Meso-Global” which is a set of 26 Dec-Jan and June-July simulations started from initial conditions provided by another global uniform 1.0 degree GEMCLIM simulation.

  3. Physics • Kain-Fritsch deep and CONRES shallow convections • Sundqvist large scale condensation • Li-Barker short- and long-wave Correlated K radiation (different from operations) • ISBA Land-Surface Scheme • McFarlane (1987) gravity-wave drag • Lott & Miller low-level blocking SST and Sea-ice conditions are provided by AMIP2 1 degree monthly means for the whole period. The AMIP2-recommended Liang & Wang (1995) monthly ozone climatology is also used. Northern Hemisphere winter model results are compared with their corresponding ERA40 time means.

  4. SGMIP2 NA SGMIP2 EU 304x204 Global 79x110 Core Meso-Global 720x360 Only one out of every five grid points are shown MODEL GRIDS

  5. Zonal mean Temperature (-ERA40) (C) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  6. 500hPa GZ (-ERA40) (dm) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  7. Zonal mean Zonal Wind (-ERA40) (m/s) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  8. 200hPa Zonal Winds (-ERA40) (m/s) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  9. Surface Winds (-ERA40) (m/s) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  10. 200hPa Temperature (-ERA40) (C) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  11. Surface Temperature (-ERA40) (C) SGMIP2 North America SGMIP2 Europe GEMCLIM Meso-Global

  12. Zonal Mean Precipitation (mm/day)

  13. SGMIP2 NA, DJF 79-03SGMIP2 EU, DJF 79-03 Meso-Global, JAN 79-04ERA40, DJF 79-01 JAN_79-04

  14. Part 1 Summary The three simulations generally produce very similar (and good) results in the large scale although there are some indications of localized sensitivity to changes in resolution, both in the negative and positive sense: That is to say that the uniform high-resolution model does not always out-perform the variable-resolution models when we compare them with ERA40 re-analysis time means This ends Part 1 of our GEMCLIM diagnostics presentation. Part 2, coming up immediately, will discuss another subset of GEMCLIM simulations, concentrating on regional North-American and European aspects.

  15. www.ec.gc.ca GEM day – June 12, 2007 Ouranos, Montréal Diagnostic results from regional-scale climate simulations with the GEM model: Part 2 – GEM-LAM As provided by Ayrton Zadra Meteorological Research Division Environment Canada

  16. EXPERIMENTSdiagnostic period: 1978-2002(3) CMOS 2007 Related talks / posters Storm track climatologies from GEMCLIM – K. Winger et al.C05-3DP.11 Evaluation of cloud-radiation interaction in the Canadian GEM model using ARM observations – D. Paquin-Ricard et al.C05-3B5.2 Issues arising from the transferability of two Canadian regional climate models to non-native domains – Z. Kothavala et al.C05-3C5.6 An evaluation of cloud and radiation processes simulated by GEM-LAM for the Arctic SHEBA year – D. Simjanovski et al.C05-3DP.3

  17. First experiment Numerics • 3x2 MPI domains & 4 OpenMP threads • 1800s time step • 6 horizontal diffusion Grid • 0.5° uniform, rotated 140x120 grid • Includes 10-point pilot area and 10-point blending area at boundaries (1 in 5 shown here) • 53 vertical hybrid levels; top 10 hPa Physics • Kain-Fritch deep convection • Sundquist large-scale condensation • ISBA land-surface scheme • Li & Barker Correlated-K radiation • McFarlane gravity-wave drag • Lott & Miller low-level blocking Pilot information • atmosphere: ERA40 1957-2002 (2.5°, every 6h) • oceans/lakes: AMIP2 (1°, monthly) • no spectral nesting, but temperature piloted at model top Results for 1978-2002 presented

  18. T_surf (deg C) DJF JJA 500hPa GZ (dam) JJA DJF

  19. DJF JJA 1000hPa spec. hum. (g/kg) Precip (mm/day) JJA DJF

  20. Summary of N. America/ERA40 experiment • > Provides a reference GEM-LAM simulation of the climate over N. America: • documentation (paper in preparation) • driving data for other simulations (e.g. hydrology models) • data for regional climate studies (e.g. student projects) • statistics of model behavior (also for NWP development) • > Issues currently under investigation: Simulated climate too dry and warm, mainly in the summer and over central N. America; possible deficiencies in • geophysical fields (e.g. roughness length, land cover) • physical parametrizations • surface scheme • boundary layer • vertical diffusion • radiative transfer • convection

  21. Second set of experiments N. America • local extension of • SGMIP2 experiment • - two 26-year, 0.5° GEM- • LAM simulations: North- • America and Europe • (same high-resolution • core areas of SGMIP2 • simulations, same period) • - lateral boundary • conditions: 1.5° global • uniform-resolution GEM • simulation (same physics) Europe

  22. WINTER 79-03 SGMIP2 NA LAM NA Surface Temperature – ERA40 [C] LAM NA SGMIP2 NA Total precip – Xie-Arkin [mm/day]

  23. SUMMER 79-03 SGMIP2 NA LAM NA Surface Temperature – ERA40 [C] LAM NA SGMIP2 NA Total precip – Xie-Arkin [mm/day]

  24. WINTER 78-03 SGMIP2 EU LAM EU Surface Temperature – ERA40 [C] LAM EU SGMIP2 EU Total precip – Xie-Arkin [mm/day]

  25. SUMMER 78-03 SGMIP2 EU LAM EU Surface Temperature – ERA40 [C] LAM EU SGMIP2 EU Total precip – Xie-Arkin [mm/day]

  26. Summary of limited-area versus variable-resolution • experiments over N. America & Europe • > comparison of climatologies: • remarkably similar in winter • some differences in summer (LAM slightly closer to analyses) • > cost/benefit favors LAM approach over both domains (note: not necessarily true for other domains, e.g. tropics) • > main weaknesses (e.g. too warm/dry regions, excessive precip. over mountains, etc.) most likely related to deficiencies in geophysical fields and physical parametrizations

  27. www.ec.gc.ca GEM day – June 12, 2007 Ouranos, Montréal … thank you / merci

  28. in case anyone asks for more results…

  29. 200hPa Winds (-ERA40) (m/s) DJF JJA

  30. DJF 850 hPa Winds (-ERA40) (m/s)

  31. JJA 850 hPa Winds (-ERA40) (m/s)

  32. From: Comparison of Cloud Microphysics between GEM and ARM-SGP Observations. By Danahe Paquin-Ricard, Colin Jones and Paul Vaillancourt. Submitted to WGNE Blue Book 2007.

  33. Distribution of ISR and DLR 3-hourly radiation fluxes from RCMs and observations. The period 15-21 UTC is analyzed due to cloud observations only being available during sunlight: • Winter ISR all sky, • Winter ISR cloud free, • Winter ISR cloudy, • Winter DLR all sky, • Winter DLR cloud free, • Winter DLR cloudy, • Summer ISR all sky, • Summer ISR cloud free, • Summer ISR cloudy, • Summer DLR all sky, • Summer DLR cloud free, • Summer DLR cloudy. From: The Simulated Surface Radiation Budget over North America in a Suite of Regional Climate Models by Marko Markovic, Colin Jones, Paul Vaillancourt and Dominique Paquin. Submitted to WGNE Blue Book 2007.

More Related