1 / 61

Radiation and fog in complex topography - Research and operational forecasting with NMM

NOAA/NCEP Washington, Aug. 2005. Radiation and fog in complex topography - Research and operational forecasting with NMM. Z.Janjic, D. Scherer & A. Bott. Mathias D. Müller Institute of Meteorology, Climatology & Remote Sensing University of Basel, Switzerland Mathias.mueller@unibas.ch.

pcarmen
Télécharger la présentation

Radiation and fog in complex topography - Research and operational forecasting with NMM

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. NOAA/NCEP Washington, Aug. 2005 Radiation and fog in complex topography - Research and operational forecasting with NMM Z.Janjic, D. Scherer & A. Bott Mathias D. Müller Institute of Meteorology, Climatology & Remote Sensing University of Basel, Switzerland Mathias.mueller@unibas.ch

  2. WHAZZZUPP ! The forecast system at the University of Basel A radiation parameterization of topographic effects Towards a 3D fog model Coffee and Cookies :-)

  3. Realtime Forecasts / Nesting GFS 00/12 UTC NMM-22 & ETA-22 00/12 UTC 5 grids 72 h NMM-4 03/15 UTC ETA-4 15 UTC NMM-2 15 UTC 45 h

  4. Hardware Weather Forecast & Post Processing Beowulf Cluster Theoretical Performance 386 Gflops1 Master 2 x 2.4 GHz Xeon Cpu's, 2GB Mem1 Fileserver 1.1TB Disk36 Nodes 2 x 2.4 GHz Xeon Cpu's, 2GB MemGigabit Ethernet and SCI interconnect Visualisation & Preprocessing 2 x 3.0 GHz Xeon CPU‘s, 4 GB Mem, Linux OS IDL, 2 TB Disk

  5. NWP at the University of Basel - Flow chart NOAA/NCEP 3D-VAR / GFS Sea-temperatue Snow RS - Bern NDVI Snowmask Meteoblue MeteoSwiss Observations Meteomedia Observations Visualisation Current forecast Archive short-term Archive Initalization phase 1 Verification Kalman Cronos File-Server / Cluster tasks management Forecast Model on Cronos Cluster Final Initialization Post processing 3D-VAR not implemented! EXTERNAL URZ WS-ETA Synchronized communication (MPI) Co VAR Data

  6. NDVI initialization with NOAA-AVHRR CLIMATOLOGY AND REALTIME NDVI CLIMATOLOGY NDVI REALTIME NDVI AUGUST, 13 2003 AGGREGATED TO 4 KM MODEL RESOLUTION

  7. 22 km NMM 4 km NMM Operational Model Statistics Forecast output of 5 Model Grids: 35 GB / 24h Total number of Images produced: ~ 20‘000 / 24h Programming language for all Graphics: IDL Operating System: LINUX Communication and time synchronisation with forecast Models, running on Cronos Cluster. Computing Time: NMM22 ~ 60 min (critical !) NMM4 ~ 90 min NMM2 ~ 4 hours 2km NMM Computing Time of Models is upper limit for Visualisation time

  8. Forecast Visualisation – key points Efficiency of Visualisation Data and Information Reduction / Aggregation (storage and resolution restrictions) (e.g. Smoothing, Skipping, 24h-precipitation) Conversion / Computation of Variables for visualisation from raw model output (e.g. Vorticity, CAPE, Helicity, Streamlines) Precomputation of domain dependent but constant layers (Topographic Shading, Continent outlines,….) Model Coordinate Systems (Lambert, Eta-Grid, …, Lat/Lon) Internal Coordinate System (Geographical Lat/Lon) Projection / Mapping to different Coordinate Systems for Visualisation Scaling: constant, image optimized, constant-optimized for forecast range

  9. Maps – transparent, constant and content Layers

  10. Point Forecasts – Meteograms

  11. Point Forecasts – Tephigrams

  12. Internet – meteoblue.ch

  13. Pictograms – it‘s for children 3. Web-page compilation: 1. Local field statistics: 2. Decision Tree: no rain rain Rain Cloud Cover Temperature ….. Rain Cloud Cover Temperature ….. clouds Rain Cloud Cover Temperature ….. Rain Cloud Cover Temperature …..

  14. 3D visualisations – nice and useless?

  15. A radiation parameterization of topographic effects Mathias Müller, Dieter Scherer

  16. Parameterization features Radiation fluxes from Mesoscale Model Land-Surface Model of Mesoscale Model Considers: Slope and azimuth angle of absorbing surface Shadows Sky-view restriction Negligible computational costs Computation of radiative mean fluxes for model grid cells based on higher resolution DEM Interface between land-surface model and radiative transfer model Relatively easy to include in other models (sandwich-like) Parameterization features

  17. Shortwave radiation - direct & diffuse Radiation fluxes from Mesoscale Model Land-Surface Model of Mesoscale Model

  18. Longwave radiation Radiation fluxes from Mesoscale Model Land-Surface Model of Mesoscale Model

  19. Effects of slope angle, slope aspect and shadow • affects: • direct shortwave

  20. Effects of slope angle, slope aspect and shadow • affects: • direct shortwave

  21. Effects of slope angle, slope aspect and shadow • affects: • direct shortwave

  22. Effects of slope angle, slope aspect and shadow • affects: • direct shortwave

  23. Sky-view factor 0.5 • affects: • diffuse shortwave • longwave

  24. Case Studies Parameterization was tested in the Nonhydrostatic Mesoscale Model (NMM) of NOAA/NCEP (version: Z. Janjic) 220 Alpine stations

  25. 2m Temperature Differences +2 K Sun facing -3 K shadow

  26. 2m Temperature Differences 0.5 +2 K in Valleys • Skyview affects: • diffuse shortwave • longwave

  27. Verification – Clear sky summer conditions (alpine stations) 22 June 2003 Nighttime warming reduces model cold bias (0.5K) Problem of verification: Stations are located in valleys but radiation affects slopes

  28. Verification – overcast conditions (alpine stations) 24 December 2003 Nighttime warming reduces model cold bias

  29. Benefits from subgrid topography doubles the effect ! 4 km Grid with 1 km subgrid topography

  30. Parallel run intercomparison (4 km grid)

  31. Conclusions Radiation fluxes from Mesoscale Model Land-Surface Model of Mesoscale Model Mean radiation flux based on high resolution DEM Modifies temperature in a full physics NMM run up to +/-3 K Reduces nighttime cold bias by about 0.5-1 K No computational costs during time integration Easy to implement, sandwich-like module for mesoscale models

  32. 3D fog with NMM and PAFOG Mathias Müller, Andreas Bott, Zavisa Janjic

  33. NMM_PAFOG Droplet number concentration Liquid water content NMM dynamical framework PAFOG microphysics Condensation/evaporation in the lowest 1500 m is replaced by PAFOG

  34. PAFOG microphysics Supersat. where S is the Supersaturation Assumption on the droplet size distribution : Log-normal function D droplet Diameter Dc,0 mean value of D σc Standart deviation of the given droplet size distribution (σc=0.2)

  35. GRID of NMM_PAFOG 50 x 50 x 45 27 layers in the lowest 1000 m 11 soil layers Thickness(cm): 0.5 0.75 1.2 1.8 2.7 4.0 6.0 10 30 60 100

  36. Boundary conditions for dNc HEIGHT PAFOG TOP 1000 m σc 1000m PAFOG TOP

  37. Nesting GFS NMM-22 NMM_PAFOG GRID: 50 x 50 x 45 (+11 soil layers) dx: 1 km dt: 2s (dynamics) / 10s (physics) CPU: 40 min/24hr on 9 Pentium-4 (very efficient!) NMM_PAFOG NMM-4 NMM-2 15 UTC

  38. 19:00 MEZ (3 hr forecast) PAFOG STANDARD 27 Nov 2004

  39. 22:00 MEZ (6 hr forecast) PAFOG STANDARD 27 Nov 2004

  40. 02:00 MEZ (10 hr forecast) PAFOG STANDARD Accurate sedimentation in PAFOG due to dNc computation. 28 Nov 2004

  41. 08:00 MEZ (16 hr forecast) PAFOG STANDARD 28 Nov 2004

  42. 10:00 MEZ (18 hr forecast) PAFOG STANDARD 28 Nov 2004

  43. qc at 5m height (01:00 MEZ) PAFOG STANDARD

  44. qc at 5m height (06:00 MEZ) PAFOG STANDARD

  45. Nc at 5m height (06:00 MEZ)

  46. Cold air pooling (05:00 MEZ) too cold !

  47. Cold bias problem Z.Janjic

  48. 19:00 MEZ (3 hr forecast) PAFOG STANDARD 27 Nov 2004

  49. 22:00 MEZ (6 hr forecast) PAFOG STANDARD 27 Nov 2004

  50. 02:00 MEZ (10 hr forecast) PAFOG STANDARD Accurate sedimentation in PAFOG due to dNc computation. 28 Nov 2004

More Related