1 / 16

University of Arizona 1.8km WRF Verification

University of Arizona 1.8km WRF Verification. 2Lt Erik Neemann Weather Operations Officer 30 Apr 08. Purpose. To analyze data from University of Arizona 1.8km WRF model for use in 25 OWS Forecast Process Provide feedback for use in model improvements

soo
Download Presentation

University of Arizona 1.8km WRF Verification

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. University of Arizona1.8km WRF Verification 2Lt Erik Neemann Weather Operations Officer 30 Apr 08

  2. Purpose • To analyze data from University of Arizona 1.8km WRF model for use in 25 OWS Forecast Process • Provide feedback for use in model improvements • Study included forecasts for winds, temperatures, and stability (surface-based CAPE) • Additional purpose of deriving tool to approximate wind gust speed from model sustained winds

  3. Methodology • Compared actual observations to 1.8km WRF forecasts • GFS and NAM output from 12z model runs • Three 25 OWS AZ forecast locations used: • Davis-Monthan AFB (KDMA) • Ft. Huachuca (KFHU) • Luke AFB (KLUF)

  4. Winds • Model wind forecasts compared to 18z and 00z observations only • Data was thrown out if observed gusts or sustained winds were at less than 15 knots • If gusts were reported, gusts were used; if no gusts were reported, sustained winds were used) • Results are from 13 Dec 07 to 30 Apr 08 • KDMA: 38 obs • KFHU: 58 obs • KLUF: 26 obs

  5. Winds • Speed error determined by average difference in observed and forecast speed • Ratio of observed speed to forecast speed used to determine “gust coefficient” • Direction error determined by absolute value of difference between observed and forecast wind direction

  6. Wind Speed (knots)

  7. Wind Speed Trends • Davis-Monthan had the best at model forecasts of wind speed, while Luke had the worst • Average for all locations was an underestimation of about 10.5 knots • Both DM and Ft. Huachuca observed/forecast ratios around 2:1 while Luke was closer to 4:1

  8. Wind Direction (degrees)

  9. Wind Direction Trends • Forecasts most accurate for Ft. Huachuca, and worst at Luke AFB • Overall, the models did fairly well for all three locations with error generally less than 35 degrees

  10. Temperature • Compared model forecasts for Max/Min temperature to observed daily Max and Min • Number of cases limited to when both GFS and NAM model runs were available from 13 Dec 07 to 30 Apr 08 (78 days) • Absolute value or error used to examine accuracy • Average error used to determine potential model bias

  11. Min Temperature (Celsius)

  12. Min Temperature Trends • Min temperatures did the best and DM, and the worst a Luke • GFS more accurate at DM and FHU; both models about the same at Luke • Cold biases at DM and FHU; strong warm bias at Luke • Absolute error for all locations was about 1.5° C

  13. Max Temperature (Celsius)

  14. Max Temperature Trends • Max temperature did the best at Luke and DM, while performing poorly at Ft. Huachuca • Both models had similar accuracy • Warm biases DM and Luke, strong cold bias at FHU • Absolute error for all locations was 1.5° C

  15. Stability • Stability results inconclusive due to scarcity of positive CAPE at any location during time period • More robust dataset expected during summer months

  16. Conclusion • NAM model forecast wind direction well for all locations • Wind speed gusts may be applicable for DM and Ft Huachuca with a corrective adjustment, but performed poorly at Luke • Max temps most reliable at DM and Luke using NAM model

More Related