Atmos 5210/6210 Group Project
Objectives:
Ensemble modeling systems are powerful tools for weather forecasting, but are typically run at modest resolution, limiting their utility for quantitative precipitation forecasting (QPF) over the western U.S. In this lab, students will use and evaluate forecasts produced by two experimental experimental modeling systems over the western United States. The first are ensemble forecasts run by the Naitonal Center for Atmospheric Research (NCAR) using the Weather Research and Forecasting model (WRF) configured with 3-km grid spacing. The second are downscaled ensemble forecasts based on forecasts produced by North American Ensemble Forecast System that are downscaled using climatological precipitation analyses. After completing this project, students will:- Be able to more effectively use ensemble prediction systems for short- and medium-range for quantitative precipitation forecasting
- Recognize the strengths and limitations of ensemble prediction systems for quantitative precipitation forecasting
- Have greater understanding of the challenges of evaluating ensemble model accuracy and value, including issues related to observational error and uncertainty
Part 1
Using the available (sometimes nearby) climatological data and studies of regional weather and climate, develop and understanding of the topography and climatology of the three forecast sites:- Snoqualmie Pass (SNO30), a precipitation measurement system operated by the Northwest Avalanche Center at 917 m MSL in the central Washington Cascades.
- Central Sierra Snow Lab (CSSL), a precipitation measurement system operated by UC Berkeley at 2098 m near Donner Pass in California's Sierra Nevada.
- Alta-Collins (CLN), a precipitation measurement system operated by Alta Ski Area at 2945 m in the central Wasatch Mountains. Summarize the topography and climatology of each site in 2-3 paragraphs. Include illustrations or figures if desired.
Part 2
By 0900 MST (1600 UTC) each morning from Jan 24 to Feb 8, produce a precipitation forecast for each site that includes a deterministic QPF amount (inches) and a range representing the middle 50% of the probability distribution (also in inches) for the following periods, where Day 1 is the day (UTC) on which you are issuing the forecast:- 1800 UTC Day 1 - 0000 UTC Day 2
- 0000 UTC Day 2 - 1200 UTC Day 2
- 1200 UTC Day 2 - 0000 UTC Day 3
- 0000 UTC Day 3 - 0000 UTC Day 4
- 0000 UTC Day 4 - 0000 UTC Day 5
- 0000 UTC Day 5 - 0000 UTC Day 6
Part 3
After the 14-day forecast period, evaluate the strengths, weaknesses, and utility of the human and ensemble forecasts. This should be done subjectively and objectively. The objective validation should include calculations for each site and forecast period of:- The mean absolute error the deterministic precipitation forecasts and the mean and median from the NCAR ensemble and NAEFS for periods with (1) observed precipitation, (2) forecast precipitation, and (3) either observed or forecast precipitation.
- The bias ratio (i.e., forecast/observed) of the deterministic precipitation forecasts and the mean and median precipitation from the NCAR ensemble and NAEFS
- The fraction of periods in which the observed precipitation fell within the 50% range anticipated by the meteorologists, NCAR ensemble, and NAEFS.
- Any further analysis you feel is necessary to identify the strengths and weaknesses of the modeling system, including additional statistical analysis and in-depth subjective assessment of the forecast discussions and major successes and busts.
Part 4
Provide a summary of your results and what you have learned in either written or oral form in a format TBD in consultation with the instructor.