Summary of MRF-based statistical quantitative precipitation forecast guidance developed using
the MOS technique
Developed by Mark Shirey
- •12- and 24-h Probabilistic Quantitative Precipitation Forecast (PQPF) guidance has been produced for several cutoffs,
including the accumulation of 0.01, 0.10, 0.25, 0.50, 1.00, and 2.00 inches of precipitation (a cutoff of 3.00 inches is
also available for 24-h PQPF's). A PQPF value can be defined as the probability that precipitation accumulation will
equal or exceed its corresponding cutoff value.
- •Forecasts containing both 12- and 24-h PQPFs are available every 12 hours for projections valid 24 to 156 hours after
0000 UTC (the PoP, or .01 PQPF, is available out to 192 hours).
- •Categorical forecasts will also be available (similar to the QPF system in the current operational NGM MOS text
- •Several MRF forecasts were offered to the regression analysis including relative humidities, precipitation amount, u- and
v- wind components, moisture convergence, various advections, vertical velocities, k index, and relative vorticity.
- •Climatic relative frequencies (valid for the 1992-1999 period) were also offered.
Developmental sample for the cool season (October-March)
- •The developmental sample contained various dates from 1992-1999. For the years of 1992-1996, MRF data was
obtained from NCEP's reanalysis project. This data set provided us with a model run every 5th day during this time
- •For the 97-98, and 98-99 season, all MRF runs available were used.
- •For seasonal stratification, a two week overlap at the beginning and end of the cool season was added.
Test sample (independent data set)
- •During the test development for verification, the last 15 days of each month in the 98-99 season were held out and used
as independent data.
- •All of the verification scores show PQPFs percent improvement over the Brier score of climate. When obtainable, the
newly developed MRF PoP is compared to the current operational PoP. Please note that for regional results, some of the
heavier precipitation amount verification is not shown due to a lack of events to test on.
- •Preliminary test results have shown that we can provide skillful forecasts into the medium range.