An Examination of Frozen Precipitation Impacts on MRMS Q3 during Winter Precipitation Events

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 9:00 AM
127ABC (Phoenix Convention Center - West and North Buildings)
Steven M. Martinaitis, CIMMS/Univ. of Oklahoma and NOAA/NSSL, Norman, OK; and S. B. Cocks, Y. Qi, B. T. Kaney, J. Zhang, and K. Howard
Manuscript (1.5 MB)

An evaluation of the Multiple Radar-Multiple Sensor (MRMS) Quantitative Precipitation Estimate (QPE) system (denoted as “Q3”) performance was conducted across the entire MRMS domain during the 2013-2014 cool season. Initial comparisons between hourly gauge observations and collocated Q3 radar-only QPE (Q3Rad) values quickly indicated that a large quantity of the gauges, numbering in the hundreds per hour, had likely become “stuck” during solid, winter precipitation events (i.e., G = 0, R > 0). This study assessed the spatial coverage and quantity of gauges impacted during winter precipitation events and the resulting impact on any multi-sensor precipitation estimates using gauge data as a local bias correction application. Comparison classifications between gauge observations and Q3Rad values were also assessed based on hourly surface wet-bulb temperatures (Twb) from the Rapid Refresh (RAP) model and Radar Quality Index (RQI) values. Periods of significant winter weather impacts on gauge observations recorded over 100 hours of at least 1000 gauge sites were classified as G = 0, R > 0 when RAP surface Twb was at or below freezing. Gauge sites that reported precipitation for at least six hours prior to becoming stuck showed a degradation of quality observations in the final two hours, with a decrease in the gauge-to-Q3Rad mean bias ratio from 1.0 to 0.6 and a reduced correlation coefficient to less than 0.2. A secondary impact from post-event thawing was also prominent in the hourly analysis. While gauge thawing during non-precipitation periods (i.e., G > 0, R = 0) were easily identifiable, thawing that occurred during both liquid and solid precipitation introduced errors in QPE and local bias corrections. The MRMS gauge quality control (QC) algorithm was modified to mitigate the impacts of stuck gauges in regimes where surface Twb was at or below freezing and post event thawing. Adjustments to the gauge QC algorithm successfully removed many of the affected gauges, thus improving the accuracy of MRMS Q3 during winter weather events.