Tracking and Verifying Heavy Precipitation Objects from NSSL’s Warn-on-Forecast Ensemble

Verification of Quantitative Precipitation Forecasts (QPF), especially at higher
thresholds, can be challenging due to the discrete nature of heavy precipitation. Standard
grid-based verification approaches do not necessarily quantify specific forecast errors,
such as displacement, intensity, orientation, and timing. Hence, it is important to explore
the utility of object-based tracking approaches with heavy precipitation objects. This talk
will focus on tracking QPF objects through space and time to quantify and display heavy
precipitation object attributes using the Warn-on-Forecast System (WoFS) at high
temporal and spatial scales.
The National Severe Storms Laboratory’s (NSSL) WoFS is a regional domain,
experimental, on-demand convective-scale ensemble analysis and prediction system. The
goal of WoFS is to provide National Weather Service (NWS) forecasters with a more
continuous flow of probabilistic model guidance in the 0 to 6 hour timeframe to enable
earlier and more accurate communication of hazardous weather threats. The WoFS heavy
precipitation objects are identified and tracked using the Model Evaluation Tools Version
8.1 (METv8.1) Method for Object-Based Diagnostic Evaluation Time-Domain (MTD).
This study will present results from tracking 15-minute QPF objects from the
WoFS both operationally (e.g. to create novel ensemble visuals for forecasters) and
retrospectively (e.g. to aggregate object-based bias statistics) for the 2019 warm season.
In addition, this talk will highlight some of the sensitivity results important to selecting
the optimal MTD parameter configuration for properly identifying and tracking heavy
precipitation objects. The importance of tracking 15-minute accumulated precipitation
compared to 1-hour accumulated precipitation during flooding events will be discussed.