Improving anticipation of the influence of upstream convection on QPF

Operational forecasting on the US East Coast can be complicated when conditions appear favorable for cyclogenesis to occur in the Southern United States or along the Southeast coast since many large population centers reside downstream in the Mid-Atlantic and Northeast United States along favored storm tracks. Despite tremendous advances with numerical weather prediction (NWP) and its ability to forewarn of significant weather impacts on the East Coast, in situations where organized convection associated with a midlatitude cyclone develops upstream (a situation dubbed “upstream convection”, or UC), significant model errors have been observed. In UC situations, previous studies have documented that model quantitative precipitation forecasts (QPF) can be highly unreliable.

Forecasters in the Mid-Atlantic and Southeast US have long noticed cases where convection initiates and rapidly translate across the Gulf Coast states, effectively cutting off moisture transport to parts of the Mid-Atlantic and Southeast, and resulting in a positive QPF bias in NWP forecasts. Other instances have been documented in which convection initiates, and moves slowly across the Gulf Coast states and Southeast, producing significant precipitation over the Mid-Atlantic and Southeast. These cases often feature an underestimation of precipitation amounts by NWP models.

The goal of this study is to identify the key features that distinguish the fast convection (FC) and slow convection (SC) scenarios. A database of cool season (1 Oct – 31 May) UC cases from 2001 to present has been developed, with composites generated for characteristic fast- and slow-moving sets of UC cases. This represents an attempt to identify which parameters contain crucial signals for operational forecasters to utilize in the forecasting of downstream QPF.

Parameters examined include the nature of the 500-mb trough axis (digging or lifting), the number of MCSs present, UC description (bowing linear, linear, etc.), the surface front nature (katafront or anafront), the 12-hour absolute speed of UC, and the 12-hour system-relative (cold frontal) speed judged by examining NARR data. Archived composite NEXRAD data is examined from UCAR over the web to determine the absolute UC speed, while the system-relative speed is computed from NARR data. It is hypothesized that these parameters may assist in the discovery of key discriminants between FC and SC cases.