Abstract
Previous research has shown that a lower-tropospheric diabatically generated potential vorticity (PV) maximum associated with an area of incipient precipitation (IP) was critical to the moisture transport north of the PV maximum into the Carolinas and Virginia during the 24–25 January 2000 East Coast cyclone. This feature was almost entirely absent in short-term (e.g., 6–12 h) forecasts from the 0000 UTC 24 January 2000 operational runs of the National Centers for Environmental Prediction (NCEP) North American Mesoscale (NAM, formerly Eta) and Global Forecast System (GFS, formerly AVN) models, even though it occurred over land within and downstream of a region of relatively high data density. Observations and model analyses are used to document the forcing for ascent, moisture, and instability (elevated gravitational and/or symmetric) associated with the IP, and the evolution of the IP formation is documented with radar and satellite imagery with the goal of understanding the fundamental nature of this precipitation feature and the models’ inability to predict it. Results show that the IP formed along a zone of lower-tropospheric frontogenesis in a region of strong synoptic-scale forcing for ascent downstream of an approaching upper trough and jet streak. The atmosphere above the frontal inversion was characterized by a mixture of gravitational conditional instability and conditional symmetric instability over a deep layer, and this instability was likely released when air parcels reached saturation as they ascended the frontal surface. The presence of elevated convection is suggested by numerous surface reports of thunder and the cellular nature of radar echoes in the region. Short-term forecasts from the Eta and AVN models failed to capture the magnitude of the frontogenesis, upper forcing, or elevated instability in the region of IP formation. These findings suggest that errors in the initial condition analyses, particularly in the water vapor field, in conjunction with the inability of model physics schemes to generate the precipitation feature, likely played a role in the operational forecast errors related to inland quantitative precipitation forecasts (QPFs) later in the event. A subsequent study will serve to clarify the role of initial conditions and model physics in the representation of the IP by NWP models.
Short-Term, Single, Multiple-Purpose Reservoir Operation: Importance of Loss Functions and Forecast Errors
