Extended Abstract (332K)11.5
Statistical assessment of explicit model forecasts of convection using a new object-based approach
Cody L. Phillips, NCAR, Boulder, CO; and J. O. Pinto, M. Steiner, R. M. Rasmussen, and N. Oien
Short term forecasts for aviation for strategic planning (leadtimes of 3-8 hrs) require information describing the approximate locations of storm systems as well as the most useful storm characteristics such as coverage, orientation, and permeability (or presence of gaps). In this study we use a modified version of the Thunderstorm Identification Tracking Analysis and Nowcasting (TITAN) to evaluate the performance of high-resolution model forecasts of storm characteristics. TITAN is used to identify storms and catalog their characteristics (sizes, shapes, orientation, motion, etc). It has recently been modified so that it can also be used to determine the properties of storm areas and relationships between the storms such as the spacing between storm cells. A set of 4-km WRF ARW simulations (run twice daily) available from summer 2006 are evaluated to determine the model skill in predicting storm attributes in the Southeastern US. Initial results indicate that the model produces too many small storms regardless of the forecast leadtime. A set of 1-km WRF simulations will be performed for this region (using the 4-km WRF runs as input forcing at the lateral boundaries) to assess the resolution dependence of this bias. Additional modeled storm characteristics will also be assessed and related to large-scale model parameters such as environmental shear and stability.
Session 11, Verification
Thursday, 24 January 2008, 11:00 AM-12:15 PM, 226-227
Previous paper
Recorded presentation