Monday, 12 January 2004: 9:00 AM
Multiscale evolution and predictability of a warm season climate anomaly in the U.S. southern Great Plains
Room 6C
It is usually assumed that the resolution of extended range global model forecasts may be truncated beyond the first week because of predictability limits on smaller scale transient waves. Although unpredictable, we argue that such features should be retained to help improve forecasts of sudden shifts in the large scale flow regime that may lead to regional changes in seasonal climate. To support this argument, we study the onset and maintenance of a seasonal climate anomaly that developed over the U.S. Southern Great Plains during summer 2002. The onset of cooler and wetter than average weather was caused by flooding rains in the vicinity of San Antonio, TX on 1-2 July 2002. The origins of this flooding event can be traced to a mesoscale convective vortex that formed over Kansas and Oklahoma on 15 June 2002. We report on the multiscale evolution of convective weather towards the shift in regional climate, emphasizing the predictability of such a regime change. NCEP/NCAR reanalyses are used to show that the climate anomaly is detectable on weekly and monthly timescales. The analyses highlight the multiscale evolution of conditions leading to the onset and persistence of the regional anomaly. Numerical model studies are used to identify the sensitivity of the anomaly to soil moisture feedback processes. We also discuss the impact of model resolution on the predictability of this event by studying variance spectra from operational NCEP 15-day global model forecasts.
Supplementary URL: http://cimms.ou.edu/~pnutter/15thGlbChng.pdf