Tropospheric Static Stability and Central North American Growing Season Rainfall

Abstract

This study investigates the relation between tropospheric static stability and central North American growing season (May?August) rainfall for the highly contrasting years of 1975. 1976, and 1979. It uses two extensive sets of meteorological data (individual rawinsonde soundings for 38 stations; hourly rainfall totals for 854?944 locations) for the region extending from the Rocky to the Appalachian Mountains and from the Gulf Coast to approximately 55°N in Canada. The major objectives are to: (i) ascertain which of the many available methods of parameterizing static stability are most strongly related to the above (predominantly convective) rainfall; and (ii) quantify the rainfall variance fraction explained by static stability alone, as opposed to other atmospheric processes/conditions. Forty static stability indices and related thermodynamic parameters (SSITPs) are treated. The results pertaining to objective (i) are definitive and those concerning (ii) are encouraging. The SSITPs that correlate most strongly with rainfall amount consistently include the lifting condensation level (LCL) (near-regionwide) and the convective condensation level (CCL) (western U.S. Great Plains) for the afternoon half-day, and K-type and SWEAT indices (eastern United States) and the CCL and convective temperature (U.S. Great Plains) for the morning half-day. In contrast, the SSITPs developed for forecasting severe thunderstorms and tornadoes correlate poorly with rainfall amount. Except on the U.S. Great Plains, the maximum SSITP-rainfall amount correlation magnitudes tend to be larger for the afternoon half-day (average of 0.47?0.49) than the morning half-day (0.37?0.39). Particularly high maximum afternoon SSITP-rainfall amount correlation magnitudes were obtained for the eastern United States (0.50?0.70); earlier work of this type seldom yielded correlation magnitudes above 0.32. For the SSITPs that correlate most strongly with rainfall amount on a regionwide basis (ICL variant for afternoon; modified-K index for morning), we also document the considerable spatial and intraseasonal variability of the thresholds beyond which the probability of rainfall exceeds that of no rainfall.