Initial soil moisture as a predictor of subsequent severe summer weather in the cropped grassland of the Canadian Prairie provinces

Abstract

AbstractSoil moisture, along with the type and stage of the vegetation, influences the thermodynamic structure of the atmosphere by regulating heat and moisture fluxes in the planetary boundary layer (PBL). This study examined whether the modelled aerial‐average root‐zone soil moisture (RzSm) in ‘wet’ and ‘dry’ areas of the cropped grassland of the Canadian Prairie provinces had predictive value in determining if these areas would subsequently have above‐ or below‐normal occurrences and event days of severe summer convective weather (i.e. tornadoes, hail, heavy rains, and/or strong winds). RzSm, simulated by the Prairie Agro‐climate Model, for the 1997–2003 growing seasons was analyzed three times per season. Dry areas with RzSm ⩽50% of available water holding capacity (AWHC) and wet areas with RzSm > 50% of AWHC were delineated post‐snowmelt, on 15th June, and on 15th July. The aerial‐average RzSm levels in the dry and in the wet areas were calculated, and plotted against the relative number of occurrences and number of event days that were recorded during the remainder of the growing season for the various types of severe summer convective weather. In each case, the best‐fit linear regression line and the variance that it explained (r2 value) were computed. The hypothesis that the slope of each regression line was significantly different from zero was then tested. A value of r2 close to or greater than 0.25 was arbitrarily used as a cut‐off point—a relationship with an r2 close to or greater than this value, and with a regression line slope that was significantly different from zero, was selected as one which could have potential value in the climatological forecasting of severe summer convective weather. For most of the severe weather types, the relative number of occurrences and the relative number of event days, which were recorded subsequent to the three dates on which the aerial‐average RzSm was determined were greater in the wet areas than in the dry areas. Copyright © 2008 Royal Meteorological Society