O the Subsynoptic Pre-Tornado Surface Environment.

The pre-tornado surface environment over a limited domain was studied during the six hours prior to reported tornado development. This was accomplished by applying composite analysis techniques to a surface tornado proximity collection. The collection consisted of 23 tornado case studies in which the tornado was first reported to have touched down within 10 nautical miles of an hourly reporting surface station. To obtain consistency in the averaging process from case to case the irregularly spaced surface data was objectively interpolated to a regularly spaced 560 x 560 nautical mile grid network with 20 nautical mile grid spacing. The grid was centered on the "tornado proximity" station. The positioning of the grid on the proximity station insured spatial and temporal continuity from case to case. The interpolated data were then averaged for each of the six hours prior to the reported tornado development. This provided a composite view of subsynoptic variability of the surface patterns in the immediate pre-tornado environment. In addition, a band-pass filtering scheme was applied to the interpolated data. The analysis was performed to that weather patterns whose spatial dimensions of about 240 nautical miles were enhanced while larger and smaller features were greatly smoothed. This procedure enhanced those subsynoptic weather patterns hidden or masked in the larger synoptic systems. All of the basic parameters (temperature, moisture, pressure, and wind) were analyzed as well as many derivative fields (gradients, advection, etc.). Many of the well known patterns associated with severe weather were evident in the composite results. On the other hand several new parameters were shown to have forecast potential. The v-component of the wind, the 3D index, and the warm-moist parameter all tended to maximize near the proximity station. In addition, band-pass pressure and temperature results showed significant features which were hidden in the regular composite analysis. Finally, the usefulness of band-pass analysis in a forecast setting was stressed.