PATTERN ANALYSIS OF ONE SUMMER'S MULTILEVEL MAPS OF MONTREAL RAIN

Abstract

For the summer of 1964, precipitation intensity in the vicinity of Montreal, observed by CPS-9 radar, was recorded on constant-altitude maps at six heights. To allow for attenuation (at ??=?3.2 cm) and other radar uncertainties, the distribution with rainfall rate R at 5,000 ft was matched to that for one rain gage at the surface. The extent (area ? time) in excess of a given R decreased rapidly with increasing R and with increasing height: a factor 1000 from 0.4 to 100 mm hr?1, a factor 10 per 15,000 ft. The pattern in plan at any height (with resolution limited to about 2 mi in the recorded maps) can be described in terms of cells with intensity decreasing outward approximately exponentially. A logarithmic function of intensity was used: T?=?1.66 log (R/R0) where R0?=?0.25 mm hr?1. The value of T for a contour bounding a central area A is given by T?=?Tc?x, where Tc is the maximum and x?=?(A/17.5 n.mi.2)1/3. (With this relation, one can estimate peak intensity from the area bounded by any given intensity.) The number of cells having Tc greater than a given value of T, per 108 n.mi.2 of map area (as for hourly maps of area 20,000 n.mi.2 throughout a 5,000-hr summer) is N0, where log N0?=?8.8?T?H and H?=?height/15,000 ft. As to shape, the maximum and minimum extents were always within a factor 6 of each other. Usually, the area in the horizontal section of a cell decreased with height. Occasionally it increased to a maximum, at about 20,000 ft, as much as four times greater than the area at 5,000 ft. These maxima aloft were generally identifiable with storms reaching 40,000 ft, which is tall for Montreal.