OCR Text |
Show aid. (10) Vertical profiles of radar AGC voltage are a valuable analysis (11) Good PPI coverage, with photographs at 10 minute intervals, could be profitably used to help explain natural variations in precipitation rate and the number of snow crystal replicas containing Agl particles at their centers. (12) There is usually positive correlation between silver-in-precipitation values and the number of snow crystal replicas containing Agl particles at their centers. 5. 4 Statistical Analysis of 1968-69 Precipitation Data Statistical testing of the 1968-69 season precipitation data, for determining the sign and magnitude of seeding effect, was accomplished by first stratifying the data into seeded and unseeded categories by time and location, and also by cloud top temperature. Then a non-parametric sum of squared ranks statistical test (Mielke, 1967) was made by individual station for each stratification. Stratification was performed according to the following steps: (A) Stratification Into Seeded and Unseeded Time Periods The time of arrival of seeding effect at each station was computed as already explained, with one modification to give timing to stations lying outside the computed area of effect. For stations lying outside the 4 a half-width seeding plume, timing was accomplished by the argument that the half-width was much greater than 4 a and thus included all stations in the data network. The three hour period following arrival of seeding effect was designated the seeded period for each six hour operational block. (B) . Stratification by Location of Computed Area of Seeding Effect After each station's data had been divided into seeded and unseeded time periods, it was further stratified by noting whether the station was located inside or outside the computed area of seeded precipitation. (The original 4 a plume half-width was used for determination of seeding effect location, of course.) (C) Stratification by Cloud Top Temperature Grant, et al (1968), have shown highly significant positive seeding effects from seeding of winter orographic clouds when 500 mb temperature (assumed -20 °C. Thus, Park Range data to represent cloud top temperature) is were further stratified into two categories according to whether the Mt. Harris radiosondes taken during the operational period indicated cloud top 294 1 1 |