WMOD00472 - Laserfiche WebLink

Home Browse Search

When the model was run on CVR soundings and INW 750-mb wind speed was included as a second independent variable in MLR, the multiple corre~ation was 0.72, approaching the 0.78 obtained with local sounding data. The model physics probably could be improved (in some future effort) to at least attain the level yielded by the more statistical approach. Performance of a followup task to study cloud PE would involve application of the forecasting equations to the 10-winter sounding data set. It would be prudent to also employ the model approach and compare results with those obtained from the equation rendered by the statistical method. A decision is required on whether an R2 of about 0.45 (statistical approach) is ad.equate to justify the performance of task 2 involving the estimation of the PE. The determination depends somewhat on the value to Arizona ofthe knowledge ofPE. Clearly, additional water is of substantial value to Arizona. Thus far, studies indicate cloud seeding to be promising for increasing winter precipitation over the Mogollon Rim. Consequently, a numerical procedure that can contribute to the quantification of cloud seeding possibilities should be of value and interest. The procedure can then be used in future assessments of the benefits of cloud seeding. About half of the variance is left to noise from the procedures described in this study. However, the analysis proposed for a followup task 2 is largely of a climate nature. Other forecasting systems, such as the seasonal forecasting of hurricanes, have also encountered substantial noise. In these types of problems, explaining about one-half of the variance offers beneficial results because of the high value of the product to society. Performing task 2 The primary purpose of performing task 2 is the development of relationships, applicable to a historical period, that estimate PE in winter storms on the Mogollon Rim. For Arizona, the definition ofPE must be loosely applied because the ice water that crosses the barrier cannot be estimated in historical data. Yet, ice water does influence PE and, in some cases, will likely substantially increase estimate errors. However, the error from exclusion of transient ice water is not likely to render pointless the performing of task 2. The rationale for this impression and the general approach to performing the task follows. The PE can be estimated by an expression such as the following (excluding water vapor in the water budget because saturation is assumed within clouds and the temperature changes are not large in the immediate vicinity of a ridge crest): P PE = I (1) PI+IT+LT The PI represents ice precipitation; that is, all precipitation captured by the gauge is assumed to be snow, though if some consisted of raindrops one should assume inclusion in the PI. The IT represents ice in transit that does not precipitate at least in the local study area. Similarly, the LT stands for cloud liquid water in transit for points downwind beyond the geographical area of interest. For Arizona, the expression must be modified to the following: P PE= I PI + LT (2) 17