WMOD00309 - Laserfiche WebLink

Home Browse Search

point and elevations were interpolated to these points. The resulting terrain was run through a two-dimensional nine- point smoother twice to insure stability within the model. An overly rough terrain can generate noise in the model when the pressure solver is applied. Three-dimensional depictions of the three domains and their relative positioning are shown in Fig. 2. The vertical scales of Fig. 2 are exaggerated. To get a better perspective of the terrain, the reader is referred to Fig. 3, which shows the terrain of the innermost domain with the vertical scale the same as the horizontal scale. The innermost domain shows good detail of the terrain, however, the canyons on the west slope of the Plateau, more specifically, Birch Creek and Fairview Canyons, are not well-defmed. The ramifications of this are discussed later. b. The Initializing Sounding The model as run for this application was initialized from a single sounding. The sounding was applied homogeneously over all domains and the outermost boundary had the equation of continuity relaxed to provide zero net mass flux into the model for the variable terrain. The best sounding would be well upwind of the innermost domain The nearest upwind sounding site was at Ely, NY, approximately 290 km to the west of the target area. Because Ely is a routine National Weather Service site, its times did not necessarily coincide with the experimental times. Initially a single sounding was input to the model, however, a majority of the soundings had one or more nuances, e.g., a shallow inversion layer or a superadiabatic zone, to which the model was sensitive. For the two cases presented herein, composite soundings were derived by overlaying each of the project and aircraft-derived soundings. In so-doing the irregularities could be discerned and information taken over the plateau during each experiment could be included. The sounding input to the model was hand-drawn and represented a somewhat smoothed profile for the time of the case study. For the 6 March case, the Ely sounding was included, but due to the LORAN failing, no aircraft-derived winds were available. To insure computational stability, the highest available level from the composite sounding was repeated to 5 mb. c. Sulfur Hexaflouride Data The aircraft SF6 data applied to this report were the post-season processed data provided by NA WC, whose technicians were responsible for both the airborne and van detectors. A Scientech LBF- 3 instrument was used. This corresponds to a sensitivity of approximately 1 parts per trillion (ppt) by volume. False signals, such as those induced by sudden changes in cabin pressure, were found by visual inspection of the data and flight notes. The process was complicated by the baseline's drifting and noise. Data points which were known to have zero SF6 signal were identified from plots of the SF6 channel and flight notes. Lines were fit between pairs of these zero points. Voltage differences from the fitted lines were converted to ppt by volume using pre- and post-flight experimental calibrations done by injecting known concentrations into the SF6 sampler. The threshold of detection for this device is 10 ppt, however in some instances concentrations as low as 1 ppt were discerned. The time for the SF6 analyzer to process a given sample was assumed to be negligible. Length and flow measurements of the sample train from -6- I I I I I I I I I I I I I I I I I I I