Laserfiche WebLink
<br />002491 <br /> <br />Although the primary objective of treating the southerly storms <br />would be redistribution and for the westerly storms increased <br />efficiency, all three hypotheses would be tested for each of the <br />two storm types. <br /> <br />Statistical Design <br /> <br />The statistical design recommended by MBA is a 50-50 randomized <br />procedure in which approximately one-half of the events identified <br />as suitable for seeding are seeded and the remaining one-half are <br />left untreated. This procedure provides an opportunity for <br />comparison of seeded and un seeded events selected at random, <br />possibly through a randomized computer program to avoid human <br />bias. <br /> <br />MBA has recommended that a 24-hour period, or "day," be chosen as <br />the experimental unit, with a 24-hour purge period whenever a <br />seeded event is not followed by another seeded event. The purge <br />period would provide time for the seeding material to leave the area, <br />thus avoiding the possibility of a "seed day" contaminating a <br />foll owi ng "no-seed day." When a second experimental peri od occurs <br />directly after a seeded event, the day would become a purge day after <br />which a new randomized draw would be made. <br /> <br />In the recommended procedure, days would be identified in the <br />forecast as to whether they are characterized by westerly or <br />southerly storms and the randomizations for each done separately. <br />The recommendation for selection of the 24-hour period as the <br />experimental unit is under consideration by Project Skywater <br />scientists. A decision on this point awaits analysis of additional <br />data. <br /> <br />MBA said the primary measure of effect would be the water content <br />of the precipitation; however, storm intensity, storm duration, <br />and seasonal streamflow would also be monitored. Due to the extremes <br />of natural variability, the SCPP is expected to require 5 to 7 years <br />to statistically verify an effect with the proposed 24-hour randomized <br />periods. <br /> <br />Operational Forecasting <br /> <br />The operational forecasting program suggested by the initial MBA <br />design would predict the onset of a seedable situation and deter- <br />mine the mode of seeding for that situation. Since a project <br />purpose is to augment the Sierra Nevada snowpack, the first objective <br />could be refined to define a seedable situation as one capable of <br />increasing snow water content either by snow or rainfall. Since <br />the snowpack can absorb a certain amount of liquid water, deter- <br />mining a seedable situation depends not only on meteorological <br />conditions, but also on the condition of the snowpack. <br /> <br />16 <br />