Laserfiche WebLink
<br />Executive Summar~ <br /> <br />The San Juan Mountains, Colorado operational seeding program of 2005-2006 was scientifically <br />evaluated using a snowpack model and data assimilation system. This system is known as the <br />Snow Data Assimilation System (SNODAS). The change in snowpack snow water equivalent <br />(dSWE) over 24 hour periods, during and after seeding, was the primary variable for measuring <br />any snowfall from seeding. The dSWE from the San Juan target area was compared with that <br />from two nearby control areas, which were assumed to be unaffected by seeding. This <br />comparison was attempted to distinguish any seeding-induced snowfall augmentation from <br />natural (unseeded) snowfall. <br /> <br />The analysis did not indicate any systematic increase or decrease of snowfall in the target (vs. <br />control areas) that might be attributable to cloud seeding. This finding could be the result of: a) <br />No actual snowfall augmentation from seeding; b) lack of similar meteorological characteristics <br />between target and control areas (resulting in an "apples-to-oranges" comparison); c) inability of <br />SNODAS to distinguish a small seeding-induced snow increase within the overall natural <br />snowfall; d) SNODAS model errors. Item c) has been an ongoing major problem in the <br />evaluation of operational seeding programs. SNODAS dSWE maps for each seeding event <br />showed that larger-scale storm characteristics were the dominant controls on spatial distribution <br />of snowfall, a factor that might have masked any seeding effect. <br /> <br />The SNODAS currently outputs a continuous grid of I km snowpack data across the United <br />States on at least a daily basis. Since seeding is aimed at a target area and not just a point, the <br />use of a variable that is averaged over target and control areas (as with SNODAS) should be <br />superior to a variable measured at a few points (as is traditionally done in operational seeding <br />evaluations with SNOTEL data). Although it is unknown whether SNODAS could diagnose an <br />actual seeding effect in future applications, the system may prove useful in providing snowpack <br />information relative to the meteorological conditions that influence seeding. Given the results <br />and uncertainties of this study, however, several other evaluation methods are recommended that <br />might better diagnose the presence of a seeding effect, if it exists. These other evaluation <br />methods are statistical or physical. <br /> <br />2 <br />