WMOD00447 - Laserfiche WebLink

Home Browse Search

. . . is + 4 dBZ. This maximizes SAA performance during important storms but also increases "virga contamination." However, the forecasters have other information with which to judge what is virga and what is light snowfall at mid- and far-ranges. They have decided that dealing with false indications of snowfall when virga is present is mom acceptable than having degraded SAA performance, and especially lack of range correction, during actual snowfall events. The NWS Forecast Offices located near the other four IN EXRAD radars also appear to be making use of the SAA products, and some of these offices have provided valuable feedback. Daily precipitation and snow depth observations, and rawinsonde data, are being saved for each significant stonn period affecting some or all of the 5 radars in Minnesota and the Dakotas. NIDS volume scan data are being saved for all days of the current winter. These data will allow post- winter comparisons of surface observations and SAA estimates on a daily basis. Such comparisons should be made with care because ground observations often differ from unknown "ground truth" because of wind related errors in precipitation gage catch and scouring and drifting of newly fallen snow. It is best to compare the general agreement among many observing points, remembering that gages almost always underestimate snowfall, even with light-to-moderate wind speeds. It is expected that such post-season analysis will improve SAA performance in future winters. Turning to Goal 7 above, the difficulties of comparing raldar estimates of SWE with those provided from aircraft gamma radiation sampling by NOHRSC have been discussed in a recently provided brief annual report on a closely associated GCIP project, Proposal GC97-377 (Super 1999). For reasons discussed there, it is not planned to pursue this approach any further. An exception might be made if a special large storm case becomes available with a week or so between sampling of the same gamma flight lines and negligible melting between the sampling dates. But typically 3 or more weeks elapse between gamma radiation sampling over the same flight lines. Also in reference to Goal 7, work in progress with the large Minnesota volunteer precipitation gage network was noted by Super (1999). Two major storms from November and December 1996, respectively, now have complete gage storm total records. Considerable editing was required to develop these data sets in spite of the fact that they were chosen in part because of the absence of much precipitation during the days immediately before or after them. Daily precipitation observations from the standard Cooperative Station neltwork have also been developed into data sets for these two storms. These data will be used to gain insight into the Ze-S relation previously developed with hourly observations from a limited number of recording gages located in fairly protected locations (Super and Holroyd 1998). It is hoped that these data will provide an alternative approach to examination of the range correction problem. Results will be presented after this study is completed. Further information on this work is given by Super (1999). It is planned to continue to improve SAA performance during and after the current winter by comparing SAA estimates with surface observations of precipitation and new snow depth, and by implementing input from the NWS Forecast Offices. It is planned to continue to provide merged HRAP SAA SWE accumulations to NOHRSC next wintE~r, and it is expected that additional radars will be added to the overall area of coverage. With GCIP field emphasis shifting to the LSA-NW (the Upper Missouri Basin), it appears reasonable to add radars in the western Dakotas and Montana, where Reclamation has considerable interest in water resources. Possible NEXRAD 5