Laserfiche WebLink
<br />002733 <br /> <br />augmentation ratio, and (2) randomizing storms for seeding or not. The San <br /> <br /> <br />Juan seeding is cut off whenever precipitation reaches 125% of normal, in <br /> <br /> <br />order to minimize the risk of disbenefit. Kahan: Persons in the path of an <br /> <br /> <br />avalanche or flood are more concerned about caution than a statistician might <br /> <br /> <br />be; having a set cut-off point makes weather modification more socially <br /> <br /> <br />acceptable than a philosophy of "seed and be damned." Tyner: The second <br /> <br /> <br />restriction comes from 80/20 randomizing (only 80% of the storms are seeded). <br /> <br /> <br />When a storm comes, Hjermstad opens an envelope from CSU that says either <br /> <br />"seed" or "don't seed". Kahan: One can not obtain a permit in Colorado <br /> <br /> <br />unless your operation promises an increase in scientific knowledge, and that <br /> <br /> <br />is the reason that all projects in the state use randomization. A 50/50 <br /> <br /> <br />randomization would allow more efficient experimentation, so 80/20 represents <br /> <br /> <br />a compromise between ideal experimentation and all-out operation. However, <br /> <br /> <br />there are emergency conditions in the Colorado law. As long as you are trying <br /> <br />to learn something, people get less upset by disbenefits. <br /> <br /> <br />Summers: Do you use a rain gage network to evaluate the results of <br /> <br /> <br />seeding? Kahan and Hurley: The Bureau's money is being used to run six <br /> <br /> <br />recording precipitation gages, monitoring data for use in a cloud model on the <br /> <br />Bureau's computer and for analysis of Pilot Project precipitation/runoff <br /> <br /> <br />relationships. Summers: In the 80/20 randomization, is there a method to <br /> <br /> <br />determine the difference of precipitation in your seed and no-seed storms? <br /> <br /> <br />~: The 80/20 randomization prolongs the time to reach any valid decision <br /> <br /> <br />over what the time would be for 50/50 randomization. The evaluation is done <br /> <br /> <br />after the fact by stratifying the randomized data according to some meteor- <br /> <br /> <br />ological parameter, i.e. cloud top temperature. Making legitimate comparisons <br /> <br /> <br />between seeded and non-seeded storms requires specifying beforehand how the <br /> <br /> <br />data will b~ stratified, rather than waiting until the seeding effect shows up. <br /> <br /> <br />Keyes: Is there any 'more information concerning the Colorado Pro- <br /> <br /> <br />gram? Tyner: Our intent in starting the program this year was to set the <br /> <br />stage for the Colorado River Weather Modification Demonstration Project of the <br /> <br /> <br />Bureau of Reclamation (also see November 19, 1976 letter in Attachments D). <br /> <br /> <br />Morrill: The Legislature also provided $2,200 to compliment the project on <br /> <br /> <br />Grand Mesa, northeast of Grand Junction. This money extended the Grand Mesa <br /> <br /> <br />Water Users Project from March to May 15th. The base pay for that project has <br /> <br /> <br />been $5,000, without seeding. The project operates only if the snowpack is low. <br /> <br />34 <br /> <br />