Laserfiche WebLink
<br />) <br /> <br />Perhaps the most commonly employed of these techniques is the "target" and <br />"control" comparison. This technique is based on the selection of a variable that would <br />be affected by seeding (such as snow). Records of the variable to be tested are acquired <br />for an historical period of several years duration (20 or more if possible). These records <br />are divided into those that lie within the designated "target" area of the project and those <br />in a nearby "control" area. Ideally the control area should be selected in an area which <br />would be unaffected by the seeding. The historical data, e.g., precipitation!snowpack, <br />in both the target and control areas are taken from past years that have not been subject <br />to cloud seeding activities, since past seeding could affect the development of a <br />relationship between the two areas. Obviously, these data are also taken for the same <br />period of time (usually months) when the seeding will be conducted. These two sets of <br />data are analyzed mathematically to develop a regression equation which predicts the <br />amount of target area precipitation! snowpack, based on precipitation! snowpack in the <br />control area. This equation is then used during the seeded period to estimate what the <br />target area precipitation! snowpack should have been based on that observed in the control <br />area. A comparison can be made between the predicted target area <br />precipitation!snowpack and that which actually occurred. Any resulting difference can <br />be tested for its significance through statistical tests. This target and control technique <br />works well where a good correlation can be found between target and control area <br />precipitation. Generally, the closer the two areas are together the higher will be the <br />correlation. Areas selected too close together, however, can result in an underestimate <br />of the seeding effect. For precipitation!snowpack assessments, a correlation coefficient <br />(r) of 0.90 or better would be considered excellent, and would indicate that over 80 <br />percent of the variance (1") in the historical data set would be explained by the regression <br />equation used to predict the variable (expected precipitation! snowpack) in the seeded <br />years. <br /> <br />One of the hazards of this type of analysis is that a "dilution" of the seeding effect <br />is encountered due to the fact that seeding occurs on only some of the storms during any <br />particular month, and in some instances seeding does not occur on any of these storms. <br /> <br />5-2 <br />