Maximum destination between stations of control region and target region’s border should be 50 km to 100 km and should be located in areas which continentally are similar to the target region. Precipitation volume of January, February, April and December of 2009-2010 was measured through kriging method and related raster maps were made through normal kriging method and total volume of precipitation for target region was specified. Time series of monthly precipitation for each station for exact time periods was provided and was take in to regression model, as independent transitive.
Historical regression approach:
According to national and international evaluations, historical regression is one of the most reliable approaches in cloud seeding project’s evaluation. In this approach, best regression proportion between control and target region during years which no cloud seeding has been done, is determined so that having precipitation volume of control region, the precipitation of target region can be prospected. In this approach, proportion data between control and target region can be very variable and possibility of estimating the changes to be accidental is probable. One of preparations which should be done before evaluation is to normalize monthly precipitation data. If the actual data is not normalized, the presumption of mistakes in adherence of changes in regression line from normal line will be common. The other point which should be considered precisely is equality of time scale between control and target region. If the time scale in both regions is not equal, correlation and reliable estimations are doubted.
Determination of precipitation in target area:
First, raster maps for each month of the year (1973-2006) were generated using interpolation kriging method and then precipitation values for each year were extracted from these maps. In the next step, the precipitation volume of December, January, February and April for target stations were calculated and (using interpolation method) used as dependent variables for estimation of precipitation volume of target regions.
Determination of confidence levels:
In creditable sources, the confidence level of results of cloud seeding projects is determined to be 95%. So to determine that in confidence level of 95% , the increase of precipitation is caused by cloud seeding or any other factor has effect on this increase, some other confidence tests will be done on results.
Determination of confidence intervals:
The Determination of confidence intervals of natural changes in precipitation of region is achieved through statistic methods. For better understanding of effect of confidence intervals on confidence of results, it can be said that if precipitation volume in target region (seeding materials affected area) is higher than prospected volume, it means that this increase (with confidence of 95%) is caused by cloud seeding project and if precipitation volume is lower than minimum prospected volume, it means that (with confidence of 95%) the cloud seeding project has caused reduction of precipitation, also if the precipitation is within prospected volume, it means that the effect of cloud seeding project can’t be judged.