WMOD00464 - Laserfiche WebLink

Home Browse Search

B-7 experiment. A criticism of the earlier Israel study (Isareli I 1960-1967) that has received some atten- tion in the literaturl~ pertains to the changes in daily schedule for seeding and recording ov~r the course of the study. Despite the fact that a subset of results for a period for which the seeding and recording times were fixed showed signific:ant seeding effects, and despite the fact that the shifting of times for seeding and recording could hav~ no systematic influence on results if seeding had been ineffective, the fact that procedures were changed raised some doubts, probably unnecessarily, about the value of the study. This nicely illustrates the sub- stantial importance of maintaining a fixed experimental design for confirmatory research, in order to assure a properly skeptical audience that results could not have been influenced by pro- cedural changes. * Alberta hail * The operational changes made in the Alberta hail project during its first season re:sulted in the loss of a randomized design in favor of operational seeding. * Santa Barbara II * Phase 1 had depended upon ground based seeding while Phase 2 was designe:d for air- borne seeding. Phase 1 used a fixed control area, while Phase 2 used no control area. (This change was reflected in the use of double ratios in analyzing Phase 1 and single ratios in Phase 2,) Apparently no other critical changes were made during the course of Santa Barbara II. * South Africa hail * There is no evidence for substantive changes, other than the acquisition of more capable aircraft, in the South African hail suppression program since its inception in 1972. * North Dakota * The operational design seems to have been maintained without substantive change throughout the term of the North Dakota pilot project, 1969 to 1972. Radiosondes and cloud modelling wen: introduced in 1971, allowing more sophisticated definitions of seedability. * Tasmania * The report on this project (Smith, Veitch, Shaw and Miller, 1977) is to be commended for its careful documentation of the days on which operations were suspended. We detect no sign of any serious issue here. ~ 10. Remote effect issues. In this section we summarize briefly issues involving the possible occurrence of remote effects which could influence reported results. * FACE * No control areas were used, thus remote effects of any importance would have had to arise as effects in the target area from other seeding operations. * ,NHRE * No control areas, remote effects could only have arisen from other seeding operations. * Colorado River Basin * No control areas, remote effects could only arise from other seeding operations. ~