Observations and science questions pertaining to cloud seeding over Tasmania

An analysis of intermittent cloud seeding activity for the period 1960-2005 over a hydroelectric catchment (target) area located in central Tasmania is presented. The analysis is undertaken on monthly and seasonal area averaged rainfall and is one of the first studies of economically viable rainfall enhancement over a large land area, approx. 10,000 square km. The current long-term investigation is consistent with earlier field experiments by Smith et al. (1979) and Shaw et al. (1984).

Results indicate that periods of cloud seeding activity are associated with increased rainfall over multiple regions within Tasmania. The probability of such increases being attributable to random variation is assessed using a bootstrap randomisation technique and found to be around the 0.10 level for spring. Further analysis using double ratios indicates that the increase in precipitation in the target (relative to the highest correlating control) is around 6% during winter and spring, with four out of the six months achieving probabilities of occurrence of less than or equal to 0.06. There is evidence to suggest that the western most control used in the analysis may have been inadvertently seeded, this result indicates that there could be an issue regarding the timing of release and targeting of the seeding material.

Further analysis is undertaken to examine some of the targeting issues highlighted by the long-term analysis. Typically, the seeding agent is released by aircraft near cloud top in the post-frontal airmass associated with cold fronts passing over the western coastal region of Tasmania. This situation is unique in that cloud seeding in Tasmania is not typical winter orographic seeding, the agent is released often upwind of the relatively low-lying mountains that are approx. 1.5km in height and the seeding elevation is typically around 2.5-3.5km. Aircraft soundings made off the western coast of Tasmania regularly show regions of shear in the free troposphere. This feature is absent in soundings from Hobart Airport and NCEP reanalyses. This shear layer is not considered by the cloud seeding team when deciding what area to target, it therefore could be a significant source of targeting errors as its presence could easily displace hydrometeor trajectories so that regions other than the intended target area receive increased rainfall. An analysis is presented regarding the sensitivity of seeding track placement to precipitation fall out location with and without the observed shear layer.