The rain enhancement experiment in Puglia, Italy: Statistical evaluation

A randomized rain enhancement experiment was carried out during 1988-94 in the area of Bari and Canosa, Italy, on the Adriatic coast. It was commissioned by the Italian Department of Agriculture and Forestry and the region of Puglia, with TECNAGRO, a nonprofit Italian company, as overall manager, and with EMS, an Israeli company, as field operator. The original purpose was to study rain-producing weather systems in southern Italy, establish similarities with Israel, and transfer Israeli technology. The experiment was a cross-over design with two alternating target areas, a buffer in between, and two additional control areas. Seeding was by injection of silver iodide into clouds by aircraft flying near the bases of clouds along predetermined tracks upwind of the target area. The experimental units were rainy days. Based on historical rain gauge data, it was estimated that 303 rainy days were required to establish a 15% rain increase at a significance level of 0.05 and 90% power. In 1995, TECNAGRO asked the Scientific Committee for a statistical evaluation to investigate if a seeding effect could be established before the original goal of 303 seeding days was reached. The results of the analysis of the 260 available rainy days were that no discernable seeding effect could be found. This was evident from the root double ratio (RDR) and root regression ratio (RRR), which yielded RDR - 1 = -0.083 +/- 0.089 and RRR - 1 = -0.004 +/- 0.057, respectively (the +/- sign represents the standard error of the estimate). Based on that result, it was decided to terminate the Puglia seeding experiment. Preliminary exploratory studies suggest that the two target areas might have been affected differently by seeding and that an apparent substantial seeding effect occurred in the Bari rea under conditions of moderate precipitable water between 700 and 850 mb. If these findings are confirmed by the recommended meteorological analyses and airflow studies, a new experiment with an appropriate design might be justified.