The intent of the Oklahoma Weather Modification Demonstration Program was to be an evaluation the potential effects and benefits of an operational program rather than an experiment attempting to "prove the concept" of weather modification. Since the operational aspect of this program did not allow for a statistical determination of any impact on the amount of precipitation, this report provides some general analysis and case studies. Although this is an excerpt of the 1997 evaluation, the results of the 1996 and 1998 evaluations were similar.

Key findings can be summarized as follows:

• Qualitative analysis of specific rainfall enhancement seeding operations suggests that they are consistent with the seeding hypothesis. Specifically, in most cases, precipitation and increased cloud development are present after and downwind of the rainfall enhancement seeding activities. These results agree with the efforts of existing research projects which have demonstrated a seeding signature based upon long-term studies. However, the available evidence from the 1997 project is insufficient to suggest that this represents a seeding signature.

• Significant emphasis is placed on the analysis of a period in mid-May during which some seeding was randomized in conjunction with the collection of cloud physics data during a series of research flights. An analysis of the liquid water content (LWC) and droplet concentration show a decrease in the LWC after seeding, which supports the static seeding hypothesis. These results are consistent with research efforts of other states such as North Dakota, Kansas, and Texas, which have shown increased precipitation associated with the cloud changes described above. Since the cloud microphysics are consistent with the seeding hypothesis, and also agree with previous efforts, the seeding may have played a role in enhancing the rainfall duration or intensity. Analysis of the WSR-88D radar for one case study shows that it began to rain prior to and upwind of the seeding; thus, it is certain that the seeding did not cause the rain in that particular event. It is possible that the noted decrease in LWC and change in droplet and ice concentration are associated with naturally-occurring phenomena.

• Hail suppression results are inconclusive as there is no evidence that those flights resulted in a decrease in the intensity or size of specific hail events. An analysis of the hail reports for 1997 does yield some suggestion that the operator is capable of identifying those systems most likely to produce significant hail damage, and the methods used by the operator are consistent with efforts in other states, which have shown a reduction in hail damage. Finally, fewer than half of the significant hail events were in areas where the clouds were seeded, suggesting that perhaps more attention needs to be placed on hail suppression in future years.

As part of the 1997 OKWMDP a cloud physics research plane made several flights in May. The research plane made initial passes through suitable seeding candidates which were then seeded or not seeded on a random basis. On May 15 the research plane made several passes through a cloud located over southwestern Canadian county. Looking at the first page of the flight log for this flight you can see that the first pass was begun at 1648Z (11:48 AM CDT), the cloud was seeded at 1705Z (12:05 PM CDT) and the last pass was begun at 1717Z (12:17 PM CDT). The supercooled liquid water (SLW) content for this cloud decreases after seeding, and the SLW distributions show a similar trend. These facts are consistent with the static phase seeding hypothesis but they are not conclusive because the SLW content in clouds will naturally decrease as the cloud matures.

Archived products from the Frederick (KFDR) and Twin Lakes (KTLX) WSR-88D radar sites show the reflectivity (BREF2) and storm total precipitation (PRETX) before the reasearch plane's first pass, during the first pass, during the last pass, and after the last pass:

The reflectivity images show that the cloud did grow in intensity during the course of the research/seeding activites, which is to be expected if the seeding is having the desired effect. The precipitation images show that this cloud produced 0.3 inches of rain. From this it is easy to jump to the conclusion that the seeding worked--the cloud was seeded and it rained. However, the images show that the reflectivity was increasing and it was raining before the cloud was even seeded. So, does this mean that the seeding did not work? We can't make that conclusion either because it's possible that the seeding accelerated the growth of the cloud and enhanced the amount and/or duration of the rain. This case illustrates the main difficulty of evaluating cloud seeding: it is impossible to know with certainty what would have happend if the cloud had not been seeded.  

5/8/2000