6.3 An Overview of Southern Sierra Nevada Cloud Seeding Programs

Thursday, 16 January 2020: 12:00 AM
105 (Boston Convention and Exhibition Center)
Richard H. Stone, RHS Consulting, Ltd., Reno, NV; and D. Munn, M. Larsen, and D. L. Newsom

This presentation will provide an overview of RHS cloud seeding operations and discuss statistical streamflow assessments of the Kern and Kaweah River Basin cloud seeding programs. Results from these programs will be compared with past results, including those found by Silverman (2010), and how fundamental differences in statistical analysis techniques affect confidence intervals. Recent streamflow trends in these target areas will also be compared with river basins located north of the two target areas. Both programs are seeded using aircraft and have been operating in the Southern Sierra Nevada of California, USA for more than forty winter seasons.

This presentation will also discuss a new confirmatory experiment designed to determine the economic efficacy of these operational programs by focusing on water on the ground. The purpose of the experiment will be to conduct a confirmatory test of the current operational programs. The proposed experiment will not be randomized because it is not a viable option for these operational programs. New California regulatory requirements mandate sustainable ground water supplies that have forced increased accounting levels and the cost of surface water to record levels, even during above average water years, making non-seeded storms a luxury.

Over the last two decades the level of proof defined by the scientific community for cloud seeding programs has evolved to include randomization, robust statistical evidence, extensive physical measurements, understanding and replication. This is a very high standard for a system as complex as the atmosphere. The requirement for extensive physical measurement including the serial link precipitation process chain of events is desirable and undoubtedly helps prevent misinterpretation of data, but it isn’t a necessary condition for proof. Proof can be provided using other controls and covariates, including physical and/or chemical tracers, that eliminate the need to follow the detailed links in the precipitation process chain.

The level of proof required by water managers is less rigorous and does not require the robust definition and standards advocated by the scientific community. Water managers view cloud seeding as a water management tool and recognize the need to continue operational programs during wet years to replenish aquifers, restore water table levels to prevent subsidence and bank surface water for future drought years. The decision to continue to operate these programs becomes a matter of risk assessment and cost benefit. Feedback from the scientific community is welcome regarding improvements in methodology and assessment techniques.

Most recent winter seeding experiments have been designed to demonstrate causal physical links in the precipitation process chain of events. These experiments have been very successful but fall short of demonstrating proof of cloud seeding efficacy and meaningful estimates of water on the ground for end users. This design fails when studying complex systems because of the natural variability of precipitation coupled with the difficulty of clearly detecting seeding signatures in noisy environments. To succeed these programs have focused on relatively simple systems to test hypotheses designed to follow the link in the chain process. We propose using multiple hypotheses and response variables not dependent on one another to allow seeding effects in complex storms to be evaluated at any stage of the precipitation process

To measure the impacts of cloud seeding, we propose a more pragmatic approach by selecting response independent variables to measure changes in precipitation at various points in the physical chain of events and at the ground level. These response variables in no order of the physical change of events include: 1) Measuring the depletion of integrated liquid and vapor by the seeding process from upwind to downwind of aircraft seed track; 2) Polarized doppler X band radar and ground-based radiometer measurements to document changes in cloud structure up and downwind of the seed track; 3) Precipitation chemistry to fingerprint air parcels and identify changes in seeded and non-seeded precipitation at the ground; 4) Using both aircraft and ground observations to determine snow water content across the overall target and control watersheds after significant storm events; and 5) Using a lake as large-scale precipitation gauge and streamflow to quantify seeding effects at the ground.

Physical evidence will be developed using a combination of the trace chemistry of snowfall, tagged ice-nucleating and non-ice nucleating aerosols and modeling techniques combined with other atmospheric and precipitation measurements to identify and fingerprint the end links between seeded air volumes and precipitation falling at the ground. These techniques have been used successfully in the past in several Sierra and one Australian winter seeding program. The use of a non-ice nucleating indium control tracer and major chemical “fingerprinting” as a control of natural non-seeded precipitation will allow the experiment to be piggybacked over existing operational seeding programs. Release of the tracers in seeded air parcels and subsequently identifying them at ground level in the target will provide physical evidence of successful transport and relative removal rates. It is not necessary to identify the specific chain of events that lead to removal of the seeding materials or scavenging of the control tracer at the ground level.

These measurements will allow more complex cloud systems to be studied successfully. This approach is analogous to many other complex systems that require calibration to understand the basic input output functions of a system. Real world examples include neural nets, biomedical drug testing, chemical reactions, engine tuning, load forecasting, financial trends, etc. These systems only require rudimentary knowledge of how the system works, not the fine details and processes. With enough data proof can be established by conducting confirmatory studies.

The design of the confirmatory experiment will include a discussion of the experimental framework, sampling requirements, instrumentation, seeding materials, and conceptual model used to conduct these operational programs and other pertinent details.


Silverman, B. A., 2010: An Evaluation of Eleven Operational Cloud Seeding Programs in the Watersheds of the Sierra Nevada Mountains. Atmospheric Research, 97, 526-539.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner