Assessing Glaciogenic Seeding Impacts in Australia’s Snowy Mountains: An Ensemble Modeling Approach

Australia’s Snowy Mountains form the headwaters of several major river systems and provide water resources for hydroelectric power, irrigation, and winter recreational activities. Wintertime glaciogenic seeding is undertaken in this region for the purpose of enhancing snowfall to increase water storage. To better quantify the precipitation change due to seeding and gain a process-level understanding of the seeding impacts on clouds and precipitation, we employed a cloud seeding model – WRF-WxMod® – to facilitate realistic simulation of cloud and precipitation processes in natural and seeded conditions in complex terrains.

Our recent numerical investigations demonstrate model sensitivities to varied configurations across different seeding cases, suggesting no one-size-fits-all model configuration. Therefore, this study aims to identify an ensemble design to quantify the range of uncertainties provided by the ensemble members for better estimating the overall seeding impacts. The ultimate goal of using this ensemble approach is to evaluate the long-term seeding impact over the past decades.

A total of nine distinct seeding cases from 2016 to 2019 are simulated. A nested domain setup is used, with a 4-km outer domain covering the major portion of eastern Australia and a 1-km inner domain enclosing the Snowy Mountains catchment. We take into account 12 ensemble members, each varying in initialization datasets, ice nucleation schemes, aerosol conditions, among other parameters. Each member consists of a SEED simulation and NO SEED (control) simulation, allowing for direct comparison and quantification of seeding impacts on cloud structure, cloud/precipitation process rates, and the distribution and amount of precipitation at the surface. The nine cases were clustered into four categories, featuring different meteorological regimes and varying levels of seeding potential. In this presentation, we will also show the seeding impact under the four clusters and assess their uncertainties due to different factors.