Determining Target and Control Areas in Operational Cloud Seeding Programs over Iran by Dispersion Simulation of Seeding Materials Using HYSPLIT Model

Determination of affected area by seeding agents, the so-called target area, is an essential requirement for evaluation of cloud seeding projects. The most conservative and credible estimates of seeding effects were obtained from control matches drawn from outside the operational target within 2 hours of the time that each unit was seeded initially (DeFelice et al., 2014). A coupled modeling system consisting of the mesoscale WRF model and the Hybrid Single-Particle Lagrangian Integrated Trajectory Model (HYSPLIT), provides capability to simulate the transportation and dispersion of seeding materials and to characterize target area on the map.
This study is devoted to sensitivity analysis of simulated dispersion patterns to several parameters including different configuration based on physical parameterizations used in WRF model, horizontal and temporal resolution of WRF and spatial resolution of HYSPLIT, to determine the most probable dispersion patterns.
Since temperature and wind parameters are the most important parameters in cloud seeding operations, they are measured instantaneously at 1-second intervals at the flight height of the airplane during each flight and therefore, they are very valuable data to assess the performance of the WRF model in simulating these fields. Hence, at first the WRF model outputs such as temperature and wind are validated by data measured by the airplane. Results indicate that there is an acceptable agreement between field data and WRF outputs that are going to be used as input data for dispersion model.
In this study, eight configurations of the WRF model based on different physical parameterization schemes are used for 34 flights in cloud seeding project in 2015 and HYSPLIT model is run by these types of input data and resulting target area are compared on the map. Then, HYSPLIT model is run for four selected seeding operations according to three temporal and two horizontal resolutions of input data in addition to three spatial resolutions of HYSPLIT model and the transport of seeding plumes is characterized on the geographical map.
The results indicate that dispersion model is sensitive to all mentioned parameters. Also, in most cases, dispersion model results at the flight height of cloud seeding aircraft are significantly influenced by the input data provided by the WRF model. In addition, the dispersion model results are less sensitive to other parameters. Furthermore, when the spatial resolution of the HYSPLIT model is close to the horizontal resolution of the input meteorological data provided by the WRF model, affected area of seeding agents is more integrated and therefore there is a greater degree of certainty in determining the target area.