WMOD00442 - Laserfiche WebLink

Home Browse Search

1758 JOURNAL OF APPLIED METEOROLOGY VOLUME 17 A Model of Hygroscopic Seeding in Cumulus Clouds GERARD E. KLAZURA AND CLEMENT J. TODD Bureau of Reclamation, Denver, CO 80225 (Manuscript received 15 August 1977, in final form 31 July 1978) ABSTRACT A systematic modeling exploration has been conducted to map the growth and trajectory of hygroscop- ically initiated precipitation particles. The model used is a one-dimensional, steady-state, condensation- coalescence model with adiabatic cloud water content. Drop breakup and freezing were simulated but competition among precipitation particles was not considered. Sizes of initial, hygroscopic seeds varied from 5 to 4OO,um in diameter, updraft speed ranged from I to 25 m s-', and cloud base temperature varied from 0 to 200C. The 23 July 1970 salt seeding case reported by Biswas and Dennis was also analyzed using the model. The numerical simulations reveal several complex interactions: 1) For slow updrafts, the larger hygro- scopic seeds travel through a lower trajectory and sweep out less water than small, hygroscopic seeds which are also more apt to grow large enough to break up and create additional large precipitation particles. 2) For fast updrafts, the larger hygroscopic seeds grow into precipitation and stand a better chance of breaking up and initiating a Langmuir chain reaction, while the small hygroscopic particles are carried up to the cirrus level and are lost before they reach precipitation size. 3) For very strong updrafts only large hygroscopic seeds will have a chance to convert to precipitation, and in this situation hail is produced. 4) Hygroscopic seeding produces a greater water yield from warmer based clouds. 1. Introduction Because there are many interacting factors which determine the characteristics of a cloud and how it produces precipitation, gt:neralizations must be made carefully. The purpose of this study has been to produce a generalization which necessarily is based upon a simplified set of assumptions. Any less re- strictive set of assumptions would have caused the computational requirements to increase dramatically . and increase the complexity of the interpretation. The. factors chosen for this study are well suited for com- puter rather than intuitive exploration in that they require a large number of integration steps. However, the interpretation of the output was designed to be adjusted by conceptual reasoning for making judg- ments about precipitation processes and management of precipitation in real clouds. Observation of the growth of a single hydrometeor under the influence of different cloud characteristics can lead to a good first estimate from which some general conclusions can be drawn. This observation has been simulated through the use of a simple cloud precipitation model which has in turn been used to explore the opportunity for producing artificially initiated precipitation (using hygroscopic seeding material) when the natural mechanisms are too slow to be effective. Similar modeling investigations have been carried out by Takeuchi (1975) and Rokicki and Young (1978) using one-dimensional, steady-state, parcel microphysical models. Results of the present investi- gation will be compared with findings of these two works. 2. Single hydrometeor condensation-coalescence model This study uses a one-dimensional condensation- coalescence model that follows the growth of a single precipitation particle in a specified cloud environ- ment. It is a steady-state version of the feeder cloud model described by Musil (1970). Both the dry and wet hail growth equations were utilized as applicable [see Eqs. (A6) and (A 10) , Musil (1970)]. A con- densation subroutine was added to the model [see Eq. (1), Chien and Mack (1966)]. The details of the model including derivations of growth equations and references and equations used for such items as ter- minal velocities of precipitation particles, collection efficiencies, ventilation coefficients, air viscosity, dif- fusivity and conductivity are discussed by Musil (1970). The processes simulated are as follows: non-com- peting hygroscopic particles are released in the updraft region beneath the cloud and grow by condensation until they penetrate the cloud base. Then both con- densation and coalescence occur until the particle has grown to a 100 J.lm diameter,' at which point growth 1 Unless otherwise noted, all sizes refer to diameter.