WMOD00270 - Laserfiche WebLink

Home Browse Search

In the simulation of the effects of salt-powder seeding the droplet size distributions (DSD) observed in the seeded and non-seeded parts of REAL clouds were compared with those simulated at different heights in the seeded and non-seeded model clouds. The observations of seeding effects were obtained from the second actual seeding case on May 31, 2005 that is described in detail in this report. This cloud had microphysical properties that were similar to those observed by Rosenfeld and Woodley (2000) in another vigorous cloud that was studied near Midland, Texas. During the simulations it was noted that the simulated DSD’s in the seeded cloud were wider than those simulated in the non-seed clouds. Thus, the modeled seeding signature agrees quite well with the seeding signature that was observed in SPECTRA II, as documented herein. Such agreement between observations and simulations makes the results of the simulations much more credible. The results indicate that seeding with salt powder leads to acceleration of raindrop formation. The DSD in clouds with lower vertical velocity (L clouds) are wider anddevelop at lower levels as compared with clouds with higher vertical velocities (H clouds). Further, raindrop formation is faster in the L-clouds. The particle size m found to be optimum (2-radius) in simulations with the parcel model turned out to be optimum in m simulations using the 2-D HUCM as well. Seeding with particles of 2-radius having concentrations of 2 3 cm provided a 12%-25% increase in the accumulated rain from a deep convective cloud. In the unseeded L- cloud the precipitation began earlier and was actually twice as large as compared to that in the H-cloud due to the lower droplet concentrations and wider DSD in the L-cloud. Even so, the simulated effects of seeding in the L-cloud were smaller than in the H-cloud. The effects of seeding with hygroscopic flares also were simulated, leading to the following conclusions: a) Seeding with flares turned out to be less efficient than that with salt powder. Even when the mass of flares was similar to that of the salt powder, rain enhancement was smaller than when seeding with thesalt powder. When a smaller flare mass was used, meaning a smaller concentration of large CCN in the simulations, the seeding leads to a decrease in the accumulated precipitation becausethe positive effect due to the large CCN in the flares is offset by the negative effect of the many small CCN in the flares. The small CCN lead to an increase in the concentration of small droplets, which is tantamount to making the clouds more like “continental” clouds that produce less precipitation. b) Seeding with flares is less effective for the L-cloud. The DSD in the L-cloud is relatively wide even under natural condition. Consequently, this cloud is less sensitive to thelarge CCN injected by the seeding, but more sensitive to the additional small CCN. Utilization of flares of the same mass as the mass used in the salt powder seeding led initially to a ½÷ similar radar-reflectivity seeding signal. Nevertheless, rain enhancement caused by seeding with flares turned out to be less than in the case of seeding with salt powder, as was mentioned above. Supplemental simulations with an increased maximum size of seed particles (up to 5 um in radius), but with a corresponding decrease in the concentration of large CCN in order to keep the reagent mass unchanged, led to some increase of the radar seeding signal at the time of first radar echo formation. However, this increase in the maximum size of seed particles (accompanied by the decrease in their concentration) did not lead to additional precipitation amounts, possibly because the concentration of largest drops was too small. This result agrees well with the conclusion about the existence of the optimum size (~2-2.5 um) of largest CCN in the seed reagent. 5.0 CONCLUSIONS AND RECOMMENDATIONS 9