Laserfiche WebLink
simulation of aerosol effects on cloud microphysics. The simulation of cloud seeding is performed by a <br />corresponding increase in the concentration of aerosols in the aerosols size distribution. <br />The focus of the modeling effort for SPECTRA II was on the numerical simulations of hygroscopic <br />seeding of Texas summertime clouds. These clouds are characterized by high vertical velocities (~15-25 m/s), <br />high droplet concentration (~700-1000 cm-3) and narrow droplet size distributions in the growing stage <br />(Rosenfeld and Woodley 2000, Khain et al, 2001b). Droplet spectra obtained in non-seed and seed numerical <br />experiments were compared with in situ observations made during the SPECTRA II field experiments.The <br />specifics of the cloud model and the detailed results are provided in the body of this Final Report. <br />During the simulation of hygroscopic seeding effects it was determined that the “competition effect” (as <br />defined herein) is of the secondary importance under any realistic concentrationof seed particles. The main <br />mechanism of rain enhancement is the formation of large droplets on large seed particles, which are able to <br />trigger the droplet collisions earlier than under natural conditions. Cloud seeding itselfwas simulated by adding <br />seed aerosols to the size distribution of the natural aerosols. The seeding was simulated ~100 m below cloud <br />base. The growth of all aerosols including seed particles to the cloud base level was simulated using a cloud <br />parcel model. Seed aerosol particles were added to the natural aerosol size distribution at t=0, and then, all seed <br />aerosols were excluded from the aerosol distribution below cloud base at t=20 min (10 min after cloud <br />formation). Thus, the duration of cloud seeding was about 10 minutes. <br />The numerical simulations of cloud seeding examined the effects of seeding with salt powder and the <br />effects of seeding with hygroscopic flares. <br />The salt powder was assumed to consist of similar size particles (monodisperse size distributions). This <br />assumption is based on actual measurements of the size distribution of the salt powder that had been processed <br />in the laboratory for SPECTRA II. Salt powders with different concentrations of seed particles (but a similar <br />mass of the reagent) were tested. In order to simulate also the logistic constraints of aircraft carrying capacity <br />of seeding material, the total mass of the reagent was kept constant inmost simulations. Therefore, the higher <br />concentration of seed particles corresponds to the smaller sizes of seed particles. The results of seeding with <br />different salt powders were compared with the result obtained using salt powder of optimum size, consisting of <br />mm <br />the 2 -radius particles. The concentration of 2 -radius CCN in the seed zone was assumed equal to 2 <br />3 <br />cm <br />3 <br />km <br />. This concentration corresponds to ~100 kg of the seeding mass. <br />In the second set, simulations of cloud seeding with hygroscopic flares were performed. The major <br />difference between salt powder and flares was in the shape of the size distributions of the seed particles. While <br />salt powder was assumed to consist of particles of similar size, flares have a wide size distribution of particles. <br />The size distributions of seed particles in different flares were examined. In the simulations the commercial so- <br />called "French" flare having the size distribution with the maximal fraction of large particles was used. In <br />different simulations the masses of the seed flares were changed within a wide range. <br /> To investigate the effects of seeding on clouds with different top heights and vertical velocities, clouds <br />were triggered with initial temperature heating of 15 min and 5 min durations. Because the sounding used had <br />significant atmospheric instability, the clouds reached high levels in all cases, but differed by the rate of cloud <br />growth and the maximal vertical velocity values. <br /> The assessment began with a comparison of the simulated cloud properties with those documented by <br />Rosenfeld and Woodley (2000) with the result that the agreement was quite good for the simulated cloud having <br />the longer heating duration. The assessment of seeding effects followed. <br />8 <br />