WMOD00318 (2) - Laserfiche WebLink

Home Browse Search

APRIL 1995 MEYERS ET AL. 845 the peak precipitation rates in the seeded run were 1.0- l.5 mm h-I (not shown). 5. Summary and conclusions This study has described numerical studies of a seeded orographic precipitation event, with particular detail to implementing explicit ice initiation by artificial ice nucleus aerosols into mesoscale models. In this ef- fort, the results from laboratory studies at CSU (DeMott 1990, 1994) were parameterized, quantifying four ice nucleation mechanisms for the seeding aerosols used in SCPP. Parameterizations were first imple- mented in a microphysical parcel model compare to DeMott's detailed descriptions of artificial ice forma- tion processes for a representative seeded orographic cloud parcel. Excellent agreement was found. These experiments also indicated that condensation-freezing and deposition nucleation should initially dominate artificial nucleation, with contact-freezing nucleation becoming dominant during parcel ascent and cooling at later times. The contribution of immersion-freezing was deemed insignificant. Earlier papers (DeMott 1991; Deshler and Reynolds 1991) have debated whether or not the contact-freezing mechanism alone could ex- plain ice initiation following seeding of orographic clouds. These new results show that the implementa- tion of more recent laboratory results, including other possible modes of ice formation, does adequately de- scribe ice initiation following seeding. The parcel model simulations also suggested that predicted ice crystal concentrations would be underestimated by a factor of about 3 if one does not account for the change in the aerosol generator PSD during airborne generation. As a consequence of the preliminary calculations, im- mersion-freezing nucleation was not considered in the mesoscale model simulations, and initial ice nucleus aerosol concentrations and subsequent activation were calculated assuming a PSD characteristic of aircraft generation conditions. The seeding parameterizations were applied to 3D versions of RAMS that examined the 18 December 1986 SCPP case. This explicit depiction of ice initiation by cloud seeding aerosols represented a unique appli- cation for mesoscale models. Three-dimensional sim- ulations of the SCPP case study showed a profound sensitivity to seeding. The model produced good agreement between the simulated natural cloud pro- cesses and the observed microphysical structure. The seeded simulation produced large areas with ice crystal concentrations exceeding 20 L - I and peak pristine ice crystal concentrations of 120 L -1 , an order of mag- nitude greater than the ice crystal concentrations pro- duced in the nonseeded run. These simulated concen- trations were similar to the observed maximum ice crystal concentrations (100 L -I). Seeding also pro- duced increased pristine ice crystal, aggregate, and graupel mass downstream of the seeded regions in the ~~ seeded simulation. Overall precipitation increases due to seeding were 0.1-0.7 mm, similar to values inferred from the observations. Precipitation enhancement in the seeded simulation resulted from increased precip- itation efficiency since no marked regions of precipi- tation deficit occurred due to seeding. These precipi- tation maxima were collocated with regions of SL W that were enhanced by topographic forcing. At least two factors for future development of cloud- seeding parameterizations in mesoscale models were identified in this study. First, it was noted that repre- sentation of seeding in mesoscale models will probably always be compromised to some extent by the ability to resolve initial concentrations and dispersion of ice nucleus aerosols. Namely, the width of the initial seed- line was on a scale much smaller than the very fine resolution (~x = 1 km) used in this simulation. Subgrid-scale dispersion schemes are needed to address this problem. The ice nucleation schemes used are also very sensitive to small changes in the temperature and humidity, so that subgrid-scale changes in these ther- modynamic quantities could also be important. This study has demonstrated the feasibility for using laboratory results to parameterize the cloud seeding response in an explicit cloud model. These simulations also show promise for a priori evaluations of seeding effects on different cloud types given complete descrip- tions of ice nuclei behavior. Uncertainties remaining from the standpoint of operational seeding are due to a lack of more complete knowledge of the transient adjustment of temperature and humidity (supersatu- ration) following combustion, as well as, a lack of better knowledge of particle sizes generated. Resolution of these uncertainties appears quite feasible. Acknowledgments. Thanks are extended to Drs. Da- vid Rogers, Johannes Verlinde, and Robert L. Walko for valuable suggestions. We also appreciate the co- operation of SCPP, the WMO, and other participants of the Third International Modeling Workshop who provided field data and valuable insights on this case. Brenda Thompson helped with the processing of this manuscript. This research was supported by Na- tional Science Foundation Grants A TM9118963 and A TM9103748. The lead author would also like to ac- knowledge the DoD Augmentation Awards for Science and Engineering Research Training under Grant F49620-92-J-0331 M and the Air Force Office of Sci- entific Research under Grant AFOSR-91-0269 for supporting his graduate research. REFERENCES Blumenstein, R. R., R. M. Rauber, L. O. Grant, and W. G. Finnegan, 1987: Application of ice nucleation kinetics in orographic clouds. J. Climate App/. Meteor., 26,1363-1376. Cooper, W. A., 1974: A possible mechanism for contact nucleation. J. Atmos. Sci., 31, 1832-1837. Cotton, W. R., M. A. Stephens, T. Nehrkorn, andG.J. Tripoli, 1982: The Colorado State University three-dimensional cloudjme-