Extended Abstract (32K)P1.5
Southern Hemisphere precipitation; a study of model sensitivity to parameter settings
Howard R. Larsen, National Institute of Water and Atmospheric Research, Wellington, New Zealand; and R. Turner, D. S. Wratt, and M. Revell
The Southern Alps, rising quite abruptly from the ocean on the western side of southern New Zealand, represent a substantial barrier to the predominantly north-westerly mid-latitude winds. Precipitation development associated with the resulting orographic flow has important impacts for New Zealand. This precipitation development has been studied, using observational and modelling data, as part of the SALPEX research campaign.
Earlier numerical modelling has been unable to reproduce the observed precipitation amounts and distribution. Verification of the conditions used as input to the models, particularly the RAMS model which is widely used for mesoscale modelling, was therefore one of the goals of SALPEX. During the SALPEX field campaign, in-cloud measurements of cloud microphysics and dynamics were made using a research aircraft.
These measurements suggest the clouds off the coast west of New Zealand differ significantly from those (generally North American) used to establish the default parameterisation of the cloud and precipitation microphysics in RAMS.
Revised microphysical parameter settings were therefore derived and used as the basis for further numerical modelling using RAMS. With these new parameter settings the model produces more rain over the Alps and slightly less at the coast, and a more extensive downwind cloud shield. This brings its output closer to the observations. The correlation between the modelled and the measured precipitation time series over three cross-alpine transect areas was increased significantly. The improvements offered by the new parameter settings were also significant when the model was run at a higher spatial resolution (5km instead of 20 km).
Poster Session 1, Poster Session
Monday, 24 March 2003, 3:00 PM-3:00 PM
Previous paper Next paper