In high-resolution numerical weather prediction models, bulk microphysical parameterization schemes play a particularly important role in the model-produced precipitation forecasts. However, comprehensive data sets needed to verify the processes and hydrometeor fields simulated by microphysical schemes, and to isolate errors in these schemes from other sources of model error, have not been available. To fill this need, a group of researchers has embarked on a study entitled IMPROVE (for Improvement of Microphysical Parameterizations through Observational Verification Experiments) to compare representations of cloud and precipitation processes in current mesoscale models with detailed observations in a variety of weather systems, with the goal of improving quantitative precipitation forecasts produced by mesoscale models.

In this presentation we review the scientific background to IMPROVE and the two series of field studies that have been carried out. We then present an overview of current research efforts to analyze cloud and precipitation processes using the field study data sets, to simulate the observed cases with high-resolution numerical model simulations, and to compare microphysical processes and production of clouds and precipitation in model simulations with those observed. Two case studies will be discussed: the 1-2 February 2001 storm over the northeast Pacific Ocean, and the 13-14 December 2001 storm over the Oregon Cascade Mountains. Several of the research efforts touched upon here will be discussed in greater detail in accompanying presentations on IMPROVE.