A new methodology for observation-based parameterization developme

We develop a methodology for identification of candidate observables that best constrain therepresentation of the most uncertain processes in physical parameterizations. This methodologyconsists of the three steps: (i) identifying processes that significantly impact parameterization results,(ii) identifying observables that best constrain the influential processes, and (iii) investigatingthe sensitivity of the parameterization results to the measurement error and vertical resolution ofthe constraining observables. This new methodology is applied to the Jet Propulsion Laboratorystochastic multi-plume Eddy-Diffusivity/Mass-Flux (JPL-EDMF) parameterization for two casestudies representing non-precipitating marine stratocumulus and marine shallow convection. Wefind that the most uncertain processes in the JPL-EDMF model are related to the representationof lateral entrainment for convective plumes and parameterization of mixing length scale for theeddy-diffusivity part of the model. The results show a strong interaction between these uncertainprocesses. Measurements of the water vapor profile for shallow convection and of the cloud coverprofile for the stratocumulus case are among those measurements that best constrain the uncertainJPL-EDMF processes. The interdependence of the required vertical resolution and error characteristicsof the observational system are shown. If the observations are associated with larger error,their vertical resolution has to be finer and vice versa. For example, it was found that for the shallowconvection case, the measurement of water vapor with a vertical resolution of 20 m and error of4 g kg􀀀1, or with a vertical resolution of 200 m and error of 2 g kg􀀀1 provide similar constraintfor the model parameters. We suggest that the methodology presented here and the results hereinprovide an objective basis for defining requirements for future observing systems such as futuresatellite missions to observe clouds and the planetary boundary layer.