Improved Representations of Cloud-Scale Processes in Meteorological ForecastModels

摘要

The functional relationship between cloud cover and relative humidity (Rh)averaged over areas comparable to grid dimensions of numerical weather models was quantified using RTNEPH and 3DNEFH observations. Cloud cover in any atmospheric level decreases exponentially as layer averaged Rh tails below 100%, and no observations support critical Rhs below which cloud cover is zero. Small cloud amounts occur at all Rhs. Therefore, current weather models probably underestimate cloud coverage, especially at Rhs below the critical humidities used by most models. At the same Rh, convection enhances cloud coverage in the upper troposphere. and decreases cloud coverage in the lower troposphere. Developed a simplified and innovative mass flux convective parameterization that was evaluated using atmospheric radon profiles, and was also used to simulate the redistribution of heat and moisture by combining the approach of stochastic mixing with detraining plumes. A public domain cloud resolving model (ARPS) was used to further refine the 1-D parameterization. Both the cloud resolving models and the convective parameterization were evaluated using GATE observations. However the ARPS model employs an advection algorithm that does not conserve water mass, making it unreliable to use for refining cloud parameterizations.