Seasonal Variation of Cloud Systems over ARM SGP

Abstract

Abstract Increased observational analyses provide a unique opportunity to perform years-long cloud-resolving model (CRM) simulations and generate long-term cloud properties that are very much in demand for improving the representation of clouds in general circulation models (GCMs). A year 2000 CRM simulation is presented here using the variationally constrained mesoscale analysis and surface measurements. The year-long (3 January–31 December 2000) CRM surface precipitation is highly correlated with the Atmospheric Radiation Measurement (ARM) observations with a correlation coefficient of 0.97. The large-scale forcing is the dominant factor responsible for producing the precipitation in summer, spring, and fall, but the surface heat fluxes play a more important role during winter when the forcing is weak. The CRM-simulated year-long cloud liquid water path and cloud (liquid and ice) optical depth are also in good agreement (correlation coefficients of 0.73 and 0.64, respectively) with the ARM retrievals over the Southern Great Plains (SGP). The simulated cloud systems have 50% more ice water than liquid water in the annual mean. The vertical distributions of ice and liquid water have a single peak during spring (March–May) and summer (June–August), but a second peak occurs near the surface during winter (December–February) and fall (September–November). The impacts of seasonally varied cloud water are very much reflected in the cloud radiative forcing at the top-of-atmosphere (TOA) and the surface, as well as in the vertical profiles of radiative heating rates. The cloudy-sky total (shortwave and longwave) radiative heating profile shows a dipole pattern (cooling above and warming below) during spring and summer, while a second peak of cloud radiative cooling appears near the surface during winter and fall.