On the Use of Probability of Clear Line of Sight Models in Parameterizing Surface Downwelling Longwave Radiation in the Tropical Western Pacific

Cloud parameterizations within Global Climate Models (GCMs) accounts for the largest uncertainty in current climate change predictions. One reason for this is that current methods for including sub-grid scale cloudiness invoke unrealistic assumptions about cloud field statistical properties. Specifically, the plane parallel horizontal (PPH) cloud assumption used in parameterizing longwave radiation under broken cloudiness unrealistically neglects three-dimensional cloud radiative effects. While there is currently no broadly accepted solution to this problem, any solution must be physically realistic and tested using a wide array of observations. This study investigates the probability of a clear line of sight (PCLoS) as a possible solution. PCLoS is estimated for single-layer tropical marine cumulus cases for 2-hour intervals using the Whole Sky Imager (WSI) observations at the Atmosphere Radiation Measurement (ARM) Program's Tropical Western Pacific (TWP) sites at Manus and Nauru Islands. PCLoS estimates are compared to calculations from a set of simple PCLoS models using measured cloud field statistics as input. The results suggest that most PCLoS models out perform the PPH assumption in parameterizing the longwave downwelling surface flux; in addition, the best performing PCLoS models exhibit mean absolute errors < 1 Wm(-2) when compared to observations. This study asserts that PCLoS models represent a viable solution to the cloud-radiation parameterization problem for small cumulus clouds; however, more work is needed to investigate the global applicability of the PCLoS models.