COMPARISON OF THE HIGHLY REFLECTIVE CLOUD AND OUTGOING LONGWAVE RADIATION DATASETS FOR USE IN ESTIMATING TROPICAL DEEP CONVECTION

Currently, there are two long-term satellite-derived datasets most are frequently used as indices for tropical deep convection. These are th e Outgoing Longwave Radiation (OLR) and Highly Reflective Cloud (HRC) datasets. Although both of these datasets have demonstrated their valu e, no direct comparison of these datasets has been conducted, to deter mine how well they agree when used to estimate tropical convection, no r has there been much work toward comparing these long-record datasets with more recently developed convection datasets. This information is vital since the inhomogeneous sampling of the in situ rainfall record makes it inadequate for many studies concerning tropical convection a nd the more modern datasets have not achieved a climatologically usefu l record length for all studies. The goal of this paper is to compare these two datasets in order to quantify their strengths and weaknesses . This information will provide guidance in choosing the most appropri ate dataset(s) for subsequent studies, interpreting the results from t hose studies, and extending more modem convection datasets backward. C omparisons are done in terms of their climatological and frequency-dep endent characteristics, their consistency in identifying deep tropical convection, and their relationships to local sea surface temperature (SST). Additionally, use is made of the more modem, shorter-term Inter national Satellite Cloud Climatology Project stage C2 dataset as a mea ns of further comparison and validation. The results of this study rev eal some important differences between the HRC and OLR in terms of the ir temporal and spatial scales of variability, their relationships to other geophysical fields, and the logistics of their use. Further, the y suggest that for many applications the HRC more accurately represent s the characteristics of cloud cluster-scale tropical convection. This is especially true in cases where 1) characterization of the spatial scales or frequency-dependent variability of convection is important, 2) the relationships between deep convection and SST or water vapor ar e being considered, and 3) the domain of interest is large enough to c ontain spatial inhomogeneities, such as land-sea contrasts or inhomoge neous SST and moisture fields. One new and important finding of this s tudy is that both the OLR-SST and HRC-SST relationships show that SSTs in excess of about 29.5-degrees-C tend to occur only under conditions of diminished convection. Thus, the maximum convective activity does not occur over the warmest (>29.5-degrees-C) water; rather, the warmes t water occurs under ''clear,'' less-convective skies. Further, our re sults empirically demonstrate that in a highly convective regime the m aximum equilibrium SST that can be supported is about 29.5-degrees-C. These results are further evidence that convective-cloud complexes pro vide a systematic and climatologically important cooling effect on the surface temperature.