Comparison of HCMM and GOES satellite temperatures and evaluation of surface statistics

Abstract

Surface temperatures detected by infrared sensors of HCMM and GOES-East satellites were compared for a clear winter nocturnal period in Florida. The purpose was to examine spatial variation of surface temperatures within GOES pixels (approximately 48 km 2) using nested HCMM pixel surface temperatures. The areas in the Florida peninsula used for comparison ranged over 3° of latitude and included dry pine forests, pastures, and crops and moist wetlands and lakes. Temperature means and standard deviations were calculated for the HCMM pixels nested in each GOES pixel for comparison. Standard deviations were plotted against mean surface temperatures for an area in central Florida of about 1900 km 2 and another area in southern Florida which was three times larger. The standard deviations of HCMM pixel (0.22 km 2) temperatures associated with mean temperature of a GOES pixel (about 48 km 2) temperature during this winter nocturnal period were smallest for homogeneous water surfaces (about 0.5 K) and slightly larger for homogeneous land surfaces (about 1.0 K), but ranged up to as high as 3.5 K for mixed water-land surfaces. Standard deviation vs. mean temperature curves of pixels nested in larger areas can characterize the homogeneity vs. inhomogeneity of a large area.