Assessment of the CALIPSO-CloudSat-CERES-MODIS (CCCM) data product at TOA and surface

Poster PDF

Description

The Clouds and the Earth’s Radiant Energy System (CERES) project calculates irradiance profiles beneath CERES observations in the Clouds & Radiation Swath (CRS) data product. Cloud properties and profiles required to execute the radiation transfer code depend upon a cloud retrieval algorithm that utilizes MODIS pixels collocated within CERES fields of view (FOV). A new product, the CALIPSO-CloudSat-CERES-MODIS (CCCM) product, merges cloud vertical profiles retrieved from these “active” sensors, providing additional information that improves estimates of atmospheric radiative heating rates and potentially the estimate of the global surface radiation budget. These CALIPSO and CloudSat cloud vertical profiles are combined as well, with aerosol properties derived from CALIPSO in addition to those derived from directly from MODIS. Regarding assessment of the irradiance calculations, two problems present themselves. The first regards comparisons at the top of the atmosphere (TOA). The clouds analyzed by the merged product represent a “slice“ through the CERES FOV, not covering its entire extent. Over time, assuming cloud fields are uniform across the FOV, this would increase the RMS of the calculations compared with CERES observations but should not introduce a bias. At the surface, another sampling problem arises. The merged data is available only along the Nadir path of the CERES FOV; sampling around the globe is significantly smaller than for the full-swath CRS product. Thus matching FOV to surface observations, given the orbit track of Aqua, is quite problematic as FOVs may never directly view surface locations where irradiance observed at the surface is available. We analyze a subset of CCCM irradiance calculations based on the surface location of CERES footprints located to within 150 km of 25 surface locations that have long-term records of observed surface irradiance. Surface sites are grouped to increase sample size, and statistics are presented at both TOA (compared to CERES observations) and at the surface (compared to surface radiometry). We find the primary improvement over nighttime skies is because the active sensor’s ability to retrieve low clouds trumps the passive methodology. This is best seen at polar surface locations where the longwave, surface-down irradiance bias (model-observation) drops from -12Wm-2 down to +1Wm-2. and RMS is reduced by approximately 5Wm-2 at the four polar site surface locations.