Observing Recent Cloud and Radiation Budget Changes over the Arctic

The warming trend in the Arctic in recent decades is approximately twice as large as the global average. The rapid decline in Arctic sea-ice is also well documented. Warmer temperatures and an increase in coverage of ice-free ocean should moisten the boundary layer, alter cloud cover, and potentially impact the top-of-atmosphere (TOA) and surface radiation budgets. Indeed, recent observational studies have shown that the decline in Arctic sea-ice coverage is highly correlated with TOA albedo, both in clear and all-sky conditions. However, a subtler question is how cloud changes influence the radiation budget. Do cloud changes offset or enhance the impact of decreasing sea-ice coverage on the radiation budget? Answering this question from satellite observations is challenging owing to well-known issues with remote sensing of clouds in Polar regions from passive instruments. Even during daytime, discriminating between clear and cloudy conditions over bright surfaces such as sea-ice and snow is difficult, and even more so during Polar night, owing to a weak thermal contrast between low clouds and the surface and persistent temperature inversions. Further complicating matters is the high degree of natural variability over the Arctic at interannual time-scales, which can mask a real trend in cloud radiative effects. In this presentation, we provide a critical evaluation of satellite observations in the context of observing cloud and radiation changes over the Arctic. TOA and surface radiation observations are from the Clouds and the Earth's Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) product. Of particular interest is whether the approach used to infer clear-sky TOA radiative fluxes from both clear and partly-cloudy CERES footprints is sufficiently robust at high latitudes to enable the inference of changes in cloud radiative effects. A key component of this is the reliability of clear/cloudy discrimination from the Moderate Resolution Imaging Spectroradiometer (MODIS), which is used in CERES data processing to provide scene information within CERES footprints (e.g., cloud mask, cloud and aerosol properties, and surface temperature). At the surface, we evaluate downward surface longwave (LW) radiation changes inferred from MODIS cloud properties and reanalysis atmospheric state data. Finally, we use the CloudSat, CALIPSO, CERES and MODIS (C3M) merged instantaneous cloud-radiation dataset to verify the cloud property variations from MODIS, as derived in the CERES processing system.