Comparison of Trends in Arctic Cloud Vertical Distribution and LW Radiation Fluxes Inferred from Ground-Based Sensors at Eureka Station, A-Train Satellite Observations and ERA-I Reanalysis

Arctic surface temperature is rapidly increasing, more than elsewhere in the globe, leading to a strong and steady decrease of sea ice cover. Observations of Arctic sea-ice extent indicate that it is decreasing at a rate of about -3%/decade in summer, and a smaller decrease is also observed during winter. Identification of the processes leading to such changes remains a research priority. Atmospheric radiative forcing, for example as modulated by variability in clouds and water vapor, plays an important role in the evolution of surface energy budget, and is therefore important also for variability in sea ice. These processes are complicated by feedbacks that, in turn, impact cloud and atmospheric properties. To this end, gridded data products, such as those derived from satellite sensors, which provide long-term monitoring with broad spatial coverage, but do not directly observe the surface energy budget, must be characterized using direct observations from the surface at locations where such measurements are available.

In this study, we first focus on analyses of cloud fraction and vertical distribution, and related surface radiation fluxes at the Eureka station in the northern Canadian Archipelago (80°N, 86.4°W) using various active and passive sensors at the surface at Eureka (e.g., lidar, radar, broadband radiometric observations) and from satellite platforms with regular overpasses in the vicinity of Eureka (e.g., products such as CERES-SSF and 2B-FLXHR-LIDAR), as well as data from the ECMWF ERA-Interim re-analysis. Our approach is to use both statistical analysis and coincident surface-satellite observations over multi-year periods to evaluate and then analyze the satellite products. In a second part, the study will be extended over other High-Arctic stations with similar instrumental capabilities, where cloud regimes are known to be different. Trends and uncertainties linked to errors in retrieved cloud properties and fluxes simulations will be quantified and discussed.