Toward a Better Surface Radiation Budget Analysis Over Sea Ice in the High Arctic Ocean: A Comparative Study Between Satellite, Reanalysis, and local‐scale Observations

Reanalysis datasets from atmospheric models and satellite products are often used for Arctic surface shortwave (SW) and longwave (LW) radiative budget analyses, but they suffer from limitations and require validation against local‐scale observations. These are rare in the high Arctic, especially for longer periods that include seasonal transitions. In this study, radiation and meteorological observations acquired during the Norwegian Young Sea Ice Cruise (N‐ICE2015) campaign over sea ice north of Svalbard (80–83°N, 5–25°E) from January–June 2015, cloud lidar observations from the Ice‐Atmosphere‐Ocean Observing System and the Cloud and Aerosol Lidar with Orthogonal Polarization are compared to daily and monthly satellite retrievals from the Clouds and the Earth's Radiant Energy System (CERES) and ERA‐Interim and ERA5 reanalysis. Results indicate that surface temperature is a significant driver for winter LW radiation biases in both satellite and reanalysis data, along with cloud optical depth in CERES. In May, the SW and LW downwelling irradiances are close to observations and cloud properties are well captured (except for ERA‐Interim), while SW upward irradiances are biased low due to surface albedo biases in all datasets. Net SW and LW radiation biases are comparable (∼20–30 Wm−2) but opposite in sign for ERA‐Interim and CERES in May, which allows for error compensation. Biases reduce to ±10 Wm−2in ERA5. In June downward LW remains biased low (8–10 Wm−2) in all datasets suggesting unsettled cloud representation issues. Surface albedo always differs by more than 0.1 between datasets, leading to significant SW and total flux differences. Surface albedo, temperature, and cloud properties contribute to bias Clouds and the Earth's Radiant Energy System (CERES) and ERA‐Interim surface irradiances, while ERA5 performs betterIn spring ERA‐Interim and CERES SW and LW biases compensate allowing estimates of total surface radiation to agree with surface observationsDifferences up to 0.1 in gridded surface albedo remain between the datasets and affect shortwave and total surface radiation budgets Surface albedo, temperature, and cloud properties contribute to bias Clouds and the Earth's Radiant Energy System (CERES) and ERA‐Interim surface irradiances, while ERA5 performs better In spring ERA‐Interim and CERES SW and LW biases compensate allowing estimates of total surface radiation to agree with surface observations Differences up to 0.1 in gridded surface albedo remain between the datasets and affect shortwave and total surface radiation budgets