Abstract A set of dedicated radiative kernels are generated to quantify the individual contributions of sea ice and clouds on the interannual variation of absorbed solar radiation (ASR) observed over the Arctic ocean. From spring to early summer, changes in ASR are associated more with sea ice than with the clouds. Conversely, the cloud contribution explains more of the ASR anomaly after June. Overall, variations in cloud cover and sea ice extent can explain 50%–93% of the ASR variability in each sunlit month. Both the sea‐ice‐ and cloud‐associated ASR variations are positively correlated with the observed ASR anomaly, but their trends over the 20‐year observational period are opposite. The positive ASR trend revealed by the kernel decomposition, primarily driven by the observed decline in sea ice extent, is dampened by clouds, particularly in June. Annually, clouds counteract 55% of the ASR trend induced by Arctic sea ice loss.

Read original article