Atmospheric aerosols influence the radiation budget, cloud amount, cloud properties, and surface albedos of sea ice and snow over the Arctic. In spite of their climatic importance, Arctic aerosol contains large uncertainties due to limited observations. This study evaluates the Arctic aerosol variability in three reanalyses, JRAero, CAMSRA, and MERRA2, in terms of the aerosol optical depth (AOD), and its relationship to the atmospheric disturbances on synoptic timescales. The AOD becomes highest in July–August over most of the Arctic regions, except for the North Atlantic and Greenland, where monthly variability is rather small. The three reanalyses show a general consistency in the horizontal distribution and temporal variability of the total AOD in summer. In contrast, the contributions of individual aerosol species to the total AOD are quite different among the reanalyses. Compared with observations, the AOD variability is represented well in all reanalyses in summer with high correlation coefficients, albeit exhibiting errors as large as the average AOD. The composite analysis shows that large aerosol emissions in Northern Eurasia and Alaska and transport by a typical atmospheric circulation pattern contribute to the high aerosol loading events in each area of the Arctic. Meanwhile, the empirical orthogonal function analysis depicts that the first- and second-largest AOD variabilities on the synoptic timescales appear over Northern Eurasia. Our results indicate that these summertime AOD variabilities mainly result from aerosol transportation and deposition due to the atmospheric disturbances on synoptic scales, suggesting an essential role played by Arctic cyclones.

Mean residence time of the seawater in the shelf region, τ, has been studied in several closed bays and inland seas around Japan. From previous estimates of 0.69 months in the Ise Bay and 6.4-14.7 months in the Seto Inland Sea, τ is expected to vary depending on area, but there is no research that reveals the whole picture of the spatiotemporal variation of τ around Japan. As the first step, we estimate τ for the entire coastal regions using the 2-km resolution Japanese coastal model “MRI.COM-JPN” that we developed for a JMA operational system. The model reproduces well tides, river inflows and many coastal currents around Japan, which are necessary to simulate the basic physical processes of coastal-offshore water exchange. In order to estimate τ, an experiment was conducted to run an “age tracer” that takes an age while existing in the shelf region and then resets the age to zero offshore. The tracer value can be regarded as the mean residence time τ of the seawater flowing into the shelf region. The model was driven for 9 years, and the results from the last 7 years were used for the analysis. Results showed that in many coastal areas, τ ranged from 20 to 100 days. In eight closed areas such as the Tokyo Bay and the Seto Inland Sea, τ reached 100-450 days. In addition, τ was as long as 100-200 days in the downstream of the two coastal currents originating from continental shelves, the Tsushima Warm Current and the East Sakhalin Current. On the other hand, τ on the Pacific side, including the southern coast of Japan where the Kuroshio Current flows offshore, was as short as several to 20 days. At presentation, we will also discuss results of a particle tracking experiment.