Quality Issues

• Non-assimilation of MGDSST

  It has been identified that MGDSST was not used for data assimilation from late November to late December 2017. Although the impact is limited, please exercise caution when using the data.

• Summer cold bias near river mouths in Japan

  A cold bias during summer has been identified in several river mouth areas in Japan. Although the bias is confined to very nearshore regions, comparisons with observations in southwestern Hokkaido show that temperatures can be underestimated by up to about 5°C. Similar biases may also occur in other river mouth regions.

• Abnormally low temperatures in summer near the mouth of Changjiang River

  A similar low-temperature phenomenon has been observed during summer near the mouth of Changjiang River. Although this is also a localized feature, users are advised to take caution when using the data.

• Issue in sea ice assimilation

  An issue in the sea ice assimilation has also been confirmed. The correction to reduce the model sea ice concentration did not work properly, resulting in overestimation of sea ice extent. After fixing this issue, we conducted a one-year test experiment and evaluated sea ice extent in the Okhotsk. The difference in sea ice extent between FORA-JPN60 and the test experiment was relatively small (within 5 %) during the sea ice expansion period from December to February, but it grew larger during the sea ice reduction period starting in March, reaching its maximum in April (approximately 15 %). Although this bias exists in the sea ice field, we confirmed that interannual variability and long-term trends are generally reproduced.