Presentation Information
[ACG68-P10]NIPR's Short-term Sea Ice Forecast Service for the Whole Arctic Ocean Based on Ice-Ocean Coupled Simulations
*Yoshihiro Niwa1, Motomu Oyama2, Takeshi Sugimura1, Hironori Yabuki1 (1.National Institute of Polar Research, 2.Graduate School of Engineering, Kogakuin University)
Keywords:
Sea Ice,Arctic Ocean,Short-term Forecast,Arctic Sea Route
With the rapid melting of sea ice in the Arctic Ocean due to global warming, the demand for accurate sea ice forecasts has been increasing. In particular, short-term sea ice forecasts, which predict sea ice distribution a few to 10 days in advance, are crucial for ship navigation, as they help identify safe and efficient routes that avoid ice-covered areas. Since the summer of 2024, the Sea Ice Information Center of National Institute of Polar Research (NIPR) has been providing a short-term sea ice forecast service covering the whole Arctic Ocean using an ice-ocean coupled numerical model.
The numerical model used has a horizontal grid resolution of approximately 5 km. It is initialized using the sea ice data, the ocean temperature and salinity data from the analysis products of RIOPS managed by Environment and Climate Change Canada. Then, the 10-day forecasts are conducted daily using the atmospheric forcing data from the weather forecast products of ECMWF. The resulting daily sea ice forecast maps can be viewed on the NIPR’s website “ADS VENUS for Mirai” (https://ads.nipr.ac.jp/venus.mirai/#/mirai) (Fig. 1). These maps were also provided to the R/V Mirai of JAMSTEC to support research operations during its summer Arctic research cruise.
In summer season, ice edge errors are estimated to range from 10 to 20km (20 to 40km) for the 5 days (10 days) ahead forecasts. Notably, the 10-day forecast errors can sometimes become exceptionally large, as inaccuracies in the atmospheric forcing data inevitably increase with forecast lead time. This result suggests that, in addition to sea ice forecast information, real-time estimation of forecast errors is necessary to better support the increasing maritime traffic in the Arctic Ocean. This information is particularly crucial for trans Arctic commercial vessels, which typically need approximately 10 days to complete their passage across the Arctic Ocean.
The numerical model used has a horizontal grid resolution of approximately 5 km. It is initialized using the sea ice data, the ocean temperature and salinity data from the analysis products of RIOPS managed by Environment and Climate Change Canada. Then, the 10-day forecasts are conducted daily using the atmospheric forcing data from the weather forecast products of ECMWF. The resulting daily sea ice forecast maps can be viewed on the NIPR’s website “ADS VENUS for Mirai” (https://ads.nipr.ac.jp/venus.mirai/#/mirai) (Fig. 1). These maps were also provided to the R/V Mirai of JAMSTEC to support research operations during its summer Arctic research cruise.
In summer season, ice edge errors are estimated to range from 10 to 20km (20 to 40km) for the 5 days (10 days) ahead forecasts. Notably, the 10-day forecast errors can sometimes become exceptionally large, as inaccuracies in the atmospheric forcing data inevitably increase with forecast lead time. This result suggests that, in addition to sea ice forecast information, real-time estimation of forecast errors is necessary to better support the increasing maritime traffic in the Arctic Ocean. This information is particularly crucial for trans Arctic commercial vessels, which typically need approximately 10 days to complete their passage across the Arctic Ocean.