Next Article in Journal
Research on Risk Assessment and Prevention–Control Measures for Immersed Tunnel Construction in 100 m-Deep Water Environments
Previous Article in Journal
Development of an Initial Burial Rate Estimation Simulator for Bottom-Contact Mines and a Reinforcement Learning-Based Mine-Laying Route Optimization Method
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JMSE
Volume 14
Issue 1
10.3390/jmse14010052
jmse-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Past and Future Changes in Sea Ice in the Sea of Okhotsk: Analysis Using the Future Ocean Regional Projection Dataset

by
Daichi Narita
and
Shinsuke Iwasaki
*
Muroran Institute of Technology, 27-1 Mizumoto-cho, Muroran-shi 050-8585, Hokkaido, Japan
*
Author to whom correspondence should be addressed.
J. Mar. Sci. Eng. 2026, 14(1), 52; https://doi.org/10.3390/jmse14010052
Submission received: 5 December 2025 / Revised: 24 December 2025 / Accepted: 25 December 2025 / Published: 26 December 2025
(This article belongs to the Section Ocean Engineering)
Browse Figures
Versions Notes

Abstract

Although subject to annual fluctuations, sea ice in the Sea of Okhotsk has decreased to a maximum extent at a rate of approximately 3.4% per decade since the 1970s. Thus far, few studies have focused on projections of sea ice in the Sea of Okhotsk. This study focused on sea ice in the Sea of Okhotsk and examined its past and future characteristics using a climate projection dataset termed the Future Ocean Regional Projection dataset. Historical sea ice areas have been reported to be larger than satellite observations, and some data contain biases of approximately double the actual value. Therefore, a simple bias correction was performed based on the ratio of historical to satellite observation sea-ice areas, and the bias was corrected. Furthermore, we performed future projections using two bias-corrected scenarios (RCP2.6 and RCP8.5). Results revealed that for the future analysis period of 2006–2100, sea ice loss would be approximately 12.3 (102 km2/year) under RCP2.6 and approximately 37.3 (102 km2/year) under RCP8.5, indicating that under both scenarios, there would be almost no sea ice in the southern Sea of Okhotsk between 2071 and 2100. The results of this study provide useful information for researchers to predict sea ice in related physical fields.

1. Introduction

Recent advances in marine research have revealed that sea ice provides significant benefits for life on Earth. For example, the albedo effect of sea ice suppresses heat absorption in high-latitude regions and is crucial in maintaining the overall energy balance of the planet. If sea ice diminished, the reduced reflectivity of sunlight would cause the ocean surface to absorb more heat, raising concerns about accelerated warming [1]. In addition, sea ice functions as a mediating layer that controls the exchange of heat, water vapor, and salt between the ocean and atmosphere. Its formation and disappearance are expected to significantly influence local climate variability and precipitation patterns. Therefore, sea ice and climate systems influence each other [2].
Furthermore, sea ice suppresses and attenuates wave propagation and functions as a “natural breakwater” in coastal areas [3]. The Sea of Okhotsk, which is adjacent to Japan, is the lowest-latitude seasonal sea ice area in the Northern Hemisphere. Sea ice formation begins in the northwestern region in early winter (November), reaches its maximum extent during late winter (February and March), and then retreats from April to June [4]. Although subject to annual fluctuations, sea ice in this region has been reported to decrease at a rate of approximately 3.4% per decade since the 1970s. Satellite observations from 2000 to 2020 have indicated that the volume of ice in the Sea of Okhotsk has decreased by 162 cubic kilometers (51.9%) during the 21st century [5].
This decline is projected to continue with progress in global warming [6]. In addition, the decline in sea ice is associated with increased wind speeds and storms [7,8]. For example, the loss of wave attenuation by sea ice increases the wave power (Pw) by approximately 12–15% per decade during the winter period from December to February of the following year [7]. This has led to concerns about increased wave energy, frequent high waves, reduced stability of coastal structures, accelerated coastal erosion, and a higher incidence of coastal disasters. From this perspective, understanding sea-ice dynamics is critical for coastal nations bordering the Sea of Okhotsk, including Japan.
One future projection dataset is FORP (Future Ocean Regional Projection dataset) [9]. FORP-NP10 is a North Pacific future projection dataset jointly developed by the Japan Agency for Marine-Earth Science and Technology and the Meteorological Research Institute of the Japan Meteorological Agency. FORP-NP10 has a relatively high spatial resolution of 10 km (2 km for FORP-JPN022) and is expected to be an optimal dataset for calculating past and future wave characteristics. However, to the best of our knowledge, no previous studies have evaluated it for this specific sea area. This study focused on sea ice in the Sea of Okhotsk and examined its past and future characteristics using the FORP dataset, a future climate projection dataset.
This study focused on sea ice in the Sea of Okhotsk and was conducted with the following objectives: First, we compared the historical data (1981–2005) of the FORP model using satellite data, estimated its bias, and approximated the satellite data by correcting the bias for each obtained model. Second, we aimed to predict the future trend of sea ice area (2006–2100) based on the RCP (Representative Concentration Pathways) scenarios using the corrected model data. The main goal of this study was to clarify the characteristics of FORP and to accurately demonstrate future changes in sea ice in the Sea of Okhotsk.

2. Materials and Methods

This study analyzed past and future sea ice area using the FORP-NP10 dataset, which covers the entire Sea of Okhotsk. FORP is a projected dataset created from high-resolution (horizontal resolution: 0.1° × 0.1°; approximately 10 km × 10 km) regional ocean model simulations using the MRI.COMv4 (Meteorological Research Institute Community Ocean Model version 4) model, with atmospheric data from multiple CMIP5 (Coupled Model Intercomparison Project Phase 5) climate models/scenarios as external forcing [10,11]. Details of the model in FORP-NP10 are shown in Supplementary Materials Table S1. Observational data such as satellite data has not been assimilated into the model.
CMIP5 is an international project that compares and evaluates coupled climate models from various countries under common experimental conditions. It includes past reanalysis experiments using observation-based external forcing (Historical; 1850–2005) and future projection experiments (2006–2100; extended to 2300 in some cases) based on different radiative forcing scenarios (RCP2.6, 4.5, 6.0, and 8.5) [12]. Furthermore, because CMIP5 comprises global climate models (GCMs), its resolution should be enhanced to capture ocean variability in detail. Therefore, MRI.COMv4 was employed for downscaling owing to its flexible modular structure and ability to accurately reproduce the unique ocean circulation systems surrounding Japan, such as the Kuroshio, Oyashio, and Tsushima Warm Currents [13]. The climate models used in the FORP include MIROC5 (Model for Interdisciplinary Research on Climate, version 5) [14], MRI-CGCM3 (Meteorological Research Institute Coupled General Circulation Model version 3) [15], and JRA55-do (Japanese 55-year Reanalysis based surface dataset for driving ocean–sea-ice models) [16] from FORP-NP10 version4, and GFDL-ESM2M (Geophysical Fluid Dynamics Laboratory Earth System Model version 2M) [17] and IPSL-CM5A-MR (Institut Pierre-Simon Laplace Coupled Model version 5A) [18] from FORP-NP10 version 4.5. Additionally, the scenarios used were historical (1970–2005) for past analysis and two future projections: RCP2.6 (2006–2100) and RCP8.5 (2006–2100). The data are available from 1 January 1970 to 31 December 2100, with a time resolution of (Daily and Monthly). The data coverage area is the North Pacific Ocean (northern limit: 63° N latitude, southern limit: 15° S latitude, western limit: 99° W longitude, eastern limit: 75° E longitude). In this study, monthly data were used to examine long-term sea ice variations (past and future).
For comparison, satellite sea ice concentration data provided by NOAA OISST at the Physical Science Laboratory used [19,20]. These data have a horizontal resolution of 0.25° × 0.25° (approximately 27.8 km × 27.8 km), cover the entire globe, and are available from 1 September 1981 to 1 June 2025.
First, we evaluated the reproducibility by comparing the sea ice distribution of each FORP-NP10 model (historical) with satellite sea ice concentration data. Because the historical period is 1970–2005 (1970–2018 for JRA55-do only) and the satellite data period is 1981–2025, the analysis period is set to 1981–2005. The analysis area is the Sea of Okhotsk, defined as 40–65° N, 135–160° E. Next, a simple bias correction was applied based on the ratio of the sea ice area between each model and the satellite data, and its effectiveness was evaluated. Finally, a bias correction was applied to each FORP-NP10 model (RCP2.6, RCP8.5) to examine future sea ice areas in the Okhotsk Sea. The future analysis period is 2006–2100. In this study, when examining changes in sea ice over time, the average sea ice area from November to June, when sea ice is observed, was used as the annual average sea ice area. In addition, the sea ice area was the total area of the grid points for ice concentration over the Sea of Okhotsk (i.e., not the sea ice extent). The significance of the linear trends was checked using the Mann–Kenndall test [21].

3. Results

3.1. Historical Model

First, we compared the historical data with the satellite data for each FORP model (GFDL-ESM2M, IPSL-CM5A-MR, MIROC5, JRA55-do, and MRI-CGCM3) (Figure 1, Table 1). Figure 1a,b shows that GFDL-ESM2M and MRI-CGCM3 overestimate sea ice area. Conversely, MIROC5 underestimates sea ice area. Quantitatively, satellite observations show approximately 3.8 × 105 km2, GFDL-ESM2M and MRI-CGCM3 show approximately 7.8 × 105 km2, and MIROC5 shows approximately 1.9 × 105 km2 (second column, Table 1). Conversely, JRA55-do and IPSL-CM5A-MR yield values relatively close to satellite data (second and third columns, Table 1). Examining the correlation coefficient for interannual variation, JRA55-do showed an exceptionally high value of 0.96 (fourth column, Table 1). By contrast, the other datasets exhibited correlation coefficients below 0.4, with GFDL-ESM2M and IPSL-CM5A-MR showing negative correlations. Regarding long-term trends, all models exhibited a decreasing trend similar to that of the satellite data, but the values differed (Figure 1c). JRA55-do and MRI-CGCM3 exceed 20.0 (10 km2/year), showing values relatively close to those of the satellite observations. By contrast, GFDL-ESM2M, IPSL-CM5A-MR, and MIROC5 are below 20.0 (102 km2/year) (especially MIROC5, which is below 50.0 (102 km2/year)). Regarding spatial distribution characteristics, all datasets reproduced the trend of abundant sea ice in the northwestern Sea of Okhotsk, decreasing southeastward (Supplementary Materials Figure S1). However, an examination of the differences reveals that GFDL-ESM2M and MRI-CGCM3 consistently overestimate sea ice across the entire domain, whereas MIROC5 underestimates it (Supplementary Materials Figure S2).

3.2. Bias Correction

These results show that the FORP sea ice data contain biases. Therefore, a simple bias correction was applied using the ratio of the sea ice area between each model and satellite data. The correction method is as follows: Focusing on the average sea ice area from November to June, when sea ice is present, correction coefficients were derived by dividing the annual average sea ice area (1981–2005) from satellite data by the annual average sea ice area (1981–2005) of each model.
“Correction coefficient for each model” =
“Annual average sea ice area from satellite data (1981–2005)”/“Annual average sea ice area for each model (1981–2005)”
Additionally, we implemented a method to derive the correction coefficients for each month separately from the annual average.
“Monthly correction coefficients for each model” =
“Monthly average sea ice area from satellite data (1981–2005)”/“Monthly average sea ice area for each model (1981–2005)”
A comparison of the two correction methods reveals that their approximation effects are consistent. Therefore, we omitted reporting experiments that used monthly corrections. For this study, we adopted a simple and clear approach of uniformly applying the correction coefficients derived by the above methods across the entire model period and spatial domain (see Supplementary Materials Table S2 for the monthly correction coefficients).
The correction coefficients obtained using this method are also presented in the fifth column of Table 1. These correction coefficients are uniformly applied to the entire period and spatial domain for the historical data of each model (1981–2005). As expected from the results shown in Figure 1a,b, models with larger differences from the satellite data also exhibited larger correction coefficients. The comparison results between the corrected models and satellite data are shown in Figure 2. This correction significantly reduces the bias in sea ice for each model (Figure 1a and Figure 2a). Furthermore, an examination of the climatological monthly averages of the sea ice area between the corrected models and satellite data demonstrates the effectiveness of the correction (Figure 1b and Figure 2b). As expected, the root mean square error (RMSE) including bias has also decreased significantly for almost models (final line of Table 1). However, for MIROC5, an underestimation bias persisted even after correction (Figure 2b, Supplementary Materials Table S3). These aspects are detailed in the Discussion section. The spatial distribution of the corrected difference shows that although the difference is smaller than that before correction, applying the correction coefficient uniformly across the domain results in regional ± differences (Supplementary Materials Figure S3). Because no significant improvement in the correlation coefficient for interannual variability or long-term trends was observed, these were omitted.

3.3. Future Forecast (Original and Corrected)

The following section describes future projections using FORP models. The FORP models used here include historical data as well as projection scenarios RCP2.6 and RCP8.5 (JRA55-do only includes historical data). Four FORP models (GFDL-ESM2M, IPSL-CM5A-MR, MIROC5, and MRI-CGCM3) analyzed both scenarios from 2006 to 2100. Briefly supplementing RCP2.6 and RCP8.5: RCP2.6 represents a scenario where the global average temperature increase by the end of the 21st century (2081–2100) is maintained below 2 °C compared with pre-industrial levels. RCP8.5 is a scenario where the global average temperature at the end of the 21st century (2081–2100) increases by 3.2–5.4 °C compared with pre-industrial levels, because of a significant failure to reduce greenhouse gas emissions. Although four RCP scenarios exist, FORP-NP10 uses two extreme examples: RCP2.6, where warming is sufficiently mitigated, and RCP8.5, where warming is the most advanced.
Figure 3 shows the results of future sea ice projections (2006–2100) based on the FORP model. Satellite and JRA55do data up to 2018 are displayed. Similar to the previous results, GFDL-ESM2M and MRI-CGCM3 showed relatively large values, whereas IPSL-CM5A-MR and MIROC5 showed smaller values (Figure 3a,b). Both scenarios exhibited a decreasing trend in sea ice area, but the magnitude of this trend differs significantly. For example, between the two scenarios, the difference is 59.9 (102 km2/year) for GFDL-ESM2M, 21.8 (102 km2/year) for IPSL-CM5A-MR, and 45.5 (102 km2/year) for MRI-CGCM3 (Supplementary Materials Figure S4a). Furthermore, before 2040, the annual sea ice area under RCP8.5 frequently exceeds that under RCP2.6 across all models. However, significant differences begin to emerge between the two scenarios after 2040 (Supplementary Materials Figure S4b). The difference between RCP2.6 and RCP8.5 is noticeable during the latter 30 years (2071–2100) when the most pronounced differences are observed (Figure 3c,d). Even GFDL-ESM2M, which retains the most ice during the peak months of February and March, failed to reach 0.5 (106 km2) under RCP8.5. Supplementary Materials Figures S5 and S6 show the sea ice changes for each model under both scenarios. GFDL-ESM2M and MRI-CGCM3, which tend to overestimate compared with the observations, exhibited reductions over a relatively large area. Conversely, MIROC5, which tends to underestimate, exhibits a decreasing trend over a comparatively smaller area. As expected, the reduction was more pronounced under RCP8.5 than under RCP2.6.
The correction coefficients derived from the historical analysis were incorporated into each FORP model to investigate future sea ice changes (Figure 4). Because of the correction effect, models such as GFDL-ESM2M and MRI-CGCM3 that previously overestimated sea ice exhibited significant reductions under both scenarios. Furthermore, these models are better aligned with satellite observations and JRA55-do values (Figure 4a,b). Furthermore, under RCP2.6, where a considerable amount of sea ice remained, the correction resulted in a substantial change in sea ice compared with the case under RCP8.5 (Figure 4c,d). However, MIROC5, which tended to underestimate sea ice compared with the observations, exhibited an extremely small effect from the correction (details are discussed in the Discussion section).

4. Discussion

For GFDL-ESM2M and MRI-CGCM3, which showed overestimation compared with satellite observations, a significant effect of bias correction was demonstrated. For IPSL-CM5A-MR, which underestimated, the correction was small owing to its inherently low bias relative to satellite observations. Although MIROC5 was significantly underestimated, the effect of bias correction was also small. This is believed to be due to the large number of areas in the original data where sea ice is absent (i.e., where sea ice concentration is zero). Therefore, when considering the most reliable projections of future sea ice changes, predictions should be made using the three corrected models (GFDL-ESM2M, MRI-CGCM3, and IPSL-CM5A-MR), excluding MIROC5. Figure 5 and Figure S7 derive and compare the average values of the three models for each scenario. RCP2.6 represents the scenario with the most restrained future warming, whereas RCP8.5 represents the scenario with the most advanced warming. Based on these two extreme scenarios, we examined future sea ice areas. No significant differences were observed under either scenario until 2040, after which differences began to emerge (Figure 5a). The overall trend values are 12.3 (102 km2/year) under RCP2.6 and 37.3 (102 km2/year) under RCP8.5. Considering monthly climatological means, sea ice area may be limited to approximately 0.5 × 106 km2 under the RCP2.6 scenario for February and March from 2071 to 2100, and the area is projected to fall below 0.2 × 106 km2 under the more warming RCP8.5 scenario (Figure 5b). Furthermore, spatial distribution revealed a significant decrease in sea ice concentration in the northwestern Sea of Okhotsk and Shelikhov Bay under both scenarios (Supplementary Materials Figure S7a,b). Sea ice in the southern Sea of Okhotsk (i.e., northern Hokkaido, Japan) is believed to be mainly generated in the northwestern Sea of Okhotsk (around the mouth of the Amur River) [22,23]; however, our study results show that it may almost completely disappear by 2100.
An option for future projections might be to use only IPSL-CM5A-MR within FORP, because it is the most similar to satellite observations. However, IPSL-CM5A-MR does not necessarily achieve a high reproduction rate for interannual variability (fourth column of Table 1). Furthermore, from the perspective of evaluating uncertainty by increasing the number of model members, using multiple model results whenever possible is desirable.

5. Conclusions

This study focused on sea ice in the Sea of Okhotsk and examined its past and future characteristics using the FORP climate prediction dataset. Comparisons between past sea ice area and satellite observations revealed that each dataset contained some bias, with some datasets exhibiting biases approximately twice the actual values. Therefore, a simple bias correction was performed using the ratio of past sea ice area to satellite observations, and future projections were then conducted using the bias-corrected scenarios (RCP2.6 and RCP8.5). The results indicate that during the future analysis period (2006–2100), sea ice is projected to decrease by approximately 12.3 (102 km2/year) under the RCP2.6 scenario and approximately 37.3 (102 km2/year) under the RCP8.5 scenario. Under both scenarios, sea ice in the southern Sea of Okhotsk is predicted to nearly disappear between 2071 and 2100. As sea ice in this region affects not only physical processes, such as waves, but also ecosystems, such as phytoplankton, these findings provide valuable insights for researchers in related fields.
Although the Sea of Okhotsk is considered a single sea, its sea ice characteristics vary significantly across different regions [5,24]. Furthermore, applying the bias correction obtained in this study may have led to overfitting in the results, potentially resulting in inconsistencies with the physical processes of each model. Therefore, although a uniform bias correction throughout the entire period can avoid temporal inconsistencies, it is considered inappropriate for analyses involving parameters such as heat balance.
We intend to clarify the characteristics of sea surface wind (i.e., CMIP5) in this region during future wave predictions. In this context, we will carefully consider whether similar corrections should be applied if biases exist in the sea surface wind or whether corrections should instead be applied to wave height and other parameters after wave calculations.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jmse14010052/s1, Figure S1: Spatial distribution of average sea ice concentration (color and contour lines) during winter (November to June) for each model (1981–2005); Figure S2: Spatial distribution of differences in sea ice concentration between each model and satellite data during winter (November to June of the following year) for each model (1981–2005); Figure S3: Similar to Figure S2, but shows the results after correction; Figure S4: (a) Trends in the interannual variability of the winter (November to June of the following year) average sea ice concentration for 2006–2100: (red) RCP8.5, (blue) RCP2.6. (b) Interannual changes in winter sea ice concentration averaged across four models for (blue line) RCP2.6 and (red line) RCP8.5, and (black line; right axis) their difference; Figure S5: Spatial distribution of sea ice concentration differences (color) among models under RCP2.6[(2071–2100)–(2006–2040)]. Contours indicate [(difference obtained above)/(average sea ice concentration in 2006–2040)] × 100 (%); Figure S6: Similar to Figure S5, but results for RCP8.5; Figure S7: Spatial distribution of sea ice concentration averaged from three FORP models for 2071–2100.RCP2.6 (a), RCP8.5 (b); Table S1: Configurations of the regional ocean model system (FORP-NP10). WOA stands for World Ocean Atlas; Table S2: Correction coefficient ratio for sea ice area calculated monthly from satellite data and past data for each model; Table S3: Sea ice area for corrected models and satellite.

Author Contributions

Conceptualization, S.I.; methodology, S.I.; software, D.N.; validation, D.N.; formal analysis, D.N.; investigation, D.N.; resources, D.N.; data curation, D.N.; writing—original draft preparation, D.N. and S.I.; writing—review and editing, S.I.; visualization, D.N.; supervision, S.I.; project administration, S.I.; funding acquisition, S.I. All authors have read and agreed to the published version of the manuscript.

Funding

This work was conducted as part of Theme 4 of the Advanced Studies of Climate Change Projection (SENTAN Program), Grant Number JPMXD0722678534, supported by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan. This study utilized the dataset ‘Future Ocean Regional Projection’ (FORP), which was produced by the Japan Agency for Marine-Science and Technology (JAMSTEC) and the Meteorological Research Institute, the Japan Meteorological Agency under the ‘SI-CAT’ project (Grant Number: JPMXD0715667163) and the ‘TOOUGOU’ project (Grant Number: JPMXD0717935561) of the Ministry of Education, Culture, Sports, Science and Technology, Japan. We thank the reviewers for their constructive and fruitful comments.

Data Availability Statement

The data used in this study are listed as follows: FORP (https://search.diasjp.net/en/dataset/FORP_NP10_version4) (accessed on 4 May 2025) and NOAA for OISST version 2.1 (https://psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html) (accessed on 4 May 2025).

Conflicts of Interest

The authors declare no competing interests.

Correction Statement

This article has been republished with a minor correction to resolve spelling and grammatical errors. This change does not affect the scientific content of the article.

References

  1. Moon, W.; Wettlaufer, J.S. On the nature of the sea ice albedo feedback in simple models. J. Geophys. Res. Oceans 2014, 119, 5555–5562. [Google Scholar] [CrossRef] [PubMed]
  2. Rysgaard, S.; Bendtsen, J.; Delille, B.; Dieckmann, G.S.; Glud, R.N.; Kennedy, H.; Mortensen, J.; Papdimitriou, S.; Thomas, D.N.; Tison, J.L. Sea-ice contribution to the air–sea CO2 exchange in the Arctic and Southern Oceans. Tellus B 2011, 63B, 823–830. [Google Scholar] [CrossRef]
  3. Iwasaki, S. Role of sea ice on winter wave power and its interannual variability in the Sea of Okhotsk: Natural breakwater modified by surface wind changes. Prog. Oceanogr. 2022, 210, 102944. [Google Scholar] [CrossRef]
  4. Cavalieri, D.J.; Parkinson, C.L. On the relationship between atmospheric circulation and the fluctuations in the sea ice extents of the Bering and Okhotsk seas. J. Geophys. Res. Oceans 1987, 92, 141–7162. [Google Scholar] [CrossRef]
  5. Pishchalnik, V.M.; Truskov, P.A.; Romanyuk, V.A.; Minervin, I.G. Interannual variability of ice volume in the sea of Okhotsk during the maximum development of ice extent for 2000–2020. Cold Reg. Sci. Technol. 2021, 189, 103326. [Google Scholar] [CrossRef]
  6. Ministry of Education, Culture, Sports, Science and Technology (MEXT); Japan Meteorological Agency (JMA). Climate Change in Japan 2020: Latest Scientific Findings and Future Outlook; MEXT: Tokyo, Japan; JMA: Tokyo, Japan, 2020; Available online: https://www.data.jma.go.jp/cpdinfo/ccj/2020/pdf/cc2020_gaiyo_en.pdf (accessed on 4 December 2025).
  7. Iwasaki, S. Increase in the wave power caused by decreasing sea ice over the Sea of Okhotsk in winter. Sci. Rep. 2023, 13, 2539. [Google Scholar] [CrossRef] [PubMed]
  8. Myslenkov, S.A.; Kruglova, E.E.; Medvedeva, A.Y.; Silvestrova, K.P.; Arkhipkin, V.S.; Akpinar, A.; Dobrolyubov, S.A. Number of Storms in Several Russian Seas: Trends and Connection to Large-Scale Atmospheric Indices. Russ. J. Earth Sci. 2023, 23, 1–31. [Google Scholar] [CrossRef]
  9. Nishikawa, S.; Wakamatsu, T.; Ishizaki, H.; Sakamoto, K.; Tanaka, Y.; Tsujino, H.; Yamanaka, G.; Kamachi, M.; Ishikawa, Y. Development of high-resolution future ocean regional projection datasets for coastal applications in Japan. Prog. Earth Planet. Sci. 2021, 8, 7. [Google Scholar] [CrossRef]
  10. Japan Meteorological Agency (JMA); Meteorological Research Institute (MRI). Future Ocean Regional Projection (FORP) Dataset, North Pacific 10 km Version 4 (FORP_NP10_v4); Data Integration and Analysis System (DIAS): Tokyo, Japan, 2024; Available online: https://search.diasjp.net/ja/dataset/FORP_NP10_version4 (accessed on 4 December 2025).
  11. Japan Meteorological Agency (JMA); Meteorological Research Institute (MRI). Future Ocean Regional Projection (FORP) Dataset, North Pacific 10 km Version 4.5 (FORP_NP10_v4_5); Data Integration and Analysis System (DIAS): Tokyo, Japan, 2024; Available online: https://search.diasjp.net/ja/dataset/FORP_NP10_version4_5 (accessed on 4 December 2025).
  12. Taylor, K.E.; Stouffer, R.J.; Meehl, G.A. An overview of cmip5 and the experiment design. Bull. Am. Meteorol. Soc. 2012, 93, 485–498. [Google Scholar] [CrossRef]
  13. Tsujino, H.; Urakawa, S.; Nakano, H.; Toyoda, T.; Fujii, Y.; Aita, M.N.; Nakaoka, S.; Nishikawa, S.; Sakamoto, K.; Yamanaka, G. Reference Manual for the Meteorological Research Institute Community Ocean Model Version 4 (MRI.COMv4); Technical Reports of the Meteorological Research Institute; Meteorological Research Institute: Tsukuba, Japan, 2017; Volume 80, pp. 1–69. [Google Scholar] [CrossRef]
  14. Watanabe, M.; Suzuki, T.; O’Ishi, R.; Komuro, Y.; Watanabe, S.; Emori, S.; Takemura, T.; Chikira, M.; Ogura, T.; Sekiguchi, M.; et al. Improved climate simulation by MIROC5: Mean states, variability, and climate sensitivity. J. Clim. 2010, 23, 6312–6335. [Google Scholar] [CrossRef]
  15. Yukimoto, S.; Adachi, Y.; Hosaka, M.; Sakami, T.; Yoshimura, H.; Hirabara, M.; Tanaka, T.Y.; Shindo, E.; Tsujino, H.; Deushi, M.; et al. A new global climate model of the Meteorological Research Institute: MRI-CGCM3—Model description and basic performance. J. Meteor. Soc. Jpn. Ser. 2012, 90, 23–64. [Google Scholar] [CrossRef]
  16. Tsujino, H.; Urakawa, S.; Nakano, H.; Small, R.J.; Kim, W.M.; Yeager, S.G.; Danabasoglu, G.; Suzuki, T.; Bamber, J.L.; Bentsen, M.; et al. JRA-55 based surface dataset for driving ocean–sea-ice models (JRA55-do). Ocean Model. 2018, 130, 79–139. [Google Scholar] [CrossRef]
  17. Dunne, J.P.; John, J.G.; Adcroft, A.; Griffies, S.M.; Hallberg, R.W.; Shevliakova, E.; Stouffer, R.J.; Cooke, W.; Dunne, K.A.; Harrison, M.J.; et al. GFDL’s ESM2 global coupled carbon–climate Earth System Models. Part I: Physical formulation and baseline simulation characteristics. J. Clim. 2012, 25, 6646–6665. [Google Scholar] [CrossRef]
  18. Dufresne, J.L.; Foujols, M.A.; Denvil, S.; Caubel, A.; Marti, O.; Aumont, O.; Balkanski, Y.; Bekki, S.; Bellenger, H.; Benshila, R.; et al. Climate change projections using the IPSL-CM5 Earth System Model: From CMIP3 to CMIP5. Clim. Dyn. 2013, 40, 2123–2165. [Google Scholar] [CrossRef]
  19. Huang, B.; Liu, C.; Banzon, V.; Freeman, E.; Graham, G.; Hankins, B.; Smith, T.; Zhang, H.M. Improvements of the Daily Optimum Interpolation Sea Surface Temperature (DOISST) Version 2.1. J. Clim. 2021, 34, 2923–2939. [Google Scholar] [CrossRef]
  20. Reynolds, R.W.; Thomas, M.S.; Chunying, L.; Dudley, B.C.; Kenneth, S.C.; Michael, G.S. Daily high-resolution-blended analyses for sea surface temperature. J. Clim. 2007, 20, 5473–5496. [Google Scholar] [CrossRef]
  21. Kendall, M.G. Rank Correlation Methods; Griffin: London, UK, 1948. [Google Scholar]
  22. Martin, S.; Drucker, R.; Yamashita, K. The production of ice and dense shelf water in the Okhotsk Sea polynyas. J. Geophys. Res. 1998, 103, 27771–27782. [Google Scholar] [CrossRef]
  23. Kimura, N.; Wakatsuchi, M. Relationship between sea-ice motion and geostrophic wind in the Northern Hemisphere. Geophys. Res. Lett. 2000, 27, 3735–3738. [Google Scholar] [CrossRef]
  24. Minervin, I.G.; Romanyuk, V.A.; Pishchal’nik, V.M.; Truskov, P.A.; Pokrashenko, S.A. Zoning the ice cover of the Sea of Okhotsk and the Sea of Japan. Her. Russ. Acad. Sci. 2015, 85, 132–139. [Google Scholar] [CrossRef]
Jmse 14 00052 g001
Figure 1. Characteristics of sea ice in the Sea of Okhotsk based on the Historical (1981–2005) model. (a) Interannual changes in sea ice area (November to June). (b) Climatological monthly mean sea ice area. (c) Trend values for interannual changes in sea ice area (November to June). (a) The dashed lines indicate the linear regression lines for each data. (a) The trends for the MIROC model are statistically significant at the 95% confidence level.
Figure 1. Characteristics of sea ice in the Sea of Okhotsk based on the Historical (1981–2005) model. (a) Interannual changes in sea ice area (November to June). (b) Climatological monthly mean sea ice area. (c) Trend values for interannual changes in sea ice area (November to June). (a) The dashed lines indicate the linear regression lines for each data. (a) The trends for the MIROC model are statistically significant at the 95% confidence level.
Jmse 14 00052 g001
Jmse 14 00052 g002
Figure 2. Same as Figure 1a,b but after correction. (a) Interannual changes and (b) Climatological monthly mean.
Figure 2. Same as Figure 1a,b but after correction. (a) Interannual changes and (b) Climatological monthly mean.
Jmse 14 00052 g002
Jmse 14 00052 g003
Figure 3. Characteristics of sea ice in the Sea of Okhotsk under future scenarios (2006–2100). (a,b) Interannual changes in sea ice area (November to June) under (a) RCP2.6 and (b) RCP8.5 (dashed lines are regression lines), and (c,d) climatological monthly mean sea ice area under (c) RCP2.6 and (d) RCP8.5 for 2071–2100. (a,b) The trends for all models (RCP2.6 and RCP8.5) except JRA55-do and satellite observations were statistically significant at the 95% confidence level.
Figure 3. Characteristics of sea ice in the Sea of Okhotsk under future scenarios (2006–2100). (a,b) Interannual changes in sea ice area (November to June) under (a) RCP2.6 and (b) RCP8.5 (dashed lines are regression lines), and (c,d) climatological monthly mean sea ice area under (c) RCP2.6 and (d) RCP8.5 for 2071–2100. (a,b) The trends for all models (RCP2.6 and RCP8.5) except JRA55-do and satellite observations were statistically significant at the 95% confidence level.
Jmse 14 00052 g003
Jmse 14 00052 g004
Figure 4. Same as Figure 3, but after correction. The semi-transparent lines in (a,b) and the dashed lines in (c,d) indicate the results before correction (i.e., the same as Figure 3).
Figure 4. Same as Figure 3, but after correction. The semi-transparent lines in (a,b) and the dashed lines in (c,d) indicate the results before correction (i.e., the same as Figure 3).
Jmse 14 00052 g004
Jmse 14 00052 g005
Figure 5. Future changes in sea ice area in the Sea of Okhotsk considered from three models. (a) Interannual sea ice change over time (November to June). The dotted lines are linear regression curves, and the values on the figure indicate their slopes. (b) Climatological monthly mean sea ice area from 2071 to 2100. (a,b) The shaded areas indicate the standard deviation for three models of each scenario. (a) The trends for RCP2.6 and RCP8.5 3 (blue and red lines) were statistically significant at the 95% confidence level.
Figure 5. Future changes in sea ice area in the Sea of Okhotsk considered from three models. (a) Interannual sea ice change over time (November to June). The dotted lines are linear regression curves, and the values on the figure indicate their slopes. (b) Climatological monthly mean sea ice area from 2071 to 2100. (a,b) The shaded areas indicate the standard deviation for three models of each scenario. (a) The trends for RCP2.6 and RCP8.5 3 (blue and red lines) were statistically significant at the 95% confidence level.
Jmse 14 00052 g005
Table 1. Annual average sea ice area from 1981 to 2005 (second column). Root mean square error (RMSE) of sea ice for each model and satellite data (third column). Correlation coefficients between satellite data and FORP models for interannual variation in sea ice area (fourth column). Correction coefficients ratio of sea ice area calculated from satellite data and historical data for each model (fifth column). RMSE analysis results for sea ice using each corrected model and satellite data (final column). The RMSE in this table includes the bias.
Table 1. Annual average sea ice area from 1981 to 2005 (second column). Root mean square error (RMSE) of sea ice for each model and satellite data (third column). Correlation coefficients between satellite data and FORP models for interannual variation in sea ice area (fourth column). Correction coefficients ratio of sea ice area calculated from satellite data and historical data for each model (fifth column). RMSE analysis results for sea ice using each corrected model and satellite data (final column). The RMSE in this table includes the bias.
Model NameSea Ice Area
[105 (km2)]		RMSE
(Original)		Correlation
Coefficients		Correction
Coefficients		RMSE
(Corrected)
OBS3.8
GFDL-ESM2M7.80.42−0.290.490.10
IPSL-CM5A-MR3.40.11−0.101.120.11
MIROC51.90.210.172.040.16
JRA55-do5.10.130.960.750.02
MRI-CGCM37.80.410.350.490.07
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Narita, D.; Iwasaki, S. Past and Future Changes in Sea Ice in the Sea of Okhotsk: Analysis Using the Future Ocean Regional Projection Dataset. J. Mar. Sci. Eng. 2026, 14, 52. https://doi.org/10.3390/jmse14010052

AMA Style

Narita D, Iwasaki S. Past and Future Changes in Sea Ice in the Sea of Okhotsk: Analysis Using the Future Ocean Regional Projection Dataset. Journal of Marine Science and Engineering. 2026; 14(1):52. https://doi.org/10.3390/jmse14010052

Chicago/Turabian Style

Narita, Daichi, and Shinsuke Iwasaki. 2026. "Past and Future Changes in Sea Ice in the Sea of Okhotsk: Analysis Using the Future Ocean Regional Projection Dataset" Journal of Marine Science and Engineering 14, no. 1: 52. https://doi.org/10.3390/jmse14010052

APA Style

Narita, D., & Iwasaki, S. (2026). Past and Future Changes in Sea Ice in the Sea of Okhotsk: Analysis Using the Future Ocean Regional Projection Dataset. Journal of Marine Science and Engineering, 14(1), 52. https://doi.org/10.3390/jmse14010052

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop