Dynamics of acidification of the Arctic Ocean through 1993–2021 and predictions of further tendencies of this process until the end of 2100 were assessed making use of both the GLODAPv.2021 and the Global Ocean Biogeochremistry Hindcast (GOBH) reanalysis data on pH. The projections of pH were performed by CMIP6 models for four scenarios of rates of socio-economic and agricultural development and emissions of greenhouse gases: SSP1–2.6, SSP2–4.5, SSP3–7.0 and SSP5–8.5.The tendencies of pH decline over the last 27 years (1993–2019) as determined from the GLODAP in situ and the reanalysis data over 1993–2021 proved to be, respectively –0.9% (from 8.18–8.11) and –0.7% (from 8.10–8.05). Thus, the annual acidification rate as assessed from both data sources proved to be –0.03%.Through the percentile method-based comparison of consistency of historical observation data on pH with GBH model hindcast four best models were identified: MPI-ESM1–2-LR, NorESM2-MM, NorESM2-LM, and CMCC-ESM2. The projection results strongly indicate that the Arctic Ocean acidification will continue till the end of this century. The highest rates of pH decrease (–4.9% and –6.2%) were forecasted, respectively, for scenarios SSP3–7.0 and SSP5–8.5 that implied the global mean temperature increases by 3.6 °C and 4.4 °C, respectively. A comparison of the results obtained with the previously made assessments is indicative that by the end of the current century the rate of acidification (i. e. pH decrease) in the Arctic should be expected to be higher than that averaged over the World Oceans: the difference for each of the SSP scenarios proved to be –0.1.

The Bering Sea and its surrounding waters are commercially and ecologically important ecosystems. Knowledge of phytoplankton phenology is crucial for understanding ecosystem dynamics. However, estimates of phenological parameters of spring phytoplankton bloom are sparse for this region. We used the Moderate Resolution Imaging Spectroradiometer (MODIS) daily data from 2003–2024 to assess the climatology of phenological parameters. A combination of data regriding, spatial interpolation, and temporal smoothing was applied. Three methods of spatial interpolation for missing data acquisition are compared: iterative first-order neighbor, inverse distance weighted interpolation, and data interpolating empirical orthogonal functions (DINEOF). We suggest that the first outcompetes the other two methods when compared to initial data. Date of the bloom initiation, bloom peak, chlorophyll-a maximum, and duration of the bloom before its peak are evaluated. The spatial distribution of mentioned phenological parameters is presented and discussed. We show that bloom starts early in Bristol Bay, in the narrow band along the eastern shelf, along the Kamchatka Peninsula, and south of the Aleutians and Alaska Peninsula. In the deep Bering Sea, bloom starts surprisingly later considering the latitude of the region. The main reason for this may be the wind mixing during the spring. The first phase of the bloom is generally longer in the deep southern areas (up to 60 days) and shorter in the northern shelf areas (less than 2 weeks in some cases).

New regional empirical algorithms were developed to obtain maximum specific photosynthetic rates of phytoplankton (\(P_{m}^{B}\)) in the surface layer of the Atlantic Ocean. These algorithms were based on the dependence of \(P_{m}^{B}\) on seawater temperature. Sea Surface Temperature remote sensing data and the PANGAEA global database of photosynthesis–irradiance parameters were used to test the algorithm. In addition, the variability in \(P_{m}^{B}\), both spatially (from 60° S to 85° N) and seasonally, (2002–2013) was estimated. The highest \(P_{m}^{B}\) was obtained in December in areas of deep convection and the interaction between the Labrador Current and the Gulf Stream, while minimum values were observed in the northern and equatorial–tropical parts of the ocean during the time intervals between the phytoplankton blooms (March to September–October). In addition, existing \(P_{m}^{B}\) and \(P_{{{\text{opt}}}}^{B}\) algorithms used in primary production models, as well as the \(P_{m}^{B}\) algorithm developed using temperature and chlorophyll a data from AMT-29, which were then tested using the PANGAEA dataset. The results show that the new \(P_{m}^{B}\) algorithm developed using seawater temperature data with regionally adjusted empirical coefficients correlated best with the in situ data.

Timing and other phenological parameters of the spring phytoplankton bloom in the waters around Kamchatka are determined on long-term satellite-derived data for chlorophyll a concentration obtained in 1998–2024. Three methods of the timing determining are compared, as the biomass threshold, the cumulative biomass threshold, and the maximum biomass growth rate; finally, a modified threshold method is proposed, more suitable for the study area. Significant negative correlation (up to r = −0.9) between the spring bloom onset date and its duration is shown. Features of the spring blooms off the western and eastern coasts of Kamchatka are described. A tendency towards later start of the bloom is revealed in the Pacific waters (up to +2 days per year), but opposite tendency occurs in local areas at both coasts, as the western Shelikhov Bay, inner Penzhina Bay, and Karaginsky Bay (up to −1.5 days per year). The bloom tends to last longer (up to +1.5 days per year) in the outer Shelikhov Bay at Yamsky Islands, and becomes more intense (up to +0.1 mg/m ³ per year) in many coastal areas of this bay. Phenological parameters of the spring bloom in the Shelikhov Bay, Karaginsky Bay, and some areas at the western coast of Kamchatka have statistically significant dependence on the timing of sea ice retreat.