Accurate, high-resolution ocean forecasting is crucial for maritime operations and environmental monitoring. While traditional numerical models are capable of producing sub-daily, eddy-resolving forecasts, they are computationally intensive and face challenges in maintaining accuracy at fine spatial and temporal scales. In contrast, recent data-driven approaches offer improved computational efficiency and emerging potential, yet typically operate at daily resolution and struggle with sub-daily predictions due to error accumulation over time. We introduce FuXi-Ocean, the first data-driven global ocean forecasting model achieving six-hourly predictions at eddy-resolving 1/12° spatial resolution, reaching depths of up to 1500 meters. The model architecture integrates a context-aware feature extraction module with a predictive network employing stacked attention blocks. The core innovation is the Mixture-of-Time (MoT) module, which adaptively integrates predictions from multiple temporal contexts by learning variable-specific reliability , mitigating cumulative errors in sequential forecasting. Through comprehensive experimental evaluation, FuXi-Ocean demonstrates superior skill in predicting key variables, including temperature, salinity, and currents, across multiple depths.

Accurately identifying ocean fronts is essential for advancing marine science and effectively managing aquatic resources. Traditional identification methods and certain deep learning approaches often struggle to capture the dynamic complexity of oceanic phenomena due to the limitations in identification efficiency, detail sensitivity, and real-world adaptability. This study develops a new residual U-network (Res-U-Net) architecture for ocean fronts’ identification to overcome these limitations, incorporating residual blocks to improve feature extraction and gradient flow of deep learning networks, which integrates deeper network architectures and comprehensive dataset analysis methods. The study utilizes daily sea surface temperature (SST) data from the Copernicus Marine Environment Monitoring Service (CMEMS), with a 0.05° spatial resolution, covering the Northwestern Pacific (100 °E–150 °E, 0 °N–50 °N) from 1981 to 2022. This strategy significantly improves the identification accuracy and generalizability under versatile oceanic conditions. The experimental results demonstrated that Res-U-Net significantly outperformed other vanilla U-Net baselines. For overall ocean fronts’ identification in the Northwestern Pacific area, the accuracy of Res-U-Net improved by 14% compared to the baseline model. Moreover, for fine-scale ocean fronts’ challenging to identify with vanilla U-Net, Res-U-Net achieved an improvement of approximately 67% in recognition accuracy over U-Net. These results are based on the best-performing models from the experimental Res-U-Net and U-Net groups. These advances not only address the degradation issues caused by gradient explosion in deeper U-Net architectures but also enhance the identification of subtle frontal details. This improvement significantly improves the adaptability of Res-U-Net for the large-scale, multiobjective task of ocean fronts’ identification.

The modeling of the LiDAR underwater echo signal is mainly divided into analytical and statistical methods. In this paper, the radiation transfer equation of LiDAR underwater transfer is established according to the QSSA approximate analytical method theory. For Katsev, the Green function is directly used to solve the radiation transfer equation, and the spectral element method is proposed to calculate the LiDAR radiation flux in the radiation transfer equation. To verify the accuracy of the LS-SEM-Katsev analytical model, the simulation results of the LS-SEM-Katsev analytical model and Semi-MC statistical model are verified under three typical seawater optical parameters. The experimental results show that the decision coefficients of the three simulation models are all above 0.99, which is highly consistent. A comparison between the computational times of LS-SEM-Katsev and Semi-MC simulation models showed that the computational efficiency of the proposed analytical model was 8 orders of magnitude higher than that of the Semi-MC model.

Mesoscale eddy is a typical mesoscale ocean phenomenon, which widely exists in all oceans and marginal seas around the world. The spatiotemporal scale of mesoscale eddies ranges from a few days to hundreds of days, tens of kilometers to hundreds of kilometers. The traditional mesoscale eddy identification is subjective and usually depends on expert discrimination or threshold setting. In this study, due to the significant advantages of YOLO series target recognition models in the field of deep learning, we propose an ocean mesoscale eddy identification algorithm for deep transfer learning target recognition based on YOLOF (You Only Look One Level Feature). Compared with traditional recognition methods, this model has better recognition effect, The influence of setting threshold on mesoscale eddy identification is avoided, and the identification speed is improved to a certain extent.

Water quality parameter is a key index to indicate the quality of water and the variation trend of materials in water. One of the parameters, biochemical oxygen demand () is an important parameter reflecting the condition of organic pollution in water. Current inversion methods for water quality parameter mainly include empirical, semi-empirical and physical method. In this study, due to the advantages of hyperspectral remote sensing with a large number of bands and high spectral resolution, we use semi-empirical method combined with hyperspectral remote sensing image to build ratio linear regression model and ratio quadratic polynomial regression model to invert the BOD5 in Qinghai Lake. The comparative analysis shows that the performance of linear model is better than the quadratic polynomial one. The results show that the R2of inversion accuracy is approximately 0.58 and Root Mean Square Error (RMSE) is below 0.16(O2, mg/L), and the inversion results are basically consistent with the spatial distribution of in situ measured BOD5 values in Qinghai Lake.

Qinghai Lake is the largest lake in China, and its water quality has a great impact on the ecological environment of the surrounding area. Chemical oxygen demand (COD) is an important indicator for water quality. In this study, a remote sensing inverse model for COD was established by using the high-spectral images of Zhuhai-1 satellite with measured sample data and a thematic map of the spatial distribution for COD was drawn in Qinghai Lake. The results showed that COD has the greatest correlation with B2 (460nm) and B14 (670nm) band ratio, and the linear regression model is best in all regression models (R2=0.8453, RMSE=1.06 mg/L), and the model results are basically consistent with the actual situation. This work showed that it is feasible to use hyperspectral images to inverse water quality parameters, and monitor the COD spatial distribution and movement of water bodies in a wide and real-time way.

Net Primary Production (NPP) is a key component of the terrestrial carbon cycle. In this study, the CASA model was applied to analyze the NPP change in the Qinghai Lake Basin in 2020, using the meteorological data and GaoFen-1 data. The results show that the NPP value changed significantly through the whole year. It has the highest NPP$(107.6 gC. m^{-2}\cdot month^{-1})$in July, while the NPP value of January is the lowest at 3.6$gC\cdot m^{-2}\cdot month^{-1}$. Spatially, the annual average NPP in Gangcha County is the highest, followed by Haiyan County and Gonghe County, and the value of Tianjun County is the lowest. This study provides an effective indicator for the environmental evaluation of the Qinghai Lake Basin.

Surface soil moisture (SSM) is an important variable in environmental studies and land surface system research. Remote sensing techniques provide a direct and convenient means to estimate SWC on a regional scale. Land surface temperature (LST) and vegetation index (VI) can be employed to construct a feature space that represents surface dry and wet conditions. A normalized soil water content model was developed to obtain comparable SWC using normalized LST (T*) and VI. The quantitative relationship among SWC, LST and VI was shown by a quadratic polynomial equation. In this study, a uncertainty sensitivity analysis for the input parameters and other factors that might affect the established model was conducted. T* and Fractional vegetation cover (FVC) are key inputs, and the uncertainties introduced by them are necessarily required for the model. The results showed that LST affects the estimation results enormously. The estimation error is approximately 0.05 m3/m3when the LST with 1K uncertainties. And the FVC has a weak influence on the soil moisture estimation. The estimation error will be 0.03 m3/m3when the FVC has an error of 0.2. For targeted analysis, LST changes from -0.2k to 0.2K and FVC has an error of 0.2. The largest error of estimation can reach 0.03 m3/m3when the two variables have the biggest uncertainties.

Soil moisture saturation index (SMSI) is an important indicator that demonstrates the status of the soil water content for drought monitoring. However, at present, most of the methods to calculate the SMSI from the in situ measurement data are inadequate or inaccurate. This paper proposed a simple method to determine the SMSI from the remotely sensed data. Combining the theory of thermal inertia and triangle method, the apparent thermal inertia and fractional vegetation cover can construct a triangular space. In this space, SMSI can be determined easily. Validation was performed with in situ measurements for 19 meteorological stations in the study area. Results indicated that the method can obtain the accurate soil water status that reflects the variation in soil moisture to some extent and is suitable for monitoring the regional surface soil moisture.