To address the low accuracy and poor stability of water level recognition in complex natural environments, which result from illumination changes, alternation between wet and dry conditions, and background interference, an intelligent water level recognition method that integrated semantic segmentation with spatiotemporal information was proposed. An FCN-RCA model was constructed by combining residual connections, a convolutional attention mechanism, and an atrous convolution module; an image segmentation optimization method based on the flooding algorithm was introduced, and highly robust semantic segmentation and structural integrity optimization for the water body area were achieved. The segmentation results were mapped to the physical space of the slope domain via an affine transformation to build a virtual water gauge, enabling precise quantification of the water level. By leveraging the spatiotemporal continuity of video frames, temporal consistency optimization was applied to the water level boundary to enhance the stability and reliability of the recognition results. The case study at the Taleldesayi Station in Xinjiang demonstrates that the proposed method achieves a mean intersection over union (mIoU) exceeding 0.92 under various meteorological, hydrological, and illumination conditions, exhibiting excellent segmentation performance and accurate extraction of water level boundaries. The dynamic water level optimization method based on spatiotemporal information effectively ensures the stability and continuity of water level boundary recognition. Compared with the pre-optimization result, the absolute mean absolute error (MAE) and root mean square error (RMSE) reduce by 0.006 m and 0.004 m, respectively, while the Nash-Sutcliffe efficiency (NSE) increases by 0.024. In validation with long-term measured data, the MAE for water level recognition remains within 0.04 m, and the NSE reaches up to 0.997. These results confirm that the method can achieve centimeter-level water level recognition in complex environments, with good temporal consistency and engineering applicability.

Based on the FVCOM of the finite volume algorithm, a three-dimensional hydrodynamic-salinity-submerged dike coupling numerical model of the tidal river estuary was constructed by using the embankment treatment method under the unstructured grid. Taking the Minjiang River estuary as the research object, which is significantly affected by saltwater intrusion, the characteristics and laws of saltwater intrusion in the tidal reach of the Minjiang River under the influence of submerged dike construction were explored. The results demonstrate that the submerged dikes effectively suppress saltwater intrusion, with a more pronounced reduction in bottom-layer salinity than surface salinity and a stronger inhibitory effect during spring tides compared to neap tides. During spring and neap tides, the salt-suppressing effect of submerged dikes on shallow and wide channels with obvious water stratification is significantly better than that on narrow and deep channels with fully mixed water. During spring tide, the freshwater-saltwater interface in the North Channel retreated approximately 5 km and 8 km downstream due to the influence of Xia’nan and Jin’gangtui submerged dikes, respectively. Further analysis of water velocity and flow direction reveals that submerged dikes alter the flow velocity and direction of bottom water body, causing the bottom water flow to deflect while generating small-scale vortices, further suppressing the continuous upstream movement of saltwater.

Based on the finite volume method, combined with measured saltwater data and climate model data, a hydrodynamic-salinity model for the dry season of tidal reaches was constructed. Taking the tidal reaches from the dam of the Minjiang Shuikou Reservoir to the estuary, which is severely affected by saltwater, as an example, the upstream characteristics and laws of saltwater intrusion under dry season climate conditions were explored. The results show that during the dry season, the average flow rate of the tidal reaches of the Minjiang River is the smallest, and the climate wind field is in the northeast direction. Under the comprehensive effect, it is most prone to the occurrence of saltwater intrusion. During the dry season climatological spring tide, the mixing of estuary water is strong, and the saltwater intrusion of the South Channel is more intense than the North Channel, which has endangered the Chengmen Waterworks when the saltwater intrusion is the most serious. In the dry season climatological neap tide, the degree of saltwater intrusion is stronger than that of the spring tide, and the South Channel intrusion is much farther than the North Channel. The water stratification is obvious in the tidal fall and the water mixing strongly at the rest of the time. When the saltwater intrusion is severe, the Mawei, Yanshan and Chengmen waterworks are affected during the peak period of water intake.