The most important motivation for streamflow forecasts is flood prediction and longtime continuous prediction in hydrological research. As for many traditional statistical models, forecasting flood peak discharge is nearly impossible. They can only get acceptable results in normal year. On the other hand, the numerical methods including physics mechanisms and rainfall-atmospherics could provide a better performance when floods coming, but the minima prediction period of them is about one month ahead, which is too short to be used in hydrological application. In this study, a deep neural network was employed to predict the streamflow of the Hankou Hydrological Station on the Yangtze River. This method combined the Empirical Mode Decomposition (EMD) algorithm and Encoder Decoder Long Short-Term Memory (En-De-LSTM) architecture. Owing to the hydrological series prediction problem usually contains several different frequency components, which will affect the precision of the longtime prediction. The EMD technique could read and decomposes the original data into several different frequency components. It will help the model to make longtime predictions more efficiently. The LSTM based En-De-LSTM neural network could make the forecasting closer to the observed in peak flow value through reading, training, remembering the valuable information and forgetting the useless data. Monthly streamflow data (from January 1952 to December 2008) from Hankou Hydrological Station on the Yangtze River was selected to train the model, and predictions were made in two years with catastrophic flood events and ten years rolling forecast. Furthermore, the Root Mean Square Error (RMSE), Coefficient of Determination (R2), Willmott’s Index of agreement (WI) and the Legates-McCabe’s Index (LMI) were used to evaluate the goodness-of-fit and performance of this model. The results showed the reliability of this method in catastrophic flood years and longtime continuous rolling forecasting.

Storm surge is one of the most destructive marine disasters to life and property for Chinese coastal regions, especially for Guangdong Province. In Huizhou city, Guangdong Province, due to the high concentrations of chemical and petroleum industries and the high population density, the low-lying coastal area is susceptible to the storm surge. Therefore, a comprehensive risk assessment of storm surge over the coastal area of Huizhou can delimit zones that could be affected to reduce disaster losses. In this paper, typhoon intensity for the minimum central pressure of 880, 910, 920, 930, and 940 hPa (corresponding to a 1000-, 100-, 50-, 20-, and 10-year return period) scenarios was designed to cover possible situations. The Jelesnianski method and the Advanced Circulation (ADCIRC) model coupled with the Simulating Waves Nearshore (SWAN) model were utilized to simulate inundation extents and depths of storm surge over the computational domain under these representative scenarios. Subsequently, the output data from the coupled simulation model (ADCIRC–SWAN) were imported to the geographic information system (GIS) software to conduct the hazard assessment for each of the designed scenarios. Then, the vulnerability assessment was made based on the dataset of land cover types in the coastal region. Consequently, the potential storm surge risk maps for the designed scenarios were produced by combining hazard assessment and vulnerability assessment with the risk matrix approach. The risk maps indicate that due to the protection given by storm surge barriers, only a small proportion of the petrochemical industrial zone and the densely populated communities in the coastal areas were at risk of storm surge for the scenarios of 10- and 20-year return period typhoon intensity. Moreover, some parts of the exposed zone and densely populated communities were subject to high and very high risk when typhoon intensities were set to a 50- or a 100-year return period. Besides, the scenario with the most intense typhoon (1000-year return period) induced a very high risk to the coastal area of Huizhou. Accordingly, the risk maps can help decision-makers to develop risk response plans and evacuation strategies in coastal communities with a high population density to minimize civilian casualties. The risk analysis can also be utilized to identify the risk zones with the high concentration of chemical and petroleum industries to reduce economic losses and prevent environmental damage caused by the chemical pollutants and oil spills from petroleum facilities and infrastructures that could be affected by storm surge.

By collecting surface sediment samples from 158 stations in the near-shore waters of eastern Guangdong, grain size analysis and grain size parameter calculations were performed to explore the characteristics and migration trends of surface sediments in the area. The analysis of the grain size results showed that the surface sediments in the nearshore waters of east Guangdong could be classified into nine sediment types, mainly including seven types of gravel sand ((g)S), gravel muddy sand ((g)mS), gravelly mud ((g)M), sand (S), silty sand (zS), sandy silt (sZ) and silt (Z). The relative percentages of gravel, sand, silt and mud were 0.7%, 40.56%, 46.7% and 12.04%, respectively. The average grain size varied from −2φ to 8φ, with an average of 4.94φ. The selection coefficient ranged from 0.44 to 3.78, with an average value of 1.8. The skewness distribution ranged from −0.34 to 0.67, with an average value of 0.07. By extracting and analyzing the spatial distribution information of grain size in the study area, using the Gao–Collins migration trend analysis method and incorporating dynamic factors such as tidal currents and waves, the transport direction and trend of surface sediments in the study area could be analyzed and inferred. The results show that the surface sediment migration trend was significant, migration on the north side of Nan’ao Island was in an east-to-west direction, and the sediment of Yifeng River was mainly deposited to the sand spout at the mouth of Lianyang River. After southward transport from the Houjiang waterway, the migration was mainly southeastward and the trend was quite significant until the 20 m isobath, where the trend gradually decreased. The sediments of the Rongjiang River were mainly deposited outside the mouth of Niutian Yang and Rongjiang River, and the surface sediments of Guang’ao Bay and Haimen Bay migrated in the northwest–southeast direction. After the 30 m isobath, the southeast corner of the study area migrated in the southeast–south direction. This sediment transport pattern revealed by the grain size migration trend is in good agreement with the physical and hydrodynamic conditions of the study area and provides an important reference for decisions regarding port dredging and waterway management in the area.

Based on the water quality monitoring results from 2002 to 2018 in the coastal waters of Shantou City, this study used eight methods, including the single-factor index method, eutrophication index method, nutrient status quality method, comprehensive index method, organic pollution index method, nitrogen–phosphorus ratio method, potential eutrophication evaluation, and dissolved oxygen saturation parameter method, to evaluate the eutrophication and organic pollution status of the area. The applicability of the first-generation eutrophication model was assessed from the perspective of red tide occurrence, and a comparison was made among the various evaluation methods. The results indicate that the coastal waters of the study area were mostly in a phosphorus-limited state, with a few periods of nutrient deficiency and a very small portion of time in a moderately nutritious state with both nitrogen and phosphorus limitations. The overall N/P ratio showed an increasing trend primarily due to the gradual increase in inorganic nitrogen content. In terms of evaluating eutrophication based on red tide symptoms, the comprehensive index method showed better performance in reflecting eutrophication conditions, while the eutrophication index method and dissolved oxygen saturation parameter method highlighted problematic areas but often exaggerated their extent. The nutrient status quality method and organic pollution index method did not provide distinctive indications and could not verify red tide symptoms. The nitrogen–phosphorus ratio method and potential eutrophication evaluation could reflect certain nutrient structures and eutrophication characteristics but also failed to verify red tide symptoms. Furthermore, this study found a certain correlation between the occurrence of red tide and the ratio of nitrate nitrogen to ammonia nitrogen, which consistently ranged from 1.15 to 1.94 during red tide events. Therefore, whether this ratio can serve as a characteristic value for red tide occurrence and be used for early warning and monitoring still requires further validation.