Fast and high-accuracy detection of underwater targets based on side scan sonar images has great potential for marine fisheries, underwater security, marine mapping, underwater engineering and other applications. The following problems, however, must be addressed when using low-resolution side scan sonar images for underwater target detection: (1) the detection performance is limited due to the restriction on the input of multi-scale images; (2) the widely used deep learning algorithms have a low detection effect due to their complex convolution layer structures; (3) the detection performance is limited due to insufficient model complexity in training process; and (4) the number of samples is not enough because of the bad dataset preprocessing methods. To solve these problems, an improved neural network for underwater target detection—which is based on side scan sonar images and fully utilizes spatial pyramid pooling and online dataset preprocessing based on the You Look Only Once version three (YOLO V3) algorithm—is proposed. The methodology of the proposed approach is as follows: (1) the AlexNet, GoogleNet, VGGNet and the ResNet networks and an adopted YOLO V3 algorithm were the backbone networks. The structure of the YOLO V3 model is more mature and compact and has higher target detection accuracy and better detection efficiency than the other models; (2) spatial pyramid pooling was added at the end of the convolution layer to improve detection performance. Spatial pyramid pooling breaks the scale restrictions when inputting images to improve feature extraction because spatial pyramid pooling enables the backbone network to learn faster at high accuracy; and (3) online dataset preprocessing based on YOLO V3 with spatial pyramid pooling increases the number of samples and improves the complexity of the model to further improve detection process performance. Three-side scan imagery datasets were used for training and were tested in experiments. The quantitative evaluation using Accuracy, Recall, Precision, mAP and F1-Score metrics indicates that: for the AlexNet, GoogleNet, VGGNet and ResNet algorithms, when spatial pyramid pooling is added to their backbone networks, the average detection accuracy of the three sets of data was improved by 2%, 4%, 2% and 2%, respectively, as compared to their original formulations. Compared with the original YOLO V3 model, the proposed ODP+YOLO V3+SPP underwater target detection algorithm model has improved detection performance through the mAP qualitative evaluation index has increased by 6%, the Precision qualitative evaluation index has increased by 13%, and the detection efficiency has increased by 9.34%. These demonstrate that adding spatial pyramid pooling and online dataset preprocessing can improve the target detection accuracy of these commonly used algorithms. The proposed, improved neural network with spatial pyramid pooling and online dataset preprocessing based on the YOLO V3 method achieves the highest scores for underwater target detection results for sunken ships, fish flocks and seafloor topography, with mAP scores of 98%, 91% and 96% for the above three kinds of datasets, respectively.

The quality of the drinking water source reservoirs has always been a research hotspot. However, few have studies focused on the water quality of reservoirs over a relatively long period with time series analysis. In this paper, based on water quality and hydrological data from 2010 to 2020, considering 8 water quality parameters, CCME-WQI with time series analysis was used to explore the interannual and seasonal changes in water quality in the Weishui Reservoir. Furthermore, the main factors affecting water quality were discussed through correlation analysis. The ARIMA model is used to predict water quality in the future. The results show that the water quality was seriously polluted from 2012 to 2013. After 2018, the water quality gradually improved and stabilized. In addition, the water quality is affected by inflow, showing the characteristics of poor water quality in summer and winter. The key parameters affecting water quality are TN and TP, which are almost 2 times higher than the grade II standard of water quality standard. Through the ARIMA model, it is predicted that CCME-WQI is maintained at 80.46 indicating that the water quality will be stable in the future.

Honghu Lake is the largest lake-type wetland in Hubei Province, China. It is also one of the largest shallow lakes in the Yangtze River Economic Zone, a key area in the relatively more developed southeast of China. However, the water quality has seen a deterioration tendency in recent decades, mostly owing to unreasonable human activities such as lake enclosure aquaculture following rapid social and economic development. Based on the water quality index (WQI) method, the water quality of Honghu Lake, by the vast amount of data collected from five observation sites monitored over ten years, was analyzed and evaluated. The results show that: (i) the water quality of Honghu Lake is in the ‘General’ grade as a whole with a WQI value of 43.41 ± 6.66; (ii) the water quality has been improving in the recent two years, reversing its decade-long deterioration; (iii) the water quality sampled at the Lantian site is the worst while that of the Guandun site is the best; (iv) the concentration of Pb and Cd are the key parameters to determine the water quality of Honghu Lake. Therefore, it can be concluded that more attention should be paid to investigate heavy metals in Honghu Lake in the future.

The security of drinking water is a serious issue in China and worldwide. As the backup source of drinking water for the Changde City in China, the Huangshi Reservoir suffers from the threat of eutrophication due to the water quality of the reservoir ecosystem being affected by the tributaries that carry Non-Point Source (NPS) pollutants. The calculation of the water environmental capacity (WEC) can provide a scientific basis for water pollution control, which refers to the maximum amount of pollutants that the water can accommodate. In this paper, according to the hydrological characteristics of the river–reservoir combination system, a one-dimensional (1-D) water quality model and the Environmental Fluid Dynamics Code (EFDC) model were chosen to calculate the water environmental capacity of each functional zone in this basin. The quantity control of pollution from the tributaries was conducted based on the combined results of the water environmental capacity calculation from the EFDC model and a one-dimensional (1-D) river water quality model. The results show that total water environmental capacity of the tributaries included a chemical oxygen demand (COD) of 421.97 tons; ammonia nitrogen (NH₃-N) of 40.99 tons; total nitrogen (TN) of 35.94 tons; and total phosphorus (TP) of 9.54 tons. The water environmental capacity of the Huangshi Reservoir region accounts for more than 93% of the total capacity. The reduction targets of the major pollutants in the Huangshi Reservoir and its four major input rivers, which are, namely, the Bamao River, the Longtan River, the Fanjiafang River, and the Dongtan River, have been determined to achieve the water quality objectives for the reservoir in 2020 and 2025. The results will be helpful for the local water quality management and will provide a valuable example for other similar water source reservoirs.

Traditional methods for water quality assessment often overlook the uncertainty of water quality data during the sample collection process, leading to limitations in their application. Therefore, this study combines the comprehensive water quality index (CWQI) method and the improved CWQI method based on CRITIC with the Monte Carlo method to evaluate the water quality in the Weishui Reservoir watershed. The results indicate that (1) there is a noticeable difference in water quality between the Shaxiping and Dayanzui sampling points. The water quality at the Shaxiping sampling point is excellent, with a water quality classification of Class I. In contrast, the water quality at the Dayanzui sampling point is comparatively poorer, with an average water quality classification of Class III. (2) Sensitivity analysis shows that TN, NH4+-N, and TP are more sensitive than other indicators, suggesting that they are the primary factors influencing the evaluation results. (3) Compared to the traditional CWQI method, combining the CRITIC-based improved CWQI method with the Monte Carlo method is more scientifically rigorous. It considers the variety of evaluation indicators, allocates weights rationally, and provides evaluation results that align better with seasonal variations, resulting in higher discriminative power.

Many accidents of releasing toxic pollutants into surface water happen each year in the world. It is believed that dam cooperation can affect flow field in reservoir and then can be applied to avoiding and reducing spread speed of toxic pollutants to drinking water intake mouth. However, few studies investigated the effects of dam cooperation on the spread characteristics of toxic pollutants in reservoir, especially the source reservoir for water diversion with more than one dam. The Danjiangkou Reservoir is the source reservoir of the China’ South‐to‐North Water Diversion Middle Route Project. The human activities are active within this reservoir basin and cyanide‐releasing accident once happened in upstream inflow. In order to simulate the spread characteristics of cyanide in the reservoir in the condition of dam cooperation, a three‐dimensional water quality model based on the Environmental Fluid Dynamics Code (EFDC) has been built and put into practice. The results indicated that cooperation of two dams of the Danjiangkou Reservoir could be applied to avoiding and reducing the spread speed of toxic cyanide in the reservoir directing to the water intake mouth for water diversions.