Water resources utilization, social economy development, and ecological environment protection are key factors in regional sustainable development. Scientific evaluation of regional coordinated development status and diagnosis of regional uncoordinated development constraints will improve the management level of decision-makers. At present, most developing countries have the problem of unbalanced regional development caused by the one-sided pursuit of a certain system. Taking 14 prefecture-level cities in Hunan Province as cases, this paper analyzed the spatial and temporal distribution characteristics of the carrying capacity level of the water resources system, the development level of the social economy system and the protection level of the ecological environment system in each evaluation unit based on entropy weight method and order parameter analysis. Based on the theory of coordinated development, a calculation model of a coordinated development degree was constructed, and the corresponding evaluation criteria were formulated. The spatial and temporal distribution characteristics of a coordinated development degree in each research unit were analyzed and evaluated. The results showed that the average coordinated development degree of Hunan Province from 2004 to 2016 evolved from “Light disorder recession” to “Nearly disorder recession”, then to “Reluctance coordinated development”. Restricted by different systems, the coordinated development degree in each research unit presented spatial and temporal differences. According to different development stages and the characteristics of different regions, corresponding development strategies can be formulated to provide the guidance for coordinated the development of regions.

Wet–dry encounters between basins and regions have an important impact on the allocation of water resources. This study proposes a multi-objective allocation model for basin water resources under full probability scenarios considering wet–dry encounters (FPS-MOWAM) to solve the problem of basin water resource allocation. In the FPS-MOWAM model, the sub-regions were merged by precipitation correlation analysis. Next, the joint probability distribution of basin runoff and region precipitation was constructed using copula functions. The possible wet–dry encounter scenarios and their probabilities were then acquired. Finally, the multi-objective allocation model of water resources was constructed using the full probability scenario for wet–dry encounters in each region. The FPS-MOWAM is calculated by the NSGA-II algorithm and the optimal water resource allocation scheme was selected using the fuzzy comprehensive evaluation method. Using the Yellow River Basin as an example, the following conclusions were obtained: (1) the Yellow River Basin can be divided into four sub-regions based on precipitation correlations: Qh-Sc (Qinghai, Sichuan), Sg-Nx-Nmg (Gansu, Ningxia, Inner Mongolia), Sxq-Sxj (Shaanxi, Shanxi), and Hn-Sd (Henan, Shandong), (2) the inconsistencies in synchronous–asynchronous encounter probabilities in the Yellow River Basin were significant (the asynchronous probabilities were 0.763), whereas the asynchronous probabilities among the four regions were 0.632, 0.932, and 0.763 under the high, medium, and low flow conditions in the Yellow River Basin respectively, and (3) the allocation of water resources tends to increase with time, allocating the most during dry years. In 2035, the expected economic benefits are between 11,982.7 billion CNY and 12,499.6 billion CNY, while the expected water shortage rate is between 2.02% and 3.43%. In 2050, the expected economic benefits are between 21,291.4 billion CNY and 21,781.3 billion CNY, while the expected water shortage rate is between 1.28% and 6.05%.

We propose a flood risk management model for the Taihu Basin, China, that considers the spatial and temporal differences of flood risk caused by the different climatic phenomena. In terms of time, the probability distribution of climatic phenomenon occurrence time was used to divide the flood season into plum rain and the typhoon periods. In terms of space, the Taihu Basin was divided into different sub-regions by the Copula functions. Finally, we constructed a flood risk management model using the Copula-based Bayesian network to analyze the flood risk. The results showed the plum rain period occurs from June 24 to July 21 and the typhoon period from July 22 to September 22. Considering the joint distribution of sub-region precipitation and the water level of Taihu Lake, we divided the Taihu Basin into three sub-regions (P-I, P-II, and P-III) for risk analysis in the plum rain period. However, the Taihu Basin was used as a whole for flood risk analysis in the typhoon period. Risk analysis indicated a probability of 2.4%, and 0.8%, respectively, for future adverse drainage during the plum rain period and the typhoon period, the flood risk increases rapidly with the rising water level in the Taihu Lake.

Simulating river discharge evolution is fundamental to water resources management in large river basins. Nevertheless, human activities such as intensive water abstraction, reservoir operation, and other anthropogenic interventions have rendered the patterns of river discharge evolution highly complex, marked by significant spatiotemporal variability and nonlinear hydrological–anthropogenic interactions. Therefore, to improve the accuracy of large river discharge evolution, taking the Yellow River as an example, we developed a discharge evolution model by employing the convolutional neural network–long short-term memory (CNN-LSTM) model. Subsequently, through parameter optimization using the shuffled complex evolution–University of Arizona (SCE-UA) algorithm, the convolutional neural network-long short-term memory coupled with the shuffled complex evolution algorithm developed at the University of Arizona (CNN-LSTM-SCE-UA) model effectively mitigates the error propagation phenomenon between river reaches. Additionally, a comparison was conducted between the deep learning model and the physical mechanism model for river discharge evolution, aiming to explore the causes of simulation discrepancies from the perspective of physical mechanisms. Results showed that (1) the hydrological method, CNN-LSTM model, and CNN-LSTM-SCE-UA model achieved accuracies of 0.627, 0.613, and 0.774, respectively (evaluated via coefficient of determination, root-mean squared error, and mean absolute error; values closer to one indicate better agreement with observed data); (2) the spatial distribution pattern of unbalanced water volume—an external manifestation of the physical mechanism governing discharge evolution in the Yellow River main stream—followed an overall trend: middle reaches > lower reaches > upper reaches; and (3) for both the hydrological method and CNN-LSTM-SCE-UA model, simulation accuracy in the upper and lower reaches exceeded that in the middle reaches, consistent with the aforementioned spatial distribution pattern of unbalanced water volume.