Abstract: Global climate change has significant influenced the meteorological,hydrological,and ecological environments of the Tibetan Plateau.Limited by the sparse field monitoring data,the river and lake hydrological forecasting and risk assessment in the Tibetan Plateau faces great challenges.To investigate the change and potential risks of the river and lake in the Tibetan Plateau,it is imperative to integrate multi-source data and construct multi-scale forecasting models.This paper introduces a cloud-native based river and lake hydrological forecasting and risk assessment system in the Tibetan Plateau.The system builds a container-based model parallel computing scheme,which can dynamically allocate and deploy containers according to the task scale,effectively improving the computing efficiency of the model and timeliness of the forecast.At the same time,the containers deploy of geographic databases and microservices which realizes an efficient response of complex decision systems to various data requests and map services.Further,tests on concurrent performance with different containerized deployment strategies have confirmed the system’s advantages in availability and performance.Implementing the cloud-native architecture for multi-model parallel computations and complex geographic information visualization,the system provides an intuitive display of hydrological situation changes and risk warning information.This study provides foundations for the advancement of river and lake risk management.