Trend detection in observations helps one to identify anthropogenic forces on natural hydrological and climatic systems. Hydrometeorological data are often persistent over time that deviates from the assumption of independence used by many statistical methods. A recently proposed Sen’s trend test claimed to be free of this problem and thereby received widespread attention. However, both theoretical derivation and stochastic simulation of the current study implies that persistence inflates the trend significance, leading to false trends. To tackle this problem, we incorporate the feature of persistence into the variance of the trend test statistic, whereby an innovative variance-corrected Sen’s trend test is developed. Two theoretical variances of the trend test statistic are newly derived to account for short-term and long-term persistent behavior. The original variance for independent data is also corrected because of its negative bias. A stepwise procedure, including steps to specify the underlying persistent behavior and to test trend with new statistic, is outlined for performing the new test on factual data. Variance-corrected Sen’s trend test can effectively restore the inflated trend significance back to its nominal state. This study may call for the reassessment of published results of the original Sen’s trend test on data with persistence.

The Hanjiang River, as a water donor basin, plays a fundamental role in supporting water supply security in northern China while maintaining the health and stability of ecosystems within the basin. However, the combined influence of climate change and anthropogenic interference has resulted in a significant change in the flow regime of the basin, challenging the sustainability of the river system. In order to understand the impact of the above factors on the river runoff, we analyzed the temporal and spatial pattern of runoff and climate factors in the basin and quantitatively assessed the contribution of climate change and human activities to the change in runoff using the elasticity coefficient method. Our results indicate that annual runoff has experienced a significant downward trend over the past 60 years, which is projected to continue into the future. It is also found that the temporal pattern of the runoff regime differed upstream and downstream of the Danjiangkou Reservoir due to the joint operation of the reservoir and China’s Middle Route Project of South-to-North Water Diversion (MRP-SNWD). A significant decrease in runoff was primarily attributed to human activities, followed by precipitation. In contrast, evapotranspiration had the least effect. In particular, the MRP-SNWD was a significant anthropogenic factor, contributing to about 20.3% of the total change in runoff. Our results highlighted the unfavorable effects of human activity on the hydrological system in the Hanjiang River and provided some constructive suggestions to turn vulnerability into resilience.

Meteorological droughts are natural disasters that have been linked to economic losses and casualties. Decision-makers need to understand the temporal and spatial variation of meteorological drought events at a daily to weekly scale to develop a more elaborate framework for drought risk management. The present study used the standardized weighted average of precipitation index (SWAP) as an indicator of meteorological droughts, computed from the daily precipitation dataset (1960–2015) of 34 meteorological stations in Chongqing, China. The multi-threshold run theory was applied to identify drought events. Variation of drought characteristics was estimated by the modified Sen’s trend test. The results suggested the following findings: (1) the onset, duration, and severity of drought events identified by the SWAP index are in good agreement with the real local records; (2) there was no significant linear trend and abrupt change in annual duration and severity of drought events, but the decadal variation was obvious. From a decadal perspective, the annual frequency, duration, and severity of drought events showed a steady decreasing trend before the 1990s, and then fluctuated upward; (3) the spatial variation of the duration and severity of ordinary drought events was quite inconsistent at different periods. The annual drought days and severity increased from 1960 to 2015 but decreased after 1990. From 1960 to 2015, the duration days and severity of persistent, long persistent, severe, and extreme drought events declined insignificantly in most parts of the middle and southeast regions but increased in the western and northeast regions. The drought situation in Chongqing shows a large range of variation and obvious spatial heterogeneity. The SWAP index is an effective tool to identify the evolution of daily scale meteorological drought events.

The western Pacific subtropical high (WPSH) is one of the key systems affecting the summer rainfall over the Yangtze River Valley in China. In this study, the forecasting capacity of the WPSH for summer rainfall and streamflow is evaluated based on the WPSH index (WPSHI) derived from the NCEP/NCAR reanalysis dataset. It has been found that WPSHI can identify extreme flood years with a higher skill than normal wet years. Specifically, exceedance probability forecasting based on WPSHI has higher skills for higher thresholds of rainfall. For streamflow, adding WPSHI as a predictor only enhances the skill for higher thresholds of streamflow relative to models based on antecedent streamflow. Under the same framework, performances of two postprocessing approaches for dynamical forecasts, i.e., the model output statistics (MOS) approach and the reanalysis-based (RAN) approach are compared. Hindcasts from Climate Forecast System version 2 from the National Center for Environmental Prediction (CFSv2) are used to calculate WPSHI, which is used as the predictor for rainfall and streamflow. The result shows that the RAN approach performs better than the MOS approach. This study emphasizes the fact that the forecasting skill of exceedance probability would largely depend on the selected threshold of the predictand, and this fact should be noticed in future studies in the long-term forecasting field.

Data-driven models have been successfully applied to flood prediction. However, the nonlinearity and uncertainty of the prediction process and the possible noise or outliers in the data set will lead to incorrect results. In addition, data-driven models are only trained from available datasets and do not involve scientific principles or laws during the model training process, which may lead to predictions that do not conform to physical laws. To this end, we propose a flood prediction method based on data-driven and knowledge-guided heterogeneous graphs and temporal convolutional networks (DK-HTAN). In the data preprocessing stage, a low-rank approximate decomposition algorithm based on a time tensor was designed to interpolate the input data. Adding an attention mechanism to the heterogeneous graph module is beneficial for introducing prior knowledge. A self-attention mechanism with temporal convolutional network was introduced to dynamically calculate spatiotemporal correlation characteristics of flood data. Finally, we propose physical mechanism constraints for flood processes, adjusted and optimized data-driven models, corrected predictions that did not conform to physical mechanisms, and quantified the uncertainty of predictions. The experimental results on the Qijiang River Basin dataset show that the model has good predictive performance in terms of interval prediction index (PI), RMSE, and MAPE.

We investigate the modulus-based matrix splitting iteration algorithms for solving the linear complementarity problems (LCPs) and propose a new model to solve it. The structure of the new model is straightforward and can be extended to other issues of type of complementarity. Convergence analysis of the new approach for the symmetric positive definite matrix is also discussed. Numerical results of the proposed approach are compared with the existing algorithms to show its efficiency.