We applied the Mann-Kendall (MK) test and Bayesian model to systematically explore trends and abrupt changes of the precipitation series in the Pearl River basin. The results showed that no significant trends were detected for annual precipitation and summer or winter precipitation totals. Significant negative trends were identified for the number of rainy days across the Pearl River basin; significant positive trends were observed regarding precipitation intensity (PI). In particular, the precipitation totals and frequencies of extremely high precipitation events are subject to significant positive trends. In addition, the number of extremely low precipitation events was also increasing significantly. Factors affecting the changes in precipitation patterns are the weakening Asian monsoon and consequently increasing moisture transport to Southern China and the Pearl River basin. In summary, the main findings of this study are: (1) increased precipitation variability and high-intensity rainfall was observed though rainy days and low-intensity rainfall have decreased, and (2) the amount of rainfall has changed little but its variability has increased over the time interval divided by change points. These finds indicate potentially increased risk for both agriculture and in locations subject to flooding, both urban and rural, across the Pearl River basin.

In this study, we propose a new multi-task learning approach for rumor detection and stance classification tasks. This neural network model has a shared layer and two task specific layers. We incorporate the user credibility information into the rumor detection layer, and we also apply attention mechanism in the rumor detection process. The attended information include not only the hidden states in the rumor detection layer, but also the hidden states from the stance detection layer. The experiments on two datasets show that our proposed model outperforms the state-of-the-art rumor detection approaches.

No abstract

No abstract