Exploring the evaluation methods and influencing factors of the effectiveness of online information relief in uncertain environments has important theoretical and practical significance for governments, relevant institutions, and media to improve the quality of information relief. Aiming at the uncertain environment caused by public events, based on the social support theory, this study divides the online information of the rainstorm event in Hebei Province into two dimensions: social support demand and social support supply. Based on the definition of information relief, the social support supply information is regarded as relief information. Based on the uncertainty reduction theory, we construct an evaluation system for the effectiveness of online information relief from the passive, active and interactive dimensions of uncertainty reduction strategies. Then, based on the heuristic system model, a regression prediction model and a SHAP model are constructed to reveal the important factors affecting the effectiveness of online information relief. It is found that significant differences exist in the effectiveness of information relief across different life cycles and theme types. The overall scores for the effectiveness of information relief during the outbreak and spread periods were relatively high, while the situational information during the incubation period, rescue information during the outbreak period, recommendation information for the spread period, and emotional information during the recovery period score higher for the effectiveness of relief in their respective stages. The influence and timeliness of publishers have significant positive influence on the effectiveness of information relief, while the emotion of information is negatively correlated with the effectiveness of information relief.