For the conservation of museum objects on display,it is important to monitor environmental risks and locate risk variables in a timely way.However,due to the differences of exhibition halls for various cultural relics,it is difficult to apply the environmental risk classification criteria for a particular exhibition hall to other exhibitions,or even to different types of cultural relics in the same exhibition hall.In addition,when a risk appears,a museum may not find the influencing factors in time,which prolongs the length of time of the risk.Therefore,in combination with the continuous monitoring data of the temporary exhibition"Jiangnan Culture and Art"in Shanghai Museum,a two-level monitoring framework for environmental risk state assessment of paper cultural relics has been proposed.The risk monitoring is carried out by four levels,namely"excellent,good,medium and poor"according to the.demands.Firstly,a principal component analysis(PCA)model is established using the"excellent"dataset and the control limit is constructed to select"excellent"data.The control limit of"poor"data is obtained using the existing PCA model.Then,a ReliefF-PCA model is established using the"non-excellent and non-poor"dataset and the control limit of"good"and"medium"classification is obtained.Finally,a contribution plot and a weighted contribution plot based on the ReliefF algorithm are used for the corresponding PCA model and ReliefF-PCA model respectively,thus quantifying the risk and initially locating the risk variables that have an impact on the environmental indicators.The monitoring framework constructed by the two evaluation models provides a more scientific scheme for the evaluation of the environmental quality of cultural relic preservation and the subsequent adjustment of the museum environment.This is helpful for improving the efficiency of the whole cultural relic conservation system.