In view of the problem that traditional methods are difficult to accurately evaluate the stability of sandy soil slopes under water level changes and ignore the coupling effect of spatial variability of soil parameters and water-induced parameter weakening,the reliability of channel slopes is studied.Random field is used to describe the spatial variability of soil multi-parameters.Python is used to realize parameter update,and stability reliability analysis is carried out on the basis of Abaqus finite element software and Monte Carlo simulation.The results are compared with indoor tests and traditional methods.The results show that the failure probability obtained by the new method is 0.125,with shallow sliding accounting for 79%,which is highly consistent with the indoor test characteristics.Rising water level reduces the failure probability,while falling water level aggravates instability.When the water level difference increases from 1 m to 4 m,the failure probability of the new method increases by 80.3%.From largest to smallest,the influence weights of spatial variability of soil parameters are the variation coefficient of internal friction angle,the variation coefficient of mutual correlation coefficient,the variation coefficient of permeability coefficient,and the variation coefficient of cohesion,among which the variation coefficient of internal friction angle increases the failure probability by 0.205.