Abstract: In this study, black locust n Mizhi area of Shaanxi Province and Changwu area of Shaanxi Province in semi-humid area were selected as the objects. By measuring the midday leaf water potential of black locust on the ground, multi-spectral images of black locust were obtained synchronously and its band reflectance was extracted to construct the vegetation index. Firstly, the difference of midday leaf water potential between the two places was analyzed. Secondly, the correlation between spectral index and ground measured data is analyzed. Finally, combine variables in three ways: 5-band combination variable, 5-band + 10 vegetation index combination variable and full subset screening variable. Random forest(RF), support vector machine(SVM) and radial basis neural network(RBFNN) models were constructed respectively. The results showed that the correlation between vegetation index and midday leaf water potential was good, in which the ratio index of red to green was good(RGRI) and Improved Enhanced Vegetation Index(EVI_reg) had the highest correlation with Ψ_md, and the correlation coefficients were 0.39 and -0.51, respectively. Normalized vegetation index(NDVI), nonlinear index(NLI), red-green ratio index(RGRI) and improved Enhanced Vegetation index(EVI_reg) were the optimal combinations of all the selected variables. RF method is the optimal model construction method, and the R² and RMSE of RF model validation set are 0.76 and 0.27 MPa. This study can provide technical guidance for the monitoring of black locust moisture based on multispectral images