Upon planning of optimum measures for prediction and removal of reasons for landslides the remote sensing images and the digital elevation models are frequently used for mapping of their special features. The landslides can occur in any degraded deep slopes under load. To analyze any 2D and 3D images the methods of geomorphometry are used for qualitative analysis of the earth surface. In the article the questions on modification of known geomorphometric model for detection of landslides using lidar are considered. The model includes the global and local procedures for detection and use methods of classification for global detection of landslides. It is noted that the great variety in geological and topographical properties of possible landslides sites lead to inevitable errors in solution of detection tasks using the global approach. To increase the accuracy of detection of local landslides the analysis of errors for minimization of possibility abnormal errors is necessary. The analysis of operation of geomorphometric model of identification of landslide using lidar has shown the necessity for its modification by way of completion of the model with devices of internal calibration. The algorithm of calibration is developed in line which upon condition of equality of topographic wetness index and complex topographic index and upon changing values of statistical parameters of latter the possibility to calculate the values of specific calibration parameters does occur.