Prediction of Landslide Deformation Region Based on the Improved S-Growth Curve Model

Quantitative research on and the prediction of a landslide deformation area is an important point to accurately comprehensively understand failure mechanism landslides degree slope failure. This study uses image processing techniques quantitatively identify volume regions during rainfall-type destabilization under multifactor conditions. The findings revealed that (1) increase in rainfall intensity angle, as well existence crest, will accelerate process instability. In our study, when was 140 mm/h reached 35.68%, most serious. (2) Slopes with high compaction subsoil those without perimeter pressure are relatively more damaged. (3) higher density vegetation cover, stronger protection ability slope, wind speed, greater slope. Furthermore, improved S-growth curve model proposed predict volumes 16 sets experiments. detail, predicted average absolute percentage error 4.34–16.77%. Compared time series analysis moving-average method (average 6.39–68.89%), not only has accuracy but also can describe three stages region development from physical perspective be applied change prediction.