Soil surface subsidence time series modeling of an area with Aridisols and Vertisols complex using surveying and drone imagery in Central Iran

Soil surface subsidence is a natural hazard that has been reported in arid and semi-arid lands of the world. From the last few decades to the present, soil surface subsidence has been a major phenomenon of most plains in Iran. The core reason of this phenomena is water extraction from ground water by pumping wells. The study area located on the clayey plain covered by the complex of Aridisols and Vertisols in the east of Yazd city in central Iran with cracks of longitudinal and polygonal shapes. This experiment had been planned to find micro-relief dynamics in a time series after rainfall and drought periods followed by soil surface subsidence and soil cracking. For modeling of soil vertical dynamics and cracking processes, a sampling area was selected with 100 points for surveying with Box Jenkins model. The topography measurements of surveying data showed soil surface height variations from a few millimeters to some centimeters (-14 to +14 mm in a year) with sinusoidal rhythms. Auto Regressive (AR) model could predict the land height variations up to 5 years ahead with high accuracy (3 mm). Based on field surveying, drone imagery data confirmed the temporal forecasting model. In the study area land depressing resulted from minerals degradation into amorphous silicas after soil alkalization. Thereupon the monthly changes of soil surface wetting and drying were major factors for land altitude dynamics, whereas the very deep level of groundwater had no effect on soil surface subsidence. It is suggested that for monitoring of soil surface subsidence and soil cracks over time, the surveying with complementary and drone imagery could be much more appropriate method, which allows predicting temporal soil surface subsidence in local scale.