Spatio-Temporal Modeling of Soil Strength Relative to Mobility

The military and civilian sectors use mobility models to plan routes, identify choke points, and assist in procurement of off-road wheeled and tracked vehicles. Off-road mobility models predict vehicle performance based on terrain data and scenario conditions. Terrain factors and scenario conditions (i.e., visibility, soil strength, snow depth) change temporally and spatially. Prediction of how quickly mobility related terrain factors change with the climate can have a major influence on decisions. This study defines temporal and spatial relationships for Newton MS such that a forecast of changing soil strength conditions is made.

A modified form of Kriging is investigated for determining spatial variability of soil strength and an extension of the Box Jenkins method is used to account for non-stationary temporal changes in the soil strength. This is an extension of soil strength forecasting as tool for predicting vehicle mobility several months into the future and generating reliable spatial information when a sparse network of data exists. A forecast of soil strength variations around a weather station is made given past weather data and sparse information on elevation and soil strength.