Correlating landslide susceptibility with landslide risk derived from statistical modelling of environmental factors using a GIS approach

An understanding and awareness of past landslide events are crucial to limit negative impacts such as injury and, in dire situations, fatalities. Landslides vary in rock movement and what triggers them. Some examples of these factors include slope characteristics, sediment age, rock type, precipitation, seismic activity, etc. This project focuses on identifying environmental and geomorphological characteristics of landslides by mapping and analyzing publicly available data. Slope angle may be the major risk factor in landslide hazard, but statistical tests show that locational context and human influences must also be taken into consideration. I compare three factors of safety models of varying dimensions to three statistical regression models to create a ranking of the landslide factors. This ranking will be used to make a weighted landslide susceptibility map. I apply this method to three watersheds located in Washington state, Northern India, and Japan. However, prolonged precipitation plays a bigger role in Washington state because of weak glacial sediments. The 3-dimensional factor of safety model has more fine-grained variation compared to the 1-dimensional model, which may not be reasonable for wide-scaled statistical analysis. With recent studies stating that the frequency and intensity of natural disasters will intensify due to anthropogenic climate change, real-time data is needed to present to the appropriate communities that will be most affected by these disasters.