Prediction of land use/cover change in Connecticut River Basin using a geographically weighted logistic regression-Markov chain model

Authors: Hui Wang*, University of Connecticut
Topics: Human-Environment Geography, Land Use and Land Cover Change, Water Resources and Hydrology
Keywords: land use/cover change, geographically weighted logistic regression, Connecticut River Basin, human activities, Markov chain
Session Type: Paper
Day: 4/11/2018
Start / End Time: 8:00 AM / 9:40 AM
Room: Maurepas, Sheraton, 3rd Floor
Presentation File: No File Uploaded

The Connecticut River Basin (CRB) experienced significant increasing runoff in the past 50 years. Previous studies show that land use/cover change (LUCC) is one of the main driving factors. It is therefore important to understand how land use/cover will be altered and which explanatory factors, such as population density and per capita income, will weigh most heavily in the future. The logistic regression (LR) technique has been widely used in LUCC prediction models. However, few studies considered spatial non-stationarity both in land use/cover categories and explanatory factors. This study explores spatial variation in the relations between explanatory factors and land use/cover categories, and then predicts future LUCC in the CRB using a geographically weighted logistic regression-Markov chain (GWLR-MC) model. Comparisons between LR model and GWLR model suggest that the latter improves the model’s goodness of fit by considering the spatial correlation of model variables, thus improving the skill of prediction of LUCC. Our results also provide insights into quantitatively assessing the contribution of various explanatory factors and help further our understanding of how runoff will be changed by human activities in the CRB.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login