Predicting Land Cover Using a GIS-Based Markov Chain and Sea Level Inundation for a Coastal Area

Abstract

New Hanover County, North Carolina, has been experiencing rapid population growth and is expected to continue this growth, leading to increased land use and development in the area. The county is also threatened by sea level rise (SLR) and its effects because of its coastal location and frequent occurrences of major storms and hurricanes. This study used a land change modeler to map the land cover change throughout the county over a period of 20 years, and predicted land cover distribution in the area in the years 2030 and 2050. Statistics revealed that the developed land in the area increased by 85 km2 between 2000 and 2010, and by 60 km2 between 2010 and 2020. Such land is predicted to increase by another 73 km2 by 2030, and 63 km2 by 2050. This increase in development is expected to occur mainly in the central area of the county and along the barrier islands. Modeling of SLR illustrated that the northwestern part of New Hanover County along the Cape Fear River, as well as the beach towns located on the barrier islands, are estimated be the most affected locations. Results indicate that sections of major highways throughout the county, including I-140 near downtown Wilmington and US-421 in Carolina Beach, may be inundated by SLR, which might delay residents during mandatory evacuations for emergency situations such as hurricanes. Some routes may be unusable, leading to traffic congestion on other routes, which may impede some residents from reaching safety before the emergency. Wrightsville Beach and Carolina Beach are estimated to have the highest levels of inundation, with 71.17% and 40.58% of their land being inundated under the most extreme SLR scenario of 3 m, respectively. The use of the present research approach may provide a practical, quick, and low-cost method in modeling rapidly growing urban areas along the eastern United States coastline and locating areas at potential risk of future SLR inundation.

Keywords

• land cover
• neural network
• land change modeler
• minimum noise fraction
• digital elevation model
• surface gradient