Incorporating neighborhood effect into land loss modeling in the Lower Mississippi River Delta

Authors: Heng Cai*, Texas A&M University, Nina Lam, Louisiana State University, Lei Zou, Texas A&M University
Topics: Coupled Human and Natural Systems, Geographic Information Science and Systems, Spatial Analysis & Modeling
Keywords: disaster resilience, land loss, spatial modeling
Session Type: Virtual Paper
Day: 4/10/2021
Start / End Time: 4:40 PM / 5:55 PM
Room: Virtual 41
Presentation File: No File Uploaded


The Mississippi River Delta (MRD) is one of the most environmentally threatened areas in the United States. The area has been suffering substantial land losses during the past decades. Using a coupled-natural and human (CNH) approach can help uncover the complex mechanism of land loss in the MRD, and provide a more accurate spatiotemporal projection of land loss patterns and probability. However, the neighborhood effect has seldom been considered in complex natural-human system modeling analysis. This case study examined the use of a semi-variogram in detecting the appropriate neighborhood size of the variables involved in land loss modeling. The goal is to find out the best combination of variables and their neighborhood sizes that best explain the variation of land loss patterns in the study area. Regression results between the model with and the model without neighborhood variables are compared. The results show that land loss modeling improved by incorporating neighborhood effects. This study uncovers the complex mechanism of land loss and provides a more accurate spatiotemporal projection of land loss patterns and probability. Knowledge gained from this study provides insight into the optimal mitigation strategies of land loss prevention and restoration and help build long-term sustainability in the Lower Mississippi River Delta.

Abstract Information

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

To access contact information login