Sensitivity analysis and accuracy assessment of simulated land-use and land-cover changes using cellular automata: the case of Bogota’s urban wetland

Keywords: urban wetlands, Cellular Automata (CA), sensitivity analysis, Land use/land cover change, FLUS model
Session Type: Virtual Poster Abstract
Day: Monday
Session Start / End Time: 2/28/2022 11:20 AM (Eastern Time (US & Canada)) - 2/28/2022 12:40 PM (Eastern Time (US & Canada))
Room: Virtual 33

Abstract

This study uses Cellular automata (CA) as a tool for modelling and simulating urban wetlands changes in the capital Bogota, Colombia. The study applied the coupled Markov-FLUS (Future Land Use Simulation model) model to simulate and analyze the land- use and land-cover changes in Bogota, Colombia over the course of the next 20 years. The analysis is executed using 3 analogic images to extract the land use/land cover for three years: 1998, 2004 and 2010. Artificial Neuronal Network FLUS module is used for calculating the relationships between the land uses and the associated drivers, and to estimate the probability of occurrence of each land use. Whenever a CA is used to model and simulate, sensitivity analysis becomes a crucial step in CA modeling to better understand the influence of parameters’ changes in the simulation outcomes. The sensitivity analysis is conducted by varying the neighborhood sizes between 3 × 3, 5 × 5 and 7 × 7, and the spatial resolution of the model inputs from 5 and to 30 meters. To assess how well the model performed, kappa coefficients, receiver operating curve (ROC), landscape metrics, figure of merit (FOM), and quantity and allocation disagreement are calculated. The results show that the model performed well, and the accuracy of the outputs improves with a 3 × 3 neighborhood size and five meters spatial resolution. This study provides insights about the different components that must be considered to evaluate sensitivity of CA models in the simulation of wetlands landscape spatial-temporal evolution.