Abstract
Benefits of a multiple‐solution approach in land change models
Highlights
INTRODUCTION AND PREVIOUS WORKLand change (LC) science seeks to understand the dynamics of land cover and land use change (Turner, Lambin, & Reenberg, 2007)
Several methods have been proposed in the literature to calibrate spatial LC models, including a visual test (e.g., Clarke, Hoppen, & Gaydos, 1997; Ward, Murray, & Phinn, 2000), multi‐criteria evaluation (e.g., Mahiny & Clarke, 2012), reusing param‐ eters from other studies (e.g., Mustafa, Saadi, Cools, & Teller, 2015), statistical analysis (e.g., García, Santé, Boullón, & Crecente, 2013), machine learning (e.g., Mileva, Suzana, Miloš, & Branislav, 2015), artificial neural networks (e.g., Basse, Omrani, Charif, Gerber, & Bódis, 2014; Pijanowski et al, 2014), and search algorithms for optimization such as genetic algorithms (e.g., Mustafa, Heppenstall et al, 2018), particle swarm optimization (e.g., Feng, Liu, Tong, Liu, & Deng, 2011), and a combination of various methods (e.g., Mustafa, Cools, Saadi, & Teller, 2017)
The model stopped working at generation 117 in the case of crowding niching genetic algorithm (CNGA) and at generation 109 in the case of Genetic algorithms (GAs)
Summary
Land change (LC) science seeks to understand the dynamics of land cover and land use change (Turner, Lambin, & Reenberg, 2007). At each time step, representing one year, the model changes the non‐urban cells with high transition potential P until they meet the required amount This model structure follows the work of Feng et al (2011) and Mustafa, Heppenstall et al (2018). All parameters are calibrated through a CNGA in order to consider multiple optimal solutions. The key to searching for multiple optimal solutions is using the crowding niching method to produce the new generation. Ci randomly replaces one of the poorer‐fitted parents in the circle This step is a self‐adaptive niche method as the algorithm dynamically selects the crowding space range for the survival by competence (Chen et al, 2014). Where FSRk (0 ≤ FSRk ≤ 100) is the fuzziness similarity rate for class k; Iikd is 1 if cell ik in the simulated map at zone d (0 ≤ d ≤ 4) has similar land use class to one cell at zone d in the observed map, otherwise it is 0; Xk,sim equals the change amount of class k in the simulated map; and Xk,actul equals the change amount of class k in the observed map
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
