Effect of the thematic resolution of land use data on urban expansion simulations using the CA-Markov model
- PDF / 3,442,058 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 6 Downloads / 169 Views
ORIGINAL PAPER
Effect of the thematic resolution of land use data on urban expansion simulations using the CA-Markov model Longyan Cai 1 & Min Wang 2 Received: 7 January 2020 / Accepted: 13 November 2020 / Published online: 23 November 2020 # Saudi Society for Geosciences 2020
Abstract Variations in the thematic resolution can impact the characterization of environmental heterogeneity and can consequently alter land use change predictions. In the current study, we quantified the impact of the thematic resolution of land use data on cellular automata (CA)-Markov model predictions by running five scenarios with increasing thematic resolutions (2-, 4-, 6-, 9-, and 11class land use maps). The proposed model was able to replicate 83.8% of a historical land use map from 2010, indicating a satisfactory calibration process. Decreasing the thematic resolution reduced the overall prediction accuracy and increased the relative prediction error. This suggests the need to employ two indicators (overall prediction accuracy and relative prediction error) to assess the model performance. Furthermore, the choice of thematic resolution for the CA-Markov model depends on the required output. For example, the overall prediction accuracy and relative prediction error indicate a 4- and 11-class thematic resolution to be optimal for the forecasting of land use change in agricultural and construction land models, respectively. However, a finer thematic resolution is generally expected to minimize uncertainties in land use prediction using the CAMarkov model. Our conclusions provide reference for land use change predictions across the globe via the CA-Markov model. Keywords Thematic resolution . CA-Markov . Logistic regression . Land use change . Model simulation
Introduction In recent years, there has been an increase in the application of geospatial models for the predictions of land use change under various driving forces (e.g., environmental and socio-economic factors) (Al-sharif and Pradhan 2016; Alsharif and Pradhan 2014; Navarro Cerrillo et al. 2020). Such modelling applications inevitably involve the characterization of environmental heterogeneity using thematic maps (Ahmed and Ahmed 2012; Smith et al. 2002). Spatial thematic maps used as input for land use models are typically derived from remotely sensed data classifications or GIS datasets (Alsharif et al. 2015). The thematic resolution (e.g., the number of land types) is decided Responsible Editor: Biswajeet Pradhan * Longyan Cai [email protected] 1
Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China
2
College of the Environment and Ecology, Xiamen University, Xiamen 361102, China
upon by the availability of the data or is subjective to the modeler (Verburg et al. 2011). The characterization of the spatial environmental heterogeneity in the prediction of land use change will vary with the thematic resolution. Previous studies have demonstrated that varying the changing thematic resolution can result in distinct characterizations of configurati
Data Loading...
