Diagnosis of drivers behind historic change in land use using receiver operating characteristic curve. (2nd January 2018)

• Record Type: Journal Article
• Authors: Manonmani, R.; Vidhya, R.
• Is Part Of: International journal of modelling & simulation. Volume 38:Number 1(2018)
• Page Start: 1
• Page End: 12
• Publication Date: 2018-01-02
• Subjects: Remote sensing -- land use -- receiver operating characteristics -- socio-economic; Mathematical models -- Periodicals; Simulation methods -- Periodicals; Mathematical models; Simulation methods; Periodicals; 003.3
• DOI: 10.1080/02286203.2017.1368843
• Languages: English
• ISSNs: 0228-6203

Abstract

Abstract: Owing to numerous drivers, selection of specific set of drivers behind historic land use is essential to diagnose before giving an input to the land use model. In this paper, logistic regression was used to quantify the relationship of climate and socio-economic with major land use between 1981 and 2011. The Receiver Operating Characteristics(ROC) method was executed to the diagnosing regression coefficient of drivers at different scale and units. The diagnosis analyses revealed that drivers have high ROC value ranges from 0.64 to 0.83. Thus, ROC method can be considered to select optimum sets of driver for land use model. Further, the Agent-based land use/land cover model was used to predict land use maps of 2001 and 2011 and was calibrated and validated with the satellite-derived land use map with an overall accuracy of 87 and 90%. Finally projected for the year 2021, 2031 and 2041 based on historical land use and drivers, this method allows modellers to identify drivers for prediction of land use.