Correlation Analysis of Urban Development and Land Surface Temperature Using Google Earth Engine in Sleman Regency, Indonesia

Abstract

The development of urban areas in Sleman Regency occurs continuously along with an increase in population, encourages changes in any kind of land cover to be built-up area to meet the needs of citizen housing. The objectives of this study were to analyse changes in land cover and its correlation with Land Surface Temperature (LST) and to determine the direction of regional development that occurs in Sleman Regency. The methods used in this research were multispatio-temporal data analysis that contained spectral transformations and supervised classification of Maximum Likelihood on Google Earth Engine, statistical analysis of Landsat 8 OLI/TIRS imagery from 2014 to 2019, and accuracy assessments to determine the accuracy of the results. The results showed there were the increase in built-up area by 1713.374 ha from 2014 to 2019 and supported by the increase in population density of 2038 inhabitants/km2. There was an increase in LST in the converted areas from 2014 to 2019 with an estimated increase of 2.73°C. The distribution of built-up area that indicated as the direction of urban area development has a tendency to head north and northeast of Sleman regency, such as to Ngemplak sub-district, Kalasan sub-district, and Berbah sub-district. The correlation between building density and LST showed the correlation coefficient of 0.61 which was considered as a strong correlation and the determination coefficient of 0.38 also regarded as significant, based on the t-Test. The accuracy assessment was done on the land cover map, generating the overall accuracy of 88.42%.