Analysis of image transformation and land use/land cover for temperature trends on landsat imagery

Abstract

The global climate change phenomenon has become a big issue, one aspect of which is change in temperature (temperature trend) in urban areas. Temperature trend can be investigated using a remote sensing data approach through multiple extractions based on transformations such as NDVI and NDBI. This transformation could represent land cover types quantitatively so the relationship between different indices, such as NDVI, NDBI, and the temperature can be seen. Processing shows that the distribution and trend of temperatures in a Landsat 5 TM thermal band for periods 1992 and 2009 has risen by an average temperature of 0.99°C, an indicator of the temperature trend in Yogyakarta city. The results of the relationship between brightness temperature and land use/cover pattern (LUCP) is seen in the results of the transformation of NDVI and NDBI correlated with brightness temperature values. The correlation value for temperature and those two parameters suggests that the NDBI transformation is the most dominant factor affecting the increase in temperature, with a regression coefficient reaching 0.5184, and that larger built up areas will create hotter temperatures.