Nowadays, global warming has become more interested for scientist, because the global surface temperature has been increased since last century. More than fifty percent of people are living in cities, in this regard, urbanization has become a key factor for global warming. The urban heat island (UHI) refers to the event of higher atmospheric and surface temperatures occurring in cities than in the surrounding rural areas due to urbanization. The annual average air temperature of urban area with almost one million people can be one to three degree warmer than its surroundings. This phenomena can affect societies by increasing summertime, air pollution, air conditioning costs, heat related illness, greenhouse gas emissions and water quality.
Tehran, a capital city of Iran is case study of this research. Additionally, Tehran is one of megacities of the world. A megacity is usually defined as a metropolitan area with a total population in excess of ten million people. Due to rapid urbanization progress that has resulted in significant UHI effect in this area. Furthermore, Tehran houses to almost twenty percent of Iranian people. In this study, new launched Landsat series (Landsat 8) was used for monitoring UHI and retrieving the brightness temperatures and land use/cover types. The Landsat 8 carries two kind of sensors: The Operational Land Imager (OLI) sensor has former Landsat bands, with three new bands: a deep blue band for coastal/aerosol studies (band 1), a shortwave infrared band for cirrus detection (band 9), and a Quality Assessment band. The Thermal Infrared Sensor (TIRS) sensor provides two high resolution (near to 30 meters) thermal bands (band 10, 11). These sensors both use corrected signal-to-noise (SNR) radiometric quantized over a 12-bit. Corrected SNR performance cause better determination of land cover type. Moreover, Landsat 8 images incorporate two valuable thermal bands in 10.9 µm and 12.0 µm. These two thermal bands improve estimation of UHI by incorporating split-window methods.
Recently, quantitative models for urban thermal environment and related factors have been studied, for example, the relation between UHI and land cover structure and established corresponding regression equation. Similar works have been done and models of the relation between the surface temperature and various vegetation Indices have been established. In order to monitor the relationship between UHI and land cover indices, this paper tried to employ a quantitative approach for exploring the relationship land surface temperature and common land cover indices and select suitable indices by incorporating supervised Feature Selection (FS) procedures, including the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Soil Adjusted Vegetation Index (SAVI), Normalized Difference Water Index (NDWI) in two definition, Normalized Difference Bareness Index (NDBaI), Normalized Difference Build-up Index (NDBI), Modified Normalized Difference Water Index (MNDWI), Bare Soil Index (BI), Urban Index (UI), Index-based Built-Up Index (IBI) and Enhanced Built-Up and Bareness Index (EBBI). In this regards, the objectives of this research are to develop a non-linear analysis model for urban thermal environment by employing Support Vector Regression (SVR) method and Multivariate Regression (MR) algorithms. In addition, providing the hazard map for Tehran city is also one of the byproducts of proposed methods for managing and mitigating UHI effects.