This study attempts to develop a functional relationship between air temperature and urban green space using Landsat data. It also aims to estimate the contributions of various forcings, namely, urban green space, population density, urban area, and automobile densities on urban heat island. Subsequently, the impact of urban heat island on temperature humidity index will be assessed quantitatively, followed by surface energy budget analysis of Jabotabek area. Air temperature series are derived from Landsat data, including the NDVI which is used as the bases in generating urban green space of the study area. Principal Component Analysis is utilized in order to establish the relative importance of forcing variables on urban heat island; in order to simplify the structure of factor loadings a varimax rotation is carried out. It is found that air temperature and urban green space for the study area is best represented by a nonlinear equation when a maximum coefficient determination (R2adj) and a minimum standard deviation (S) are to be fulfilled. A 10 to 50% reduction in urban green space would bring air temperature to raise between 0,2 to 1,8 oC. It is interesting to note that this study reveals the same percentage increase in urban green space would only lower the temperature by 0,1 to 0,5 oC. Automobile density is found to be the most important forcing for urban heat island in Jakarta; on the other hand, urban green space is the most dominant forcing in Tangerang and Bekasi. Surface energy budget analysis indicated that the an increase of 1,0 oC in urban heat island would result in additional 80,3 and 69,0 Wm-2 sensible heat fluxes to submedium (G) and to air, respectively.