Thermal and energy performance of green roofs received many attention in the past few years. Field or laboratory experiments have been conducted in different regions , and the results show that thermal performance of extensive green roofs depends on multiple factors which mainly can be divided into external and internal ones. External factors refer to climate conditions, including solar radiation, wind velocity, rainfall (irrigation), temperature and humidity. Internal factors include the geometrical and hydrothermal properties of plants and substrate, volumetric water ratio, thermo-physical properties of building structural layer and insulation layer. In order to understand these phenomenon deeply, different models have been developed . Eco-roof is one of the most cited models ,which has been also adopted by Energyplus software. However, because of various configurations and materials involved in local green roofs, many researchers have stressed the importance of further validation. What’s more, some disputes exist in thermal study of green roofs. For example, Liu et al. conducted measurement on two extensive green roofs with different substrate depth . They found that little thermal contribution came from vegetation, and thicker substrate tended to have better thermal performance. However, Orna’s results indicated that cooling effect of bare soil roofs was much weaker than that of any other extensive green roofs, and plants shading played an important role . In terms of energy balance, there are also some divergences. Lazzarin’s data showed that, 25% of total heat gain in summer was consumed by evapotranspiration and 13% by convection and long wave radiation. By contrast, Feng chi’s result indicated that evapotranspiration accounted for 51.5%, the proportion of convection and long wave radiation was 37.8% .Because of various factors affecting the process of hygrothermal transfer, including external and internal ones, further experiments and data are needed for comprehensive understanding of local green roofs.