Urban green space is one of the ecological landscapes coupling the artificial and natural elements of the city, which provides the basic environment for urban biodiversity on the one hand and the urban thermal ecology on the other, which is of great significance to the ecological barrier to build livability and effectively enhance the thermal comfort and happiness of the community. This paper focuses on the impact of urban green space systems on the thermal environment at the community scale based on the three-dimensional non-hydrostatic ENVI-met model, it is centered on “plant-surface-atmosphere” interactions. This paper takes the renewal and optimal transformation of urban green spaces in the neighborhoods in urban renewal actions as a starting point, and use the ENVI-met model to simulate the thermal environment outdoors in the neighborhoods, and the study the quantitative effects of green space transformation on temperature and humidity, wind speed, and habitat thermal comfort at the neighborhood scale in hot summer weather. Meanwhile, we conduct quantitative analysis of habitat thermal comfort before and after green space transformation in the study area based on the Universal Thermal Climate index (UTCI) model, to clarify the effects of green space combination pattern and its structural layout and vegetation configuration on the local thermal environment, and to propose corresponding optimization strategies. The model od UTCI is a comprehensive consideration of multiple factors such as human metabolic heat dissipation and clothing thermal resistance, combined with a multi node thermophysiological model and clothing models. The research results show that the UTCI is more sensitive to meteorological factors than other thermal comfort indices, and can be used for research on high temperature and human health. In order to provide a quantitative basis for the influence of green space layout and structural configuration on outdoor temperature, humidity, wind speed and thermal comfort in urban renewal. This study takes residential areas in Xi’an city as an example, the experimental data mainly include three kinds of data, including high-resolution Quickbird images, field measurement and field survey data of Xi’an city in June 2019. The field measurement data mainly include patch edge length, perimeter, area, location, shape and relative position relationship with the surrounding buildings, surface temperature and atmospheric temperature data at the sampling locations. The field survey data mainly include patch structure, vegetation type, collocation combination, distribution pattern, etc. Finally, it used the three-dimensional non-hydrostatic model ENVI-met to simulation the thermal environment effect of the “central + residential” green space model, and select three typical samples in the district to simulate the thermal environment and its influencing factors before and after the green space system transformation. The main research conclusions are as follows: (1) the results of the comparative analysis between the actual measurement and simulation programs established in the experimental area show that the model simulation has a high accuracy, with the mean variance of temperature and humidity being 0.67℃ and 2.3%, and the correlation coefficients being 0.88 and 0.91, respectively, indicating that the simulation results of ENVI-met are close to the real situation and can better reflect the changes of temperature and humidity; (2) The pattern distribution and vegetation configuration of the residential green space landscape have obvious effects on the temperature and humidity, wind speed, etc. after the renovation, the temperature decreased by 0.02~1.13 ℃, and the humidity increased by 1.6~5.87%. It shows that the appropriate increase of trees can effectively improve the local thermal environment and human thermal comfort in the renovation process for old neighborhoods; (3) The UTCI results show that trees are effective in improving the thermal comfort of residential green areas, in terms of plant mix, the model combining lawn, shrubs and trees has the best thermal comfort, and ventilation corridors can effectively increase airflow exchange in small areas. So, when planning green areas in residential areas, in addition to considering the green area ratio index, it should focus on the influence of green area pattern, tree, shrub and grass combination structure and ventilation corridor, consider the aesthetic and economic aspects of plant scenery, and also consider the improvement effect of plants on the local thermal environment, and through the reasonable configuration of green area plants in vertical space and their structural layout, so as to effectively improve the quality of living environment in residential areas. Therefore, the renovation of the layout, morphological structure, vegetation mix and ventilation corridor of the existing green areas in the district based on the existing site constraints is a feasible method to improve the thermal comfort of human living in the old district at this stage.