The existing urban construction land allocation mainly follows the guidance of spatial development, which is carried out by using static data and land allocation standards, evaluation and prediction models. There are some phenomena of "regional preference" and "seeking large scale". In the new era of emphasizing the construction of ecological civilization and high-quality development, the implementation of territory spatial planning strategy requires that the allocation method of urban construction land needs to be innovated in time. This project focuses on the inevitable trend that the urban place space will be transformed into the urban flow space under the deep influence of information or intelligent technology in the future, and considers that the spatiotemporal flow and distribution of production, life and ecological elements of different scales will be the key to the scientific allocation of urban construction land. Therefore, the project uses multi-source big data of element flow and multi-disciplinary analysis methods, trying to establish a new method framework from four aspects: evaluation of urban spatial development level, spatio-temporal simulation of urban element flow network and land allocation, identification of urban element functional area and delimitation of development boundary, and dynamic evaluation of urban construction land allocation effect. Firstly, it is found that the evaluation of urban spatial development level needs to comprehensively consider the ability of cities to control, attract and overflow the current regional elements, and construct the evaluation index system from three aspects of spatial development scale, vitality and quality. Secondly, in the process of regional distribution of urban construction land index, we pay attention to using network optimization model and social network analysis method to describe the flow law of regional elements, and find out the cities with different levels and types of combination of "spatial development level and element control" through time-space coupling model. Thirdly, the future expansion direction and boundary shape of cities are determined by the spatiotemporal matching of supply and demand between the flow changes of various internal elements and the functional space layout. It is necessary to identify the functional influence area and flow preference of elements by using models of spatiotemporal coupling, gravity, cellular automata and agent-based system. Finally, the evaluation index of urban construction land allocation effect needs to reflect the requirements and trends of future urban development, such as ecological civilization, high quality, smart governance and integration. These research results can provide reference for the delimitation and implementation of urban development boundary in territory spatial planning.