Bottom-up hierarchical detection of territorial spatial pattern based on multi-source heterogeneous data

The scientific cognition and methods for territorial space are fundamental to digital territorial governance. This study proposed a bottom-up territorial space cognition system through classifying dominant types, grading principal purposes, zoning major functions, and delineating spatial patterns. At the practical level, this study developed a grid-based method for detecting territorial spatial patterns, with Lingbao city in Henan province as a case study. The method first integrates multi-source heterogeneous datasets to classify single and mixed land use. And the land use functions were evaluated and graded into different levels. A multi-tiered iterative determination process was then devised to find the optimal scale for major functional areas. Multiple attributes of major functional areas were analyzed. The territorial spatial patterns were ultimately delineated by combining major functional areas, natural resource endowments, and projection data. The result showed that mixed land uses were located primarily along main roads or form an outer belt surrounding residential areas. There was a significant difference in the grade of land use functions in both urban and rural areas. The major functional areas demonstrated significant north-south differentiation characteristics, with ecological function areas in the south and agricultural production function areas in the north. The territorial spatial patterns are "one barrier, one corridor, and two areas, one center, three deputies, and four development axes". The theoretical and empirical analysis offered a new perspective on territorial spatial digital governance.