| Popis: |
Accurate extraction of urban built-up areas and understanding their expansion dynamics are crucial for urban planning and sustainable development. Traditional single-source remote sensing methods often have low accuracy in extraction. Our study addressed this limitation by integrating multi-source data, including Landsat land surface temperature (LST), NPP-VIIRS nighttime light data, and POI data. Through comparisons of mutation detection, reference comparison, and a multi-source comprehensive index method, we identified the most effective approach for urban built-up area extraction. Additionally, we developed an innovative spatial model based on morphological structural elements and introduced a new corridor expansion to supplement the traditional urban expansion patterns of edge, infilling, and outlying expansion. This new model enhanced the understanding of urban landscape connectivity and provided a more comprehensive analysis of urban growth processes. The results showed that the comprehensive index method produced the most accurate extraction, with overall accuracies exceeding 89% and Kappa coefficients above 0.80 across all years. The built-up area and expansion indicators displayed an increasing trend, with urban expansion extending northwest from 2013 to 2020 and then shifting southward from 2020 to 2023. The spatial pattern of urban sprawl was characterized by edge expansion, followed by corridor expansion (except between 2017 and 2020), with a trend toward increasing outlying types and fewer infilling expansions. These findings indicated that Hefei’s main urban area had reached a relatively mature stage, with built-up areas beginning to expand into surrounding towns; future efforts should focus on strengthening corridor expansion. This study filled a critical gap in accurately extracting built-up areas and analyzing spatial expansion, providing a scientific basis for urban land management, policy-making, and sustainable development. |
Full Text from ScienceDirect