Use of UAV Multispectral Imagery for Estimating Biologically Active Areas to Support Spatial Planning

The increasing development pressure observed in rural areas leads to progressive land-use transformation and a reduction of biologically active areas (BAA). Maintaining the required share of biologically active areas is an important element of sustainable spatial planning, climate-change adaptation, and ecosystem-based urban and rural land management. However, conventional field-based assessments are time-consuming and may not accurately reflect current land-use conditions. This study evaluated the potential of high-resolution UAV-derived multispectral imagery for estimating biologically active areas on residential parcels located in rural areas under increasing development pressure. Multispectral imagery was used to calculate selected vegetation indices, while Receiver Operating Characteristic (ROC) analysis and the Youden index were applied to determine optimal classification thresholds. The obtained classifications were compared with manually delineated reference data prepared for individual cadastral parcels. Among the analysed vegetation indices, NDVI and RGBVI achieved the highest agreement with the reference data and provided the most reliable estimation of biologically active areas. NDVI achieved the highest pixel-level classification accuracy, whereas RGBVI produced the lowest parcel-level area estimation errors. The results also demonstrated that ROC-based threshold selection reduced subjectivity and improved the repeatability of land-cover classification. Although classification accuracy was influenced by shadows, heterogeneous land cover, and tree canopies, the proposed workflow proved effective for parcel-scale monitoring. The developed approach can support spatial planning by enabling rapid identification of parcels where the actual share of biologically active areas may differ from planning requirements, thereby facilitating environmental monitoring, supporting sustainable spatial planning, and contributing to climate-resilient land management.