Forest growth rate assessment using aerial and satellite imagery for aerodrome hazard monitoring

Many European aerodromes are located in areas with forest vegetation reaching heights of 40–50 m or more. To comply with ICAO Annex 14 and 15 standards, trees exceeding specified heights must be lowered or removed as they become obstacles to flight operations. Monitoring irregular forest growth—driven by species diversity, stand age, and climate change—requires a synergy of expertise in geodesy, geoinformatics, and air navigation, combined with advanced remote sensing (RS) data. High-resolution aerial photography and satellite imagery (30–50 cm) are most effective for creating metrically accurate 3D models to monitor these obstacles; however, the high cost of regular data acquisition limits their frequent use. Therefore, this study aims to evaluate the feasibility of using predictive growth modeling as a cost-effective alternative for forecasting tree heights and identifying potential obstacles. The key method is evaluating the accuracy of traditional growth models in comparison with empirical data derived from multi-temporal RS measurements. The study examines coniferous and deciduous species (Pinus sylvestris, Picea abies, Betula pendula, Populus tremula, etc.) across three forest-vegetation subzones in Belarus (51°N–56°N, 23°E–32°E). We used aerial and satellite stereoscopic imagery with 30–50 cm resolution acquired by sensors ADS-100 (Leica Geosystems), Vexcel (UtraCam), and Pleiades-1B (AIRBUS) between 2017 and 2023 for different test polygons. The stereophotogrammetric 3D models were generated with precise (5-10 cm) geodetic reference and measured at the same tree height. Growth patterns were analyzed with two sets of regression equations: theoretical models based on standard forest inventory (taxation) tables (2005) and empirical data based on measurements from multi-temporal 3D models. Comparison of the regression models revealed significant discrepancies between empirical RS data and theoretical inventory tables. Specifically, for Pinus sylvestris aged 6-10 years, the mean height deviation reached 3 m over the 6-year period. The regression coefficients indicate that traditional models slightly underestimate current growth dynamics, likely due to local climatic shifts. The trees in the south of Belarus became significantly higher over 20 years, 42 m instead of 37 m maximum, as defined in the 2005 taxation tables. Cartographic models based on the growth regression equations identify airfield obstacles with high reliability, accounting for local climate conditions. This approach significantly improves land-use management in airfield areas, preventing excessive logging that often results from a lack of up-to-date and precise remote sensing data.