A novel topographic correction framework for detecting forest disturbance from annual wide-time-window Landsat time series

• Propose a post-processing topographic correction framework for disturbance mapping. • Conduct a comprehensive comparison of pre- and post-processing frameworks. • The new framework enhances forest disturbance mapping accuracy by up to 2.18%. • Quantify errors and uncertainties introduced by the pre-processing framework. Topographic effects in mountainous forested regions disrupt the spectral consistency of remote sensing imagery, hindering accurate forest disturbance detection in Landsat time series acquired over wide time intervals (WTW-LTSs). This study evaluates the necessity of topographic correction for improving forest disturbance detection and proposes a novel post-processing topographic correction framework using the sun-canopy-sensor with C corrections (SCS + C) model. The framework simulates spectral reflectance distortions from illumination variations in uncorrected WTW-LTSs before change detection and employs post-processing to remove the resulting topographic artifacts from detected disturbances. Applied in Yunnan Province, China, the results show that (1) the post-processing framework effectively distinguishes topographic artifacts caused by intra-annual variations, achieving a high accuracy of 81.65 %; (2) by removing topographic artifacts, the post-processing framework significantly enhances forest disturbance detection, improving overall accuracy (OA), user’s accuracy (UA), and producer’s accuracy (PA) by 0.38 %–0.51 %, 1.08 %–1.83 %, and 0.18 %–2.18 %, respectively; (3) the pre-processing framework introduces uncertainties, reducing OA and PA by 0.1 % and 1.93 %–2.99 %, leading to the omission of 14.15 %–16.77 % of disturbances and the false detection of 10.03 %–14.57 % of new disturbances. These findings underscore the importance of eliminating topographic effects in WTW-LTSs for accurate forest disturbance detection. The proposed post-processing framework significantly improves accuracy, particularly in complex terrains, contributing to more reliable disturbance maps. This advancement provides valuable insights for ecological monitoring and supports sustainable forest management for mountainous regions.