Sunjoo Lee

This study aims to classify and analyze forest ecosystem recovery patterns in the regions of Busan and Gyeongsangnam-do, South Korea, that were affected by landslides caused by Typhoon Sanba in 2012. To evaluate both the structural and functional recovery of the forest ecosystems, we employed two complementary indicators: the Vegetation Recovery Rate (VRR), derived from time-series NDVI (Normalized Differenced Vegetation Index) data, and the Shannon Diversity Index, calculated from the 1st and 2nd national forest health management surveys. VRR quantified the recovery of vegetation cover based on NDVI changes before and after the disturbance, while the change in the Shannon Index represented functional recovery based on Shannon Index variation. Each indicator was interpreted using the median value as a relative threshold (VRR: 100.26%, Shannon Index difference: 0.1431), allowing spatial classification into four recovery types (A ~ D). Type A (high VRR and high Shannon Index difference) indicates full structural and functional recovery, while Type D (low VRR and low Shannon Index difference) signifies areas requiring priority ecological restoration. Spatial distribution patterns of recovery types revealed that Type A areas were concentrated in Ulsan and Miryang, whereas Type D areas were predominantly found in western Gyeongnam such as Hadong, Geochang, and Sancheong. This classification framework provides a basis for developing region-specific forest restoration strategies and highlights the necessity of integrating both remote sensing and field-based ecological indicators for comprehensive post-disaster assessment

Taper equation was developed to predict diameters at specific height for Larix kaempferi species in the Central Region of South Korea. The sampled trees that were collected through destructive sampling ranged from 0.60 to 47.90 cm DBH with total height ranging from 2.00 to 33.00 m. The dataset was randomly split into two: 80% for initial model fitting and 20% for model validation. The combined that means 100% dataset was used for final model fitting. Statistics of fit were used as criteria, including the coefficient of determination (R²), standard error of the estimate (SEE), bias (Ē), the absolute mean difference (AMD), and the Akaike information criterion (AIC) values and weight (AICw), with a rank analysis being applied. Initial model fitting utilized a total of nine taper equations, with the three segmented taper equations and another three non-segmented determined for model validation and final model fitting. Kozak02 equation, which had not yet been evaluated for L. kaempferi species, indicated the best performance for the species in the Central Region of South Korea. Upon comparison, the ability of this study to predict diameters at specific height and stem volume was better than the existing taper equation for L. kaempferi in South Korea.