A novel workflow for forest regeneration prediction

Abstract Forest regeneration is crucial for forest adaptability to environmental changes and the maintenance of ecosystem services. However, assessing regeneration across different spatial scales remains a challenge. While modern remote sensing methods improve data collection, they still face technical challenges such as image interpretation, sensor calibration, and data processing. Detailed spatial recording of forest regeneration thus remains time-consuming and costly. This study introduces an innovative method based on point pattern reconstruction, which predicts the spatial distribution of regenerating trees cost-effectively over large areas. The method was tested in two 1-hectare forest areas: one in the Billenhagener Forest near Rostock and one in the Dippoldiswalder Heathland near Malter. In each area, a 0.09-hectare reference plot containing both the overstory and regeneration was analysed using terrestrial laser scanning. A reconstruction method was developed using spatial tree data and correlations between tree attributes to extrapolate the regeneration patterns to the entire area. Spatial analyses showed that the structural properties of regeneration were successfully scaled from the small reference plot to the full study site. The bimodal diameter distribution of regenerated tree species observed in the reference plot was also well reproduced throughout the area. The results confirm that the method presented enables an adequate reconstruction of regeneration based on overstory structure from a small sample plot. It provides an efficient, time- and labour-saving approach for assessing forest regeneration capacity. It has the potential to support sustainable forest management by offering scalable, data-driven insights into regeneration dynamics.