Integrating forest inventory plot and GEDI data for forest structure assessment

Forests play a critical role in the global carbon cycle, yet carbon removals in Europe are declining due to increasing wood demand, natural disturbances, and a growing share of aging forests. Sustaining and enhancing forest carbon sinks requires a better understanding of forest structure complexity, which underpins accurate carbon estimates and aligns with emerging EU policy priorities such as identifying old-growth, natural, and even-aged forests. 

Forest inventory surveys provide essential ground-based information for evaluating forest structure complexity. Remote sensing data enables consistent and timely large-scale assessments. Therefore, our objective is to assess the applicability of integrating NFI and GEDI data for characterising forest structure complexity, particularly for distinguishing low and high structural complexity forests. We evaluate the availability of matched NFI plots and high-quality GEDI shots, derive a forest structure complexity measure from integrated variables, and demonstrate a machine learning model trained on NFI-GEDI data to classify forest structure complexity. This study covers Czech Republic, Italy, and Spain, representing temperate, mountainous, and Mediterranean biomes.

We initially identified about 34,000 NFI plots that had a geographic match with almost 90,000 GEDI shots (from a total of ~64,000 NFI plots available and ~200,000,000 GEDI shots in Spain, Italy, and Czech Republic). Rigorous GEDI quality filtering and additional matching criteria reduced the dataset to a total of 2,509 NFI plots and 5,488 corresponding GEDI shots. This is 7% of the NFI plots and 6.5% of the geographically matched GEDI shots. This highlights that data quality requirements reduce the number of matched plots and GEDI shots drastically. Therefore, the data base for assessments of individual countries is low, and a pan-European assessment favourable. 

Forest structure complexity was derived at the plot level using variability in diameter at breast height, tree height, and species richness, combined into an equally weighted structure complexity score. Low variability indicated even-aged, single-species stands, whereas high variability reflected diverse, multi-aged, structurally complex forests. We selected the NFI plots within the lowest and highest 25 % structure complexity score for low and high structural complexity, respectively. 

Training a Support Vector Machine with GEDI data to differentiate between low and high structural complexity, as derived from the NFI-based score, resulted in a model accuracy of 0.81. Restricting the evaluation to the predictions with probabilities > 80% increased the accuracy to 0.94. Applying this model to high-quality GEDI shots in Italy, Czech Republic, and Spain highlights the country-wide occurrence and distribution of low and high structural complex forests. A first assessment indicates that 86%, 65%, and 26% of the forest areas are associated with high structural complex forests in Czech Republic, Italy, and Spain, respectively. 

These results demonstrate the potential of integrating ground-based data with spaceborne-lidar to characterise forest structure complexity. Even simple structure scores and models provide a reliable indication of the structural complexity distribution across Europe. This approach provides a new basis for improving carbon estimates, monitoring structural changes driven by disturbances and other changes, and supporting EU forest-policy targets related to biodiversity, climate resilience, and sustainable forest management.