Climate change is causing great threat to the boreal forests. Modern remote sensing offers tools for near-real-time monitoring of forest resources and allows building high-quality Digital Twins. However, efficient tools for monitoring disturbances are still missing. We propose a methodology that integrates the latest innovations in drones, hyperspectral imaging, and machine learning to implement an efficient and precise framework for forest health monitoring at local scale. In this context, one of the greatest bottlenecks is the requirement for extensive labeled training datasets for the deep learning methods. To solve this problem, we propose a novel simulation based approach. First, we integrate existing Digital Twin datasets, procedural forest modeling, and Monte-Carlo ray tracing to produce a forest simulation model with realistic structures and spectral signatures. Second, to enable training of disturbances caused by biotic or abiotic stresses, we use information of impacts of these factors on optical characteristics of plants and use leaf-optical model simulators to create spectral data of different health statuses. Third, we will produce simulated labeled drone image datasets in different regions in Finland with selected stress factors. Fourth, we will use this data to train machine learning models for vegetation analysis. To evaluate the usefulness of the training method, we will apply the method to analysis of forest health: First, utilizing the simulated data we will optimize the sampling methods and procedures for drone-based detection of the selected biotic and/or abiotic stresses. Second, we will study aspects of improving the methods, such as strategies for transfer learning. Third, we will validate the methods using the existing and new in-situ datasets collected using FGI’s beyond-state-of-the-art drone systems flying above and inside of forests. We believe that the proposed approach will result in a breakthrough in usability of machine learning methods in drone and hyperspectral imaging based forest health and disturbance analysis.