Coupling artificial intelligence and simulation-optimization for numeric forest management planning
Application period: 11 June – 10 August, 2026
Research project description
This research project analyzes forest inventory data from two time points, approximately 10 years apart, to detect and characterize different harvesting practices, such as clearcuts and thinning, in a large forest landscape of about 200,000 hectares. Simultaneously, this research uses conventional simulation based approaches to project forest development resulting from various management alternatives. Using this database, the research compares the simulated alternatives to the realized management to determine the management regime of the numerous forest owners in the landscape. The research applies deep learning and generative AI approaches to identify spatial drivers of forest management and predict natural processes and harvesting.
The main scientific objective is to advance the understanding of how socioeconomic, demographic, and biophysical factors interact to shape timber harvesting patterns from the scale of individual forest holdings to regions. Ultimately the aim is to better capture the forest owners’ decision-making aspects and ensure that our predictive scenarios reflect real-world preferences and constraints.
Academic background and skills of the applicant
An ideal candidate holds a master’s degree in forestry or data science, computer science, applied mathematics, statistics, geoinformation sciences, or a closely related field. The candidate is expected to have experience in simulation and optimization systems for forest planning or natural resources planning in general; and demonstrated numeric skills and ability to implement computations (preferably in a programming or scripting environment such as R). Earlier hands-on experience with deep learning and generative AI methods is an asset but not necessity. The person should have studies from a scientific field relevant for the role described above and the ability to work in international and multidisciplinary research teams, demonstrated scientific writing, communication and presentation capabilities, and an ability to work independently. The selected candidate must fulfil the language skills requirements of the Doctoral Programme in Science, Forestry and Technology at UEF (see https://www.uef.fi/en/degree-programme/doctoral-programme-in-science-forestry-and-technology (Eligibility and admission criteria)).
Doctoral programme and research group
Doctoral education in the University of Eastern Finland is arranged in seven discipline specific or thematic doctoral programmes. This research project will be located in the Doctoral Programme in Science, Forestry and Technology, and the submitting department is the School of Forest Sciences.
The doctoral candidate will join the research group of forest planning led by Professor Jari Vauhkonen. This research group is a multidisciplinary group focused on the use of forest inventory data to support multi-objective forest planning and decision analysis. The group’s research has strong emphases on advanced data analytics and modeling techniques, understanding of social and behavioral dimensions, and the heterogeneity of forest owners in the national and international forest policy context, which are studied in a number of recent research projects funded by the Research Council of Finland for example.
Partners / Secondments
Swedish University of Agricultural Sciences, Department of Forest Resource Management, hosted by Prof. Karin Öhman
Other interdisciplinary, international and/or intersectoral collaboration
Intersectoral and interdisciplinary collaboration is strengthened through DP-FOBI’s three summer schools involving academic supervisors, non-academic partners, and other stakeholders, enabling hands-on interaction and knowledge exchange.