University of Edinburgh logo British Antarctic Survey logo National Oceanographic Centre logo University of Leeds logo


Automatic Tree Recognition for ecological understanding and satellite mapping

Project summary:

Knowledge of forest structure, biomass and leaf area is essential for understanding processes in ecology, climate studies and meteorology. Terrestrial laser scanning (TLS) has been proven to be the most accurate way to characterise forest structure, biomass and leaf area, but current methods are labour intensive, limiting their application to a few sites around the world. New machine learning techniques should allow a more automated workflow to be developed, enabling information to be generated from many more sites around the world more rapidly, and so increase our understanding of the behaviour and responses of forests.



Terrestrial laser scanning (TLS) is a ground-based technique which produces high-resolution (~1 cm) information on 3D structure. This has been used to characterise forest stands in terms of woody structure (Raumonen et al. 2015) and leafy structure (Hancock et al. 2014). TLS derived products have been shown to be more accurate than traditional methods for estimating stand scale forest metrics, avoiding biases and the saturation suffered by other methods (Calders et al. 2015, Hancock et al. 2014), as well as providing much more detail on their 3D distribution. The detailed 3D information is providing new insights in to forests and will allow better calibration of satellite data for global studies.

However, the software for current methods requires manual intervention in order to tune parameters before analysis ready products can be generated from raw TLS data. The extensive time required for processing the data to a usable product has limited the application of TLS data to a small number of sites and prevented its widespread uptake.

This project aims to use modern computer science, machine learning and signal processing techniques to automate the generation of products from TLS data. The student will collect TLS data in a tropical rainforest in order to test the software developed. The streamlining of the processing will allow TLS to be more widely deployed around the world. Links will be made to the international community working in tropical forest ecology (RAINFOR, Malhi et al 2002) and developing satellite data products (eg. NASA GEDI, Hancock et al 2019), and collaborations formed in order to ensure the work stays at the cutting edge of international efforts.

The supervisory team will include Steven Hancock (Edinburgh Geosciences), who will provide expertise on terrestrial and spaceborne lidar, Bob Fisher (Edinburgh Informatics), who will provide expertise on machine learning of 3D objects, Laura Duncanson (University of Maryland Geography and NASA Goddard), who will provide expertise on lidar and its use in mapping global biomass, and Tim Baker (Leeds, Geography) who will provide expertise on tropical forests and fieldwork techniques.

This PhD is part of the NERC and UK Space Agency funded Centre for Doctoral Training “SENSE”: the Centre for Satellite Data in Environmental Science. SENSE will train 50 PhD students to tackle cross-disciplinary environmental problems by applying the latest data science techniques to satellite data. All our students will receive extensive training on satellite data and AI/Machine Learning, as well as attending a field course on drones, and residential courses hosted by the Satellite Applications Catapult (Harwell), and ESA (Rome). All students will experience extensive training on professional skills, including spending 3 months on an industry placement. See



Calders, K., Newnham, G., Burt, A., Murphy, S., Raumonen, P., Herold, M., Culvenor, D., Avitabile, V., Disney, M., Armston, J. and Kaasalainen, M., 2015. Nondestructive estimates of above‐ground biomass using terrestrial laser scanning. Methods in Ecology and Evolution, 6(2), pp.198-208.

Hancock, S., Essery, R., Reid, T., Carle, J., Baxter, R., Rutter, N. and Huntley, B., 2014. Characterising forest gap fraction with terrestrial lidar and photography: An examination of relative limitations. Agricultural and forest meteorology, 189, pp.105-114.

Hancock, S., Armston, J., Hofton, M., Sun, X., Tang, H., Duncanson, L.I., Kellner, J.R. and Dubayah, R., 2019. The GEDI simulator: a large‐footprint waveform Lidar simulator for calibration and validation of spaceborne missions. Earth and Space Science, 6(2), pp.294-310.

Malhi, Y., Phillips, O.L., Lloyd, J., Baker, T., Wright, J., Almeida, S., Arroyo, L., Frederiksen, T., Grace, J., Higuchi, N. and Killeen, T., 2002. An international network to monitor the structure, composition and dynamics of Amazonian forests (RAINFOR). Journal of Vegetation Science, 13(3), pp.439-450.

Raumonen, P., Åkerblom, M., Kaasalainen, M., Casella, E., Calders, K. and Murphy, S., 2015. Massive-scale tree modelling from TLS data. ISPRS Annals of Photogrammetry, Remote Sensing & Spatial Information Sciences, 2.