This research aims to tackle challenges related to preprocessing point clouds acquired in forestry contexts, aiming to extract fine-grained characteristics from individual trees. We'll validate and deploy methods for capturing and aligning ultra-high-density 3D data using diverse sensors across different forest types. Additionally, we'll develop tools for characterising trees within large 3D forest scenes and we'll innovate new methods for generating synthetic 3D datasets at the tree level to facilitate the training of deep learning models.
Contact: Robin Hartley, Geospatial Scientist

This research area has been investigating data capture methodologies to assess the optimal ways of collecting 3D data to phenotype trees. In this image, an individual tree was extracted from a UAV laser scanner (ULS) point cloud of a breeding trial. The tree was subsampled as part of an experiment to determine what the minimum point density was to be able to predict DBH, height and tree volume before accuracy of the measurements starts to drop. This information will be used to aid in optimising flight planning for UAVs, to make data capture times and data sizes more efficient.

The research aims to develop methods, using hyperspectral and thermal data, to characterise phenotypic traits related to disease susceptibility and drought tolerance for a series of trials in New Zealand. Combining these data with 3D structural attributes (RA1) will enable the extraction of highly detailed individual phenotypes to be combined with climatic and novel genotypic data (RA3). The application of these methods will be expanded to support cultural phenotyping in complex, indigenous forests (RA4).
Contact: Russell Main, Senior Spatial and Remote Sensing Scientist

Hyperspectral and thermal images captured over a Dothistroma Needle Blight infected radiata pine stand. Both sensors have been shown to be sensitive to stress induced changes in the trees physiology. Severely infected trees appear reddish-brown in the hyperspectral image (left) and exhibit warmer canopy temperatures (pink-yellow) than mildly infected trees (blue-black) in the thermal image (right). The inset graph illustrates the variation in visible near infrared reflectance (VNIR) between mild (green) and severely infected (red) trees. Mapping the disease severity across the stand and correlating it with genotypic information may provide insights into  genetic resistance of different tree genotypes. The two sensors have also been shown to be sensitive to pre-visual effects of disease, which could offer rapid, scalable, and early assessments of forest health, ensuring timely and targeted responses to disease outbreaks.

To address the challenge of climate change we will (1) integrate state-of-the art approaches in tree physiology, genomics, climate data science, remote sensing, and geospatial modelling to characterise key adaptive traits (i.e. drought resistance and vegetative phenology), (2) use an unprecedented data set in radiata pine to relate tree ancestries to climates and adaptive traits and (3) develop climate-based seed zones and deployment guidelines for radiata pine and at least one indigenous species in New Zealand.
Contact: Dejan Firm, Forest Ecology and Management Scientist

High precision dendrometers used for continuously monitoring water deficit and radial growth of tree seedlings exposed to drought.

To create tools that capture the different ways Māori see and value forests, enabling a distinctive Māori forestry identity to express itself with new economic platforms that also serve to offset the carbon footprint of New Zealand.
Contact: Lania Holt, Research Scientist

Vision Mātauranga is a key part of what we will do, how we will do it, and the impact we seek to enable a distinctive Māori forestry identity to express itself with new economic platforms. Our research area of Ngā Mata o Tāne (great forest of many eyes) is Vision Mātauranga centric, and it’s strongly connected to all the other research areas in our programme under the themes of:

Indigenous Innovation by creating Māori algorithms and datasets that will synthesise Māori knowledge with remote sensing applications to, for example, be able to depict the distribution of culturally important species.

Taiao aims to change how we analyse the interaction between genetics and the environment, particularly for indigenous species. This shift will help Māori invest in forests that provide various advantages such as biodiversity, lasting impact, and new opportunities. It will transform landscapes, especially where Māori-owned land is not being fully utilised.

Mātauranga Māori by researching modern applications of traditional forest-based opportunities that not only lead to economic opportunities but can also reinvigorate Māori practices and Māori cultural linkages to forests.