Skip to content

Spatial scaling

Spatial scaling and optimisation of the location and configuration of clusters of trees must be robust to support long-term planning for climate change adaptation. However, the suitability of spatial optimisation to generate stable patterns and networks under uncertainty has not yet been evaluated. We will, for the first time, analyse the impact of uncertain input data on the spatial patterns and networks generated by spatial optimsation and probabilistic modelling to inform landscape design that incorporates tree cluster configurations used currently by landowners and exploring new configurations to optimise benefits.

We will combine the findings from the process-based carbon cycling model with a suite of models for nature’s contribution to people (ecosystem services) and predict these at landscape scale using spatial modelling. Initial development of the models will focus on the four farms at the case study sites used for detailed measurements (farm scale), with the analysis extended subsequently to landscapes (landscape scale).

Landscape mapping of key plant and soil properties

For each of the case study sites at farm and landscape scale, we will use existing databases to map the current locations of trees and estimate biomass from allometric methods. The landscape potential to support clusters of trees will be evaluated by modelling habitat suitability for the different species and forest types using empirical species distribution models. For native forest regeneration, we will build on existing experience with suitability mapping to determine areas where spontaneous regeneration may occur, and the most likely forest types to establish at specific sites.

We will use spatial mapping to identify locations with the most potential for increasing soil carbon stocks. Detailed data on key soil properties will be collated from existing datasets and environmental conditions. We will map key soil properties using machine learning applied to a combination of field measurements, existing national datasets, and proximal and remote sensing at farm and landscape scale. We will then use the process-based models to predict spatial changes in biomass and soil carbon stocks at landscape scale based on establishment of trees with specific traits and incorporating edge effects.

Spatial quantification of carbon stocks and co-benefits

We will extend the application of our existing suite of models to quantify the major co-benefits of different types and configurations of trees across rural landscapes. This modelling will include provisioning (e.g. increasing fodder and timber production, shelter) regulating (e.g. erosion control, water quality, pollination), and cultural nature’s contribution to people (e.g. visual amenity, Māori aspirations).

Optimisation scenarios of benefits and trade-offs

We will use a multi-objective optimisation modelling framework to generate optimal landscape templates for spatial prioritisation of clusters of trees in relation to soil properties and plant traits under current and future climate conditions. Outputs from the model will be combined with probabilistic modelling of landowner economic and cultural decisions about tree establishment to deliver probabilistic models of adoption to achieve optimal landscape solutions across multiple co-benefits.

Key contacts