In the last section, we explored how Vibrant Planet builds a fine-scale vegetation model of your analysis area. Now, we’ll explore how the platform segments the landscape, making it easier to identify and prioritize areas for treatment.
As you know, landscape resilience or community protection is implemented through smaller treatment units with a specific management action such as variable-density thinning or mastication. Vibrant Planet gives you a head start towards implementation by planning within a landscape delineated into detailed management units.
Identify land with similar characteristics and ownership
As discussed in Section 01, Vibrant Planet uses a Synthetic Canopy Height Model (Synthetic CHM) to map current and specific vegetation structure and height:
Then, to help you identify the best areas to prioritize treatment, the platform groups areas with similar CHM, vegetative, and topographic conditions. These areas are called EcoObjects, and vary in size depending on landscape composition and configuration. Adjacent EcoObjects are grouped into minimum 3-acre sized EcoUnits. The platform overlays these EcoUnits with applicable ownership and land use designations (the Land Resume).
The result is a grouping of homogenous management units with similar biophysical characteristics and administrative jurisdictions that will similarly respond to a natural disturbance and/or mitigation. Each management unit contains raster and vector-based data, including polygon size and location, topography, administrative class, ownership, and designations, canopy cover and height, and disturbance information.
These management units will be the underlying basis for scenario planning and optimization. This is just one example of how Vibrant Planet does the front end work like planning and segmentation for you, helping you get to implementation faster.
Science Corner: EcoObjects
Two separate workflows are used to generate EcObjects: first, “treed” EcObjects are created; then, the remaining portion of the landscape is broken into meaningful segments using information unrelated to vegetation to create “un-treed” EcObjects.The general methodology to generate treed EcObjects is as follows:
- Construct a tree crown adjacency graph
- Create groups of tree crowns by:
- Selecting the tallest tree crown as the seed for a group.
- Adding adjacent tree crowns to the group, as long as the adjacent tree crown is not already part of another group and is within a given percentage of the seed tree crown height.
- Continue adding adjacent tree crowns to the grouped tree crowns until a maximum number of tree crowns is reached, or there are no further adjacent tree crowns to consider.
- For remaining single-tree tree crowns that were not grouped with other tree crowns, create single-tree crown “groups.”
- Generate smoothed, simplified single-polygon objects from the groups of tree crowns.
After this process is complete, some areas within the landscape have been delineated into treed EcObjects (where groups or large individual trees exist), and the remaining parts of the landscape. This remaining area is then delineated into un-treed EcObjects by merging together the vectorized slope, fuel models, roads, and streams layers. The treed EcObjects are erased from the untreed EcObjects, resulting in a continuous EcObject segmentation of the landscape. As a final step, waterbodies are erased from the EcObjects layer.
Now that you understand the background landscape and segmentation data that Vibrant Planet uses to display your analysis area, let’s explore how we determine risk.