Following the initial 2019 field sampling season and subsequent data analysis, scientists with the Colorado Natural Heritage Program (CNHP) have produced a scientifically robust Study Design that has been thoroughly reviewed by WBI and the Science Team.
The Study Design describes in detail the Technical Methods for Field Study (outlined here) as well as the Technical Methods for Extrapolation Analysis and Project Deliverables, outlined below.
By taking into account metrics that the Science Team have communicated would be useful in furthering their missions, the Study Design guarantees that each organization represented by the Science Team will be well served by the science produced by the Study.
The intent of our young organization is to set up and orchestrate the Study, and then hand off the results and deliverables to long standing organizations who will make use of them. To this end we are a collaborative agent, producing a Study that serves other entities’ needs.
Technical Methods for Extrapolation Analysis
Following the steps taken by CNHP scientists to define sample plots as either Excellent, Good (mid-range), or Marginal quality ungulate habitat, the next step is to assign polygons that indicate the extent of those patches of habitat. CNHP scientists use Google Earth to define these polygons.
In the example below, the sample plots are represented by the dots and the scientists at CNHP have digitally drawn boundaries that define distinct patches of similar habitat:
The classified polygons are then used to “train” an algorithm that extrapolates the ground-truthed data across the rest of the Roaring Fork Watershed using remote sensing imagery. This is a “machine-learning” algorithm, which means that the margin of error lessens as more data is added and evaluated. Multiple well-performing models are them combined based on their level of agreement to produce a 10m resolution map of ungulate habitat quality across the entire Roaring Fork Watershed.
The main component currently missing from the model is higher elevation sampling sites representing deer and elk summer forage. The CNHP scientists plan to collect data from these higher elevation points during the summer 2020 field season. They will then use this data to create additional training polygons to further refine the model.
Study Deliverables
The Study Design document also outlines the anticipated deliverables of the Study, which include the following:
Written report that describes study methods, results, suggested conservation and restoration strategies, and supporting maps/graphics, in digital format.
Maps
a. habitat suitability model for each focal species
b. habitat connectivity model for each focal species
c. field sites sampled
d. distribution of habitat patches categorized by relative quality (excellent, moderate, marginal) for each focal species
e. connectivity pinch points across the watershed
f. geographic areas of high, medium, and low priority for conservation of high quality habitats
g. geographic areas of high, medium, and low priority for restoration of impaired habitats
h. other maps as determined through ongoing collaboration with the Science Team through 2020
Spatial layers used in creation of each map, with federally compliant metadata
The rigorous peer review of the Study Design by the Science Team will help pave the way for eventual use and application of the Study deliverables by conservation and land management entities. Long term, we intend that our impact will be reflected in increases in elk, mule deer, and bighorn sheep populations in our Watershed.