Document Type

Technical Report

Publication Date



Forests and forestry, Sage grouse -- Habitat -- Conservation, Sage grouse -- Oregon -- Management


We used a predictive model to map canopy cover of vegetation over seven feet in height ("tall woody vegetation") at 30-meter resolution over nearly 29 million acres within and adjacent to the range of the greater sage-grouse in Oregon (Figure 1). Texture measures computed at various resolutions from color-infrared aerial photography provided the main source of predictor data used to produce the map. Canopy cover was treated as a categorical variable using six cover classes: absent (cover class C0), present at less than 4% (C1), 4 – 10% (C2), 10 – 20% (C3), 20 – 50% (C4), and 50% and over (C5). The map is referenced to conditions in the years 2011 and 2012.

Although the specific target of the mapping was western juniper (Juniperus occidentalis), our reference data did not permit separating juniper from other tall woody vegetation during the predictive modeling process. The majority of the tall woody vegetation within the project area is western juniper. However, in high elevation regions, riparian, wetland, and residential areas, other vegetation is occasionally represented.

The methodology discussed here produces raw modeled data. It is recommended that prior to use in most applications this raw tree cover product be additionally filtered or masked to eliminate false detections which often occur adjacent to agricultural areas and roads. For species-specific applications, an additional modeling phase is necessary to either eliminate tree cover detections likely to be species other than the target, or to model species importance values associated with each tree occurrence.

Persistent Identifier