Generating A New Montana Land Cover Framework
University Of Montana Spatial Analysis Lab
Base Raster
Landfire Remap was downloaded as two tiles which were merged, clipped to the State boundary, and projected to Montana StatePlane.
Composition Updates
Pixel modifications to improve the base raster: Transportation, Agriculture, Fires, and Classification Accuracy
Transportation
- Landfire Remap does not differentiate among types of roads and is lacking railroad;
- Road pixels stem from a classified satellite image, resulting in a pixelated effect instead of continuous lines.
The Montana Spatial Data Infrastructure (MSDI) Transportation dataset offers an alternative source of road data, with categories such as interstate, major roads, other roads and railroad, in a vector format that can be rasterized as uninterrupted lines.
Compared with the original road pixels, this update brings added information and increased accuracy. Use the slider to compare before and after!
Agriculture
Landfire agriculture pixels come from the USDA Cropland Data Layer , a crop-specific raster produced using Landsat 8 satellite imagery.
In Montana, the Department of Revenue generates each year a vector layer (Final Land Unit, FLU). Using photo interpretation and feedback from land owners, DoR Technicians map private agricultural land into one of six uses: fallow, hay, grazing, irrigated, continuously cropped and forest. To update Landfire, we used irrigated and continuously cropped (grouped) and hay.
A comparison of Landfire agricultural pixels (left panel) and FLU 2020 polygons (right panel) shows missing agricultural fields, both irrigated (red outline) and dryland hay (yellow outline).
The updated raster now has better roads and more complete agriculture.
Fires
Fires are one of the most important causes of landscape change in Montana: in 2017 alone, over one million acres burned.
The first fire update consisted of fixing misclassified Recently Burned, Herb and Grass Cover pixels that should be Recently Burned, Tree cover (as this 2011 NAIP image clearly shows).
By combining Landfire Remap with the Land Change Monitoring, Assessment, and Projection (LCMAP) 2011 Montana raster, "Recently Burned - Herb and Grass Cover" pixels overlapping "Tree Cover" pixels were reclassified to "Recently Burned - Tree Cover".
Because Landfire Remap is based on 2016 imagery, another update involved "regrowing" Recently Burned - Herb and Grass Cover to the grassland ecological system most likely to be found within a given area. As shrub and forested systems take longer to regenerate, they were not modified.
But the most important fire update consisted in adding areas that burned since 2016 (the year of Landsat imagery used to generate Landfire Remap) - red polygons on this map. Let's walk through an example, the 2017 Rice Ridge fire.
Recent (2017-2019) fire perimeters and intensity rasters were downloaded from Monitoring Trends in Burn Severity (MTBS) , a multiagency program designed to consistently map the burn severity and perimeters of fires across all lands of the United States from 1984 to 2019 (and onward).
MTBS data do not include fires smaller than 1,000 acres, so the last fire update consisted of "burning in" small 2017-2019 perimeters from the National Interagency Fire Center .
Classification Accuracy
To improve classification accuracy, classes with just a few pixels (such as bur oak savanna or basin playa) were replaced with neighboring, ecologically similar classes.
In other cases, a new class was created by grouping several. For example, Great Plains Riparian Grassland, Shrubland and Forest were merged into a single "Great Plains Riparian" class after comparison with NAIP imagery revealed many misclassified pixels.
Tabular updates
Attribute Modifications
To make visual comparison between the Montana Landcover series (2010-2017) and Montana Landcover (MTLC) 2023 easier, the Landfire Remap color palette (red, green, blue attributes) was replaced with that of Montana Landcover. New classes (found in the former but not the latter) kept their Landfire color scheme.
A series of modifications was performed to remove some of the original Landfire Remap attributes, while adding others based on a crosswalk with the Montana Landcover series.
Two of the attributes have live links: SNAME_url will take you to an ecological systems description from the Montana Natural Heritage Program Field Guide; BPS_url to the LANDFIRE Biophysical Setting Model zone report where the EVT class/system is most dominant.
After downloading Montana Landcover (MTLC) 2023 from MSDI , you can open the metadata for a more in-depth description of the various updates presented in this document.
