Remap, Landsat, and User Input - Fourth in a Series*
by Julia Deis, LF Operations Lead, Stinger Ghaffarian Technologies (SGT), Inc
Technical Support Services (TSSC), Contractor to the U.S. Geological Survey (USGS)
Earth Resources Observation and Science (EROS) Center
Using Landsat data to identify disturbances on the landscape, the original LF National product suite consisted of base layers mapped circa 2001. The data were updated (i.e., LF 2008, 2010, 2012, and 2014); however, thanks to the free and open Landsat archive, improvements in image compositing algorithms, plus faster, lower-cost computing hardware, many more Landsat images are being incorporated into the LF Remap effort.
LF Remap is not limited to the production techniques of the original LF National and subsequent update production processes. Instead, it is using newly available datasets such as Landsat 8 and airborne lidar and data processing techniques to produce new vegetation, fuels, and fire regime base layers that represent circa 2016 ground conditions.
In preparation for the LF Remap effort, users were asked what they were looking for in the new base map products, and they responded. After considering user feedback, LF is focusing its efforts on the following:
- Reference data – incorporate newest data, eliminate geographic outliers, incorporate disturbance, integrate spectral testing of plots to discard plots that likely have experienced disturbance recently.
- Image compositing –image seamlines caused by date differences between adjacent images (path/rows) were apparent in the base imagery. Solution: use more Landsat scenes along with improved preprocessing logic, resulting in more seasons of near cloud-free imagery and fewer phonological differences. To date, we have used nearly 1/2 million scenes and we're only 2/3 of the way through CONUS.
- Lifeform – better base imagery with fewer seamlines that represents a longer time horizon, combined with lidar data that provides stronger linkages with structure, result in an improved depiction of lifeform.
- Masking – unlimited free access to the Landsat archive allows for the processing of more imagery to better identify specific land cover types, such as water, barren sparse or to create mapping masks to restrict where alpine or riparian /wetland classes can be modeled.
- Existing Vegetation Type (EVT) mapping – more seasons of base imagery to better model vegetation type.
- Vegetation structure - incorporate lidar data, improved accuracy, thematic resolution.
- Disturbance – incorporate recent landscape disturbances into the base map layers for more representative mapping of EVT.
- New Products – LF is redefining products to make information more readily available. The result is six new products: Fuel Disturbance History, Dominant Cover Type, Fuel Vegetation Height, Fuel Vegetation Cover, Fuel Vegetation Type, along with crosswalks to Society of American Foresters/Society for Range Management.
LF Remap is integrating lidar data for structure mapping
Click here for access to an informative poster that explains the scope of the project.
*LF Remap Series:
Part One: Using Top Technology to Improve LF Products
Part Two: Improving Surface Water Mask
Part Three: Disturbance Modeling
Part Four: LF Remap, Landsat, and User Input
