LANDFIRE Remap: Using Top Technology to Improve LF Products


poster of lidar process

LANDFIRE Remap -- First in an ongoing series*

by Julia Deis, LANDFIRE 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. Contact Julia.

Remap is a comprehensive vegetation and fuels mapping effort designed to produce new base maps of the LF product suite. Consistent methodologies and processes - including access to the most current satellite imagery, contemporary data sources and software and hardware technologies - are being combined with broadening program partnerships to create a product that improves upon the updated versions of legacy LF National.

Particular focus is being given to mapping methodologies that span several topical areas, including  LF Reference Database (LFRDB), Satellite Image Compositing, Lifeform modeling, Existing Vegetation Type (EVT) modeling, and Vegetation Structure modeling.

Specifically, Remap is amending the existing EVH and and Existing Vegetation Cover (EVC) legends to represent continuous percent cover and height to represent the landscape structure characteristics and variability at a finer thematic resolution, on which fire fuel modeling is greatly dependent. Continuous structure products are possible by enhancing reference data through incorporating lidar data in combination with the LFRDB.

LF and Lidar

Although there are tens of thousands of LFRDB plots across the United States, structure data gaps remain in several regions. Incorporating lidar observations increases reference data and reduce vegetation structure data gaps. LF is aware that lidar data are not available everywhere and is building a modeling process that attempts to mitigate this issue.

During the remap process LF found that incorporating lidar data in the two prototype areas (Grand Canyon and Northwest) increased the amount of EVC reference data by 310% in the Grand Canyon area, and by 79% in the Northwest area. Further results of LF Remap prototyping in the two study areas confirm that incorporating lidar-derived plots increases reference data considerably, resulting in a more comprehensive reference database that better represents the continuous nature of vegetation structure characteristics than using reference plots alone. Including lidar reference plots has shown higher correlations with validation plots for both EVC and EVH, indicating the inclusion of lidar reference data increases vegetation structure model accuracies.

More information

Click here for access to an informative poster that explains the scope of the project.

The series:
Part Two: Improving Surface Water Mask
Part three: Disturbance Modeling