LANDFIRE 2016 Remap

LANDFIRE (LF) 2016 Remap is the most comprehensive mapping effort in LF's history. This years-long effort included updating the base maps across the original data product suite to create LF's second base map product suite consisting of 28 vegetation and fuels layers. LF Remap products reflect circa 2016 ground conditions.

The initial base map (LF National) product suite was developed using Landsat imagery circa 2001. The product suite was updated regularly using vegetation and fuel transition rulesets to more accurately represent current conditions and account for landscape disturbances (LF 2008, 2010, 2012, and 2014). However, with a base map more than 15 years old, LF decided it was time to remap the landscape.

Designed to produce vegetation and fuels data that inform wildland fire and ecological decision systems, LF Remap processes followed consistent methodologies to incorporate current satellite imagery, contemporary data sources, and the latest software and hardware technologies. The resulting LF Remap data products offer significant improvements to all previous LF versions.

Advances in image compositing, satellite image tiling algorithms, plus faster computing hardware ensures LF vegetation and fuels products remain relevant. To achieve this, LF Remap leveraged the Landsat archive, lidar data, and over 1 million ground control plots from a variety of partners and collaborators compiled into the LF Reference Database (LFRDB), to characterize landscape characteristics for more than 9 billion 30-meter pixels across CONUS. Learn more about contributing plot data for future mapping efforts.

Starting with LF 2016 Remap for Alaska (AK), LF products will include a 90-kilometer buffer along the borders of neighboring countries. For LF Remap AK that includes 1,538 miles of border with Canada.

All LF data product categories have been refined as part of the remap effort:   Reference  •  Disturbance  •  Vegetation  •  Fuel  •  Fire Regimes  •  Topographic

'NEW' - Capable Fuels - LF Remap enhancements to fuels products in disturbed areas

A discrepancy in Time Since Disturbance (TSD) 1 (0-1 years) for surface and canopy fuels in disturbed areas existed in previous versions of LF Update products. TSD was not truly representative of the release date. For example, fuel layers created from 2012 disturbance data were released in 2014, meaning TSD lagged two years behind and users had to adjust accordingly. This issue has been corrected in LF Remap fuel products (Fuel Disturbance, Canopy Height, Canopy Cover/Base Density/Base Height, Fire Behavior Fuel Model 13 and 40, Fuel Vegetation Cover/Type/Height, and Fuel Characteristic Classification System) by synchronizing TSDs for surface and canopy fuels in disturbed areas.

What this means is LF Remap fuels products in disturbed areas have been revised to expected 2019 or 2020 vegetation conditions for disturbances that occurred between 2009-2016 or 2010-2016, respectively, making the fuels products 2019 or 2020 capable.

The capable fuels functionality is able to calculate TSD assignments for disturbed areas using an "effective year." Utilizing fuel model transition rulesets, fuels within disturbed areas over the 10-yr period prior are transitioned to the capable year. Therefore, when fire behavior modelers encounter a disturbance (that occurred within the last 10 years) on their landscape, the fuels within the disturbance should reflect growth after the disturbance and up until the capable year, rather than a few years before.

This improves performance of fire behavior modeling and reduces the need for local LF users to update the vegetation and fuel conditions to represent current conditions.

As a result, LF Remap capable fuels layers better represent active and potential wildfire behavior on the current landscape.

LF Remap Changes and Improvements

The LF team at the USGS Earth Resources Observation and Science (EROS) Center used the latest data and processing techniques to produce new vegetation, fuels, and fire regime base layers for the LF Remap data product suite, representing a circa 2016 ground condition.

LF Remap improvements include:

• Unlimited, free access to the Landsat archive: Hundreds of thousands of Landsat 7 and 8 scenes were processed using EROS High Performance Computing resources, resulting in..
  • Better Base Imagery for vegetation mapping
    • Reduced clouds - Image composites were created using the best quality pixels from numerous images for each season to create cloud-free/near-cloud-free imagery.
  • Fewer Seamlines – By using input imagery from a longer time horizon, along with improved preprocessing logic, and hands-on post-processing, the data retain fewer phenological differences within and between imagery tiles.
  • Better map masks – Masks of specific land cover types such as water (Integration of Landsat Essential Climate Variables, i.e. Dynamic Surface Water Extent), barren /sparse that are important to fire behavior were created to exclude these areas from vegetation and fuels layers.
  • Better stratification for vegetation type modeling – This includes alpine or riparian/wetland classes which restrict where certain vegetation types can be identified.
• Vegetation Structure and Lifeform:
  • Improved plot selection - Incorporated the newest contributed data, eliminating outlier plots that may have been impacted by disturbance, and allowing models to be trained using the highest quality inputs.
  • Improved base imagery – Has fewer seamlines and represents a longer time horizon, so that specific vegetation systems can be identified.
  • Structure input information - Developed methods to leverage lidar data, which provides stronger linkages with vegetation structure, so that lifeform classification can be improved.
  • Continuous canopy structure (Existing Vegetation Cover (EVC) and Existing Vegetation Height (EVH)) products – Developed methods to create continuous cover and height products to better inform fuels products.
• Existing Vegetation Type: Flexible vegetation type legends based on data driven range maps and ecoregions. Two classifications systems are used.
  • Ecological Systems – Allows for the comparison of data over time and informs LF fuel products,
  • National Vegetation Classification (NVC) – Standard at the Group level hierarchy.
• Enhancements to Remap Vegetation:
  • Ecological application - Riparian and wetland vegetation classes and sparsely vegetated classes were aggregated in previous LF versions. In LF Remap they are mapped separately.
  • Disturbed areas - Annual disturbance data were used to identify recently disturbed areas based on expected vegetation type recovery period within the previous 10-year period.
  • Wildland-urban interface (WUI) - Developed ruderal classes are better identified by combining updated WUI data with population density information from the U.S. Census Bureau.
  • Standardized ruderal and invasive vegetation types - New ruderal classes identify semi-natural vegetation types using a standardized legend within the U.S. National Vegetation Classification and Ecological Systems.
  • New and improved Auto-Keys - Redesigned to determine existing vegetation type classification based on plot data in the LFRDB.
  • Expert review and feedback - Of draft spatial products before delivery, so that fitness of use can be ensured.
• Additional Enhancements:
  • Field Data – Unprecedented access to the best field data with Bureau of Land Management for Assessment, Inventory, and Monitoring plots and Natural Resources Conservation Service for National Resources Inventory plots primarily in non-forest areas.
  • Geographic production unit – Existing Vegetation mapping image classification based on Omernik III/IV Ecoregions instead of National Land Cover Database map zones, which better matches the distribution of major vegetation types and in turn produces better output from modeling and resulting in less manual mapping techniques.
  • Metadata for LF Remap products were improved to include more details.
• LF 2016 Remap New Products:
  • Historical Disturbance – The latest 10 years of disturbance data representing disturbance year and original disturbance type and severity, allowing multiple years of disturbance to be captured for each pixel. This allows a historical representation of disturbance on the landscape and helps classify fuel data assignments.
  • National Vegetation Classification – Allows opportunities to evaluate the vegetation classification and mapped products.
  • Fuel Vegetation Cover, Height, and Type – More accurately reflects pre-disturbance vegetation type, as well as pre-disturbance binned cover and type in order to apply the effects of disturbances on fuel to depict fuel model assignments to the capable year.
     

LF Remap Resources

LANDFIRE 2016 Remap Project Closeout Report

Vegetation dynamics models: a comprehensive set for natural resource assessment and planning in the U.S. 

Changing the mapping approach for fuels in disturbed areas - LANDFIRE introduces "Capable" Fuel - a summary of the new mapping process, called "capable" fuels.

LANDFIRE Capable Fuels - changes to the fuels assignment process for LF Remap - a technical description of the refinement to the LF Remap products now representing "capable" fuels.

LANDFIRE 2016 Remap and "Year Capable" Fuels 

LF Remap Image Processing

The New Age of LANDFIRE: LF 2016 Remap and the LF 2020 Update

LANDFIRE 2016 Remap: Improvements to the Reference Database

LANDFIRE Remap Prototype Mapping Effort: Developing a New Framework for Mapping Vegetation Classification, Change, and Structure - a narrative of the innovative methodologies in image processing and mapping used to create LF Remap LANDFIRE vegetation products

LF Remap - Utilization of Remotely Sensed Data to Classify Existing Vegetation Type and Structure to Support Strategic Planning and Tactical Response - characterizing vegetation type, cover, and height with Earth observation data enhances wildland fuel modeling.

Developing LF Remap by Production Unit - products are developed using a combination of legacy and new productions units.