LANDFIRE-FAQs

LANDFIRE Fact Sheet
Background
Organization

Sponsor
Partners & Cooperators
Exec. Oversight Comm.
Organization Chart
Project Components

Prototype
LANDFIRE National
Updating Products
Information Bulletins
Associated Websites

Overview

LANDFIRE National arrow down

Data Products

Notices

Version Alerts close

Notifications

Data Product Access arrow down

Overview

National Map
bullet

Data Access Tool
bullet

DVD

FTP

Applications
bullet

Data Product Quality

Updated Products

Data Products
Data Product Access

National Map
LF Rapid Assessment (2005)
Helpful Data Tools

National Schedule

Schedule Map
Schedule Table
Mapping Status
Updated Products Schedule
Key LANDFIRE Dates

Field Reference Database

Overview
Sharing your Existing Data
Collecting New Data
Vegetation Modeling

Overview
Workshops
Model Review
Fuel Workshops

FBFM Assignment
FBFM Calibration
Updating Products

Overview
Refresh Data Needs
Providing Event Data
Providing Other Data
Training Workshops
Key LANDFIRE Dates

Link toNational
Map LANDFIRE

LANDFIRE
National Schedule

LANDFIRE ›› Frequently Asked Questions (FAQs)

Q. How does LANDFIRE map existing vegetation?

LANDFIRE's general approach to mapping existing vegetation entails several major processes:

1) Assignment of the LANDFIRE existing vegetation type map units to plots in the LANDFIRE reference database

For each LANDFIRE mapping zone, NatureServe develops a draft rule set for classifying field reference plots to LANDFIRE existing vegetation type (EVT). The development of these "sequence tables " is initiated at regional EVT workshops organized by NatureServe, and these tables are delivered as an ordered set of floristic queries that are implemented in the LANDFIRE reference database (LFRDB) to assign EVT map units to each plot. Draft sequence tables are reviewed by vegetation ecologists at the Missoula Fire Sciences Laboratory in Missoula, Montana and by remote sensing scientists at the USGS Center for Earth Resources Observation and Science (EROS) in Sioux Falls, South Dakota. An external review of the sequence tables is solicited, provided that the LANDFIRE schedule allows. The map units are then refined according to the LANDFIRE map unit development criteria listed under "What are the LANDFIRE design criteria for developing vegetation map units?", and ecological systems are either aggregated or split into mappable NVCS alliances. The final sequence table for each mapping zone is implemented in the LANDFIRE reference database, and the plots with EVT assignments in addition to a suite of QA/QC information are compiled as map attribute tables (MATs). It is important to note that, although LANDFIRE EVT map units may prove useful in other applications, such as wildlife habitat modeling or hydrologic modeling, they are crafted specifically to meet the requirements of the LANDFIRE wildland fuel and fire regime products.

2) Assignment of the LANDFIRE existing vegetation structure (height and cover) map units to plots in the LANDFIRE reference database

Height — Height information is compiled from the LFRDB and binned into classes (see table below) by dominant life form. LANDFIRE relies on the Forest Inventory and Analysis (FIA) database and other agency monitoring data for vegetation height and cover data. Binned height classes are used to meet the LANDFIRE design criterion that vegetation map units be "mappable;" in other words, they represent classes that result in a precision that meets LANDFIRE requirements. Using training data from the LFRDB, separate decision tree-based mapping models that have height as the response variable and imagery and biophysical gradients as predictor variables are developed for each life form. The final LANDFIRE height data product depicts average canopy height per pixel for the dominant life form as defined by the EVT map. The height maps are used in developing the LANDFIRE fuel maps and for developing and implementing LANDFIRE vegetation models.

Herbaceous	Shrub	Tree
0-0.5 meters	0-0.5 meters	0-5 meters
> 0.5-1.0 meters	> 0.5-1.0 meters	> 5-10 meters
> 1 meter	> 1.0-3.0 meters	> 10-25 meters
	> 3.0 meters	> 25-50 meters
		> 50 meters

Cover — Continuous canopy cover information is compiled from the LFRDB, binned into classes, and used to develop decision tree-based mapping models for shrub and herbaceous life forms defined by the LANDFIRE EVT map. Binned canopy classes are used in order to best arrive at a precision that meets LANDFIRE requirements. For the tree life form, the LANDFIRE canopy cover map is developed separately by the National Land Cover Data (NLCD) project using reference points based on digital ortho-quad (DOQ) and high-resolution imagery and regression modeling. The results of these spatially explicit decision tree mapping models are normalized over the three life forms, so a given pixel cannot have more than 100 percent canopy cover (based on overhead viewing of the canopy). LANDFIRE wildland fuel and vegetation modeling requirements dictate that pixels defined as a tree life form have at least 10 percent canopy cover, pixels defined as shrub life form have at least 10 percent canopy cover (and tree cover of less than 10 percent), and pixels defined as herbaceous life form have at least 10 percent canopy cover (and tree/shrub cover of less than 10 percent). Pixels are defined as "barren" if canopy cover for all life forms is less than 10 percent. The final LANDFIRE canopy cover products depict average canopy cover per pixel for the dominant life form as defined by the EVT map. The canopy cover information is used in creating the LANDFIRE fuel maps and for developing and implementing LANDFIRE vegetation models. The LANDFIRE fuel products include a forest canopy height product that is scaled to maximize ease of use in the FARSITE and FlamMap fire behavior applications.

3) Development of LANDFIRE life form and life zone stratification maps

LANDFIRE reference database plot processing includes assignment of dominant life form and life zone classes for each plot in the map attribute table (MAT) for each mapping zone. This information is used to create life form and life zone maps that serve as stratification tools for defining whether a mapped pixel in the EVT data set will be a tree, shrub, or herbaceous life form and whether the same pixel lies in a lowland-foothill, montane, or alpine-subalpine life zone. For each life form / life zone group (see table below), separate binary decision tree mapping models are developed. The resulting maps from these models serve as masks in the development of the LANDFIRE EVT product.

Life form / life zone code	Description
>11	Tree/alpine-subalpine
12	Tree/montane
13	Tree/lowland-foothills
21	Shrub/alpine-subalpine
22	Shrub/montane
23	Shrub/lowland-foothills
31	Herb/alpine-subalpine
32	Herb/montane
33	Herb/lowland-foothills
41	Any/alpine-subalpine
42	Any/montane
43	Any/lowland-foothills

4) Quality assurance / quality control

Based on information contained in the map attribute tables (MATs), the EVT mapping team reviews the plot information to detect problems and errors inherent in field-referenced data derived from disparate sources. This process involves the following steps:

Detecting outdated field data — The MAT is populated with information that quantifies Normalized Difference Vegetation Index (NDVI) change on every plot between the Multi-Resolution Land Characteristics (MRLC) Consortium's 1992 and 2001 Landsat image catalogs. This information is assumed to represent vegetation change during this period and is used, along with the plot sampling date, to determine whether the plot will be used in map development.

Detecting field data with erroneous geographic coordinates — Major geo-coding problems, such as plots located on roads or located out of mapping areas, are identified through

Back to FAQs