10.4 Remote Sensing and Ecosystem Management Examples
145
high-altitude aerial photos would be useful to map
and analyze vegetation.
Vegetation Classification at the
Subregional Level
The Classification and Assessment with Landsat of
Visible Ecological Groupings (CAL VEG) was developed to provide a classification system and
mapping effort for the state of California to more
adequately assess natural resources statewide.
CAL VEG uses a hierarchical classification that allows vegetation to be divided into broad units for
region-wide or larger-area landscape planning. This
classification system was designed to describe existing vegetation at various levels of detail (USDA
Forest Service Regional Ecology Group, 1981).
CAL VEG mapped overstory size, structure, crown
closure, and plant series associations. Plant series
associations were designed to aggregate into lifeforms (i.e., shrub, conifer, deciduous, grassland).
Maps were created using automated classification
procedures based on multispectral Landsat TM imagery, field verification, aerial photography, and local knowledge.
Landscape
Landtype associations are mapped at the landscape
level. General topography, geomorphic processes,
surficial geology, type of soil, local climate, and
potential natural communities determine landtype
association boundaries (Avers et aI., 1993). At the
landscape level, forest or watershed assessments
are possible using Landsat TM, SPOT, and aerial
photos. The 185-kIn swath width and 30-m spatial
resolution of Landsat TM imagery make these data
useful for both subregional and landscape studies.
The moderate swath width provides a synoptic
view of a large area, and the mid-level spatial resolution supplies significant detail about vegetation
or land cover conditions.
Aspen Change Detection and Mapping for
Landscape Analysis in Southwestern Montana
The Beaverhead-Deerlodge National Forest conducted a landscape-level assessment of current and
historical aspen populations (Lachowski et aI.,
1996). The study area, 186,000 hectares located in
the Gravelly Mountain Range in southern Montana,
was selected to analyze current aspen locations. A
vegetation classification based on Landsat TM imagery yielded 15 vegetation classes. Further refinement was necessary to account for spectral confusion between willow and aspen stands. Aerial
photos, digital camera images, and GIS modeling
using elevation and slope were employed to reduce
classification errors.
To evaluate aspen decline, historic aerial photos
were compared to the Landsat TM classification
over a 2400-hectare subsection of the study area.
Aspen, aspen-conifer mix, and conifer cover types
were interpreted from 1947 black-and-white aerial
photos. The photointerpreted data were then digitized and overlaid onto the satellite classification
to identify changes in aspen distribution. Successional change in conifer distribution was determined to be the primary cause of aspen decline
(Figure 10.4).
Land Unit
Land units are the building blocks of landtype associations and are described as either landtypes or
landtype phases (Avers et aI., 1993). At this scale,
local topography, rock and soil types, and vegetation determine unit descriptions and boundaries.
Landtypes and landtype phases are used for managing and planning at the forest or project level.
High-resolution remotely sensed data are combined
with digital elevation models (DEMs), field data,
large-scale aerial photos, and expert local knowledge to describe ecological units at this level.
Ecological Unit Mapping with Remote
Sensing and GIS
The Bridger-Teton East Ecological Unit Inventory
team inventoried ecological units on the western
slopes of the Wind River Mountain Range in
Wyoming. The inventory process integrated information on landform, geology, potential natural vegetation, and soils to map and characterize ecological units. Satellite imagery and DEMs within a GIS
were used to expedite mapping and to stratify the
study area for sampling into preliminary map units
or landscape units. Repeating patterns were recognized across the landscape by grouping areas with
similar ecological properties, thus further expediting the mapping process (Figure 10.5). Stratifying
the study area helped to develop an effective strategy for field sampling. The boundaries of the map
units and landscape units were refined through aerial photointerpretation and extensive field data collection. Products generated for this project included
a classification scheme for ecological types, map
units that comply with Natural Resource Conservation Service mapping standards, and data layers
for the Bridger-Teton National Forest's GIS database.
145
high-altitude aerial photos would be useful to map
and analyze vegetation.
Vegetation Classification at the
Subregional Level
The Classification and Assessment with Landsat of
Visible Ecological Groupings (CAL VEG) was developed to provide a classification system and
mapping effort for the state of California to more
adequately assess natural resources statewide.
CAL VEG uses a hierarchical classification that allows vegetation to be divided into broad units for
region-wide or larger-area landscape planning. This
classification system was designed to describe existing vegetation at various levels of detail (USDA
Forest Service Regional Ecology Group, 1981).
CAL VEG mapped overstory size, structure, crown
closure, and plant series associations. Plant series
associations were designed to aggregate into lifeforms (i.e., shrub, conifer, deciduous, grassland).
Maps were created using automated classification
procedures based on multispectral Landsat TM imagery, field verification, aerial photography, and local knowledge.
Landscape
Landtype associations are mapped at the landscape
level. General topography, geomorphic processes,
surficial geology, type of soil, local climate, and
potential natural communities determine landtype
association boundaries (Avers et aI., 1993). At the
landscape level, forest or watershed assessments
are possible using Landsat TM, SPOT, and aerial
photos. The 185-kIn swath width and 30-m spatial
resolution of Landsat TM imagery make these data
useful for both subregional and landscape studies.
The moderate swath width provides a synoptic
view of a large area, and the mid-level spatial resolution supplies significant detail about vegetation
or land cover conditions.
Aspen Change Detection and Mapping for
Landscape Analysis in Southwestern Montana
The Beaverhead-Deerlodge National Forest conducted a landscape-level assessment of current and
historical aspen populations (Lachowski et aI.,
1996). The study area, 186,000 hectares located in
the Gravelly Mountain Range in southern Montana,
was selected to analyze current aspen locations. A
vegetation classification based on Landsat TM imagery yielded 15 vegetation classes. Further refinement was necessary to account for spectral confusion between willow and aspen stands. Aerial
photos, digital camera images, and GIS modeling
using elevation and slope were employed to reduce
classification errors.
To evaluate aspen decline, historic aerial photos
were compared to the Landsat TM classification
over a 2400-hectare subsection of the study area.
Aspen, aspen-conifer mix, and conifer cover types
were interpreted from 1947 black-and-white aerial
photos. The photointerpreted data were then digitized and overlaid onto the satellite classification
to identify changes in aspen distribution. Successional change in conifer distribution was determined to be the primary cause of aspen decline
(Figure 10.4).
Land Unit
Land units are the building blocks of landtype associations and are described as either landtypes or
landtype phases (Avers et aI., 1993). At this scale,
local topography, rock and soil types, and vegetation determine unit descriptions and boundaries.
Landtypes and landtype phases are used for managing and planning at the forest or project level.
High-resolution remotely sensed data are combined
with digital elevation models (DEMs), field data,
large-scale aerial photos, and expert local knowledge to describe ecological units at this level.
Ecological Unit Mapping with Remote
Sensing and GIS
The Bridger-Teton East Ecological Unit Inventory
team inventoried ecological units on the western
slopes of the Wind River Mountain Range in
Wyoming. The inventory process integrated information on landform, geology, potential natural vegetation, and soils to map and characterize ecological units. Satellite imagery and DEMs within a GIS
were used to expedite mapping and to stratify the
study area for sampling into preliminary map units
or landscape units. Repeating patterns were recognized across the landscape by grouping areas with
similar ecological properties, thus further expediting the mapping process (Figure 10.5). Stratifying
the study area helped to develop an effective strategy for field sampling. The boundaries of the map
units and landscape units were refined through aerial photointerpretation and extensive field data collection. Products generated for this project included
a classification scheme for ecological types, map
units that comply with Natural Resource Conservation Service mapping standards, and data layers
for the Bridger-Teton National Forest's GIS database.