Section 1 |
Identification Information |
| Originator |
Peter T. Wolter |
| Title |
Landsat TM Forest Classification
for Northeastern Minnesota |
| Abstract |
Wolter, Peter T. and David
J. Mladenoff, 1995. Improved forest classification of northeastern
Minnesota using Multi-temporal Landsat data and National Wetlands
Inventory data. Proceedings of the 10th Annual U.S. Landscape Ecology
Symposium, April 22-26, 1995.
Forest classifications using Landsat TM data have been only moderately
successful in separating forest cover types in the northern Lake States
region where forest cover types are notoriously heterogeneous. Forest
cover types in northern Wisconsin were classified to roughly Anderson
III precision with 80 percent accuracy by incorporating Landsat imagery
from several key phenological dates. One criticism of the Wisconsin
classification work was that the study region used was not fully representative
of what is considered classic northern Lake States cover type heterogeneity
(i.e., northeastern Minnesota). In this study we developed a more
specific forest cover type classification for northeastern Minnesota
using Landsat TM data from early summer in conjunction with key ancilliary
MSS and TM dates to capture phenological changes of different tree
species. In addition to satellite data, National Wetlands Inventory
data were used to post-classify cover types that were initially confused
in terms of spectral reflectance but were known to be associated with
different soil moisture regines. Of greatest significance, aspen-birch,
maple, black ash, and tamarack were sucessfully classified. This approach
may be useful for broad-scale forest monitoring in other areas, particularly
where ancilliary data layers are available. |
| Purpose |
Developed for the Forest
Bird Monitoring Initiative funded by LCMR and the USDA Forest Service.
|
| Time Period of Content
Date |
|
| Currentness Reference
|
Date of the base TM scene
for the:
Chippewa National Forest region 8/6/1990
western Superior NF region 7/31/1990
eastern Superior NF region 7/2/1988
Mille Lacs - St. Croix region 6/5/1987 |
| Progress |
Complete |
| Maintenance and Update
Frequency |
As needed |
| Spatial Extent of Data
|
Landsat World-wide Reference
System (WRS) coordinates 28/27, 27/27, 26/27, 27/28 in the same order
listed for the dates |
| Bounding Coordinates
|
-95.60
-89.67
48.62
45.07 |
| Place Keywords
|
Duluth, Minnesota is approximately
the geographic center of these data. |
| Theme Keywords
|
Dominant species forest
cover classification for northeast Minnesota. |
| Theme Keyword Thesaurus
|
Landsat TM |
| Access Constraints
|
|
| Use Constraints
|
Use of these data is restricted
to the purchaser; the data may not be distributed third party entities.
Any use or analysis of these data resulting in publication, technical
reports or maps must give credit to Peter T. Wolter, David J. Mladenoff,
Gerald Niemi and the Natural Resources Research Institute GIS Lab,
UMD. Credits should also include the funding source: This landuse/landcover
map was funded with dollars appropriated by the Minnesota State Legislature
from the Environmental Trust Fund as recommended by the Legislative
Commission on Minnesota Resources. |
| Contact Person Information
|
Peter T. Wolter, Research
Fellow
Natural Resources Research Institute, University of Minnesota - Duluth
5013 Miller Trunk Highway
Duluth, Minnesota 55811
Phone: 218 720 4275
FAX: 218 720 4219
E-mail: pwolter@sparkie.nrri.umn.edu
|
| Browse Graphic File
Name |
none available |
| Browse Graphic File
Description |
none |
| Associated Data Sets
|
none |
Section 2 |
Data Quality Information - - - - - - Go back
to top |
| Attribute Accuracy
|
Overall classification
accuracy ranges from 72 to 75 while individual forest class accuracies
range from approximately 65 to 95 percent. These data were validated
using the Minnesota DNR's Phase II forest inventory data (CSA). |
| Logical Consistency
|
The 'conifer, low density'
attribute is a confusion class from the eastern Superior National
Forest region (WRS 26/27). Atmospheric problems captured by the 7-2-88
Landsat TM scene precluded confident separation of sparse conifers
--largely jack pine-- that were associated with bare to brushy matrix
ground cover. Therefore, these covertypes were put into their own
class. |
| Completeness |
jack pine = dominated by
jack pine
jack pine - hardwood = jack pine with hardwood associate in overstory
jack pine - oak = jack pine with a pin oak associate
red pine = dominated by red pine
red pine - hardwood = red pine with hardwood associate in overstory
spruce/fir = balsam fir and white spruce occuring singlely or in mixture
spruce/fir - hardwood = spruce/fir with hardwood associate in overstory
cedar = dominated by northern white-cedar
cedar - hardwood = N. white-cedar with hardwood associate in overstory
tamarack = dominated by eastern larch -- tamarack
black spruce = dominated by black spruce
acid bog conifer = stagnant black spruce, tamarack or N. white-cedar
conifers, low density = sparse conifer associated with grass and brush
conifer, regeneration = primarily regenerating red pine or jack pine
black ash = dominated by black ash
black ash - conifer = black ash with conifer associate in overstory
black ash - conifer under = black ash with understory conifer associate
hardwoods, misc. lowland = undifferentiated lowland hardwoods
aspen/birch = dominated by aspen or birch singlely or in mixture
aspen/birch - conifer = aspen/birch with conifer associate in overstory
aspen/birch - conifer under = aspen/birch with conifer in the understory
northern hardwoods = dominated by Acer spp.
N. hardwoods - conifer = N. hardwoods with conifer in overstory
N. hardwoods - conifer under = N. Hardwoods with conifer in understory
red oak = dominated by red oak
oak - pine = pin oak and red oak with a conifer associate
hardwoods, transitional = sparse hardwood with brush/grass associations
hardwood regeneration = regenerating hardwood < approx. 5 years
old
bare = bare soil or rock outcrop
water = rivers, lakes, ponds and reservoirs
aquatic emergent = Scirpus spp., Nuphar spp, wild rice, etc.
emergent = Typha spp., Carex spp., etc.
Sphagnum = Sphagnum spp.
grass, native = native upland graminoids
grass, native (lowland) = native lowland graminoids
grass, cool season = native graminoids that have greened up by mid-May
grass, domestic = agriculture or domestic herbaceous vegetation
brush, alder = mesic brush dominated by alder
brush, alder (lowland) = wetland brush dominated by Alnus spp.
brush, willow = mesic brush dominated by Salix spp.
brush, willow (lowland) = lowland brush dominated by Salix spp.
brush, misc. = Corylous spp., Cornus spp., Alnus spp., Salix spp.,
etc.
brush, misc. (lowland) = primarily Alnus spp. and Salix spp.
brush, ericacious = bog rosemary, labrador tea, leather leaf, etc.
developed = commercial, residential, recreational or industrial
roads = paved and unpaved roads; augmented using TIGER DLG data. |
| Horizontal Positional
Accuracy |
RMS error < 0.5 pixels
(14.25 meters) using a second order polynomial transformation. USGS
7.5 minute topographic quadrangle maps were used as the reference
for the transformation. |
| Vertical Positional
Accuracy |
none |
| Lineage |
For an overview of the
basic techniques used to classify this region see:
Wolter, P.T., D.J. Mladenoff, G.E. Host and T.C. Crow, 1995. Improved
forest classification in the northern Lake States using multi-temporal
Landsat imagery, Photogrammetric Engineering & Remote Sensing,
61(9):1129-1143.
Data used:
Chippewa footprint (WRS 28/27);
TM 8-6-90
MSS 5-13-88, 9-26-85, 8-9-90
Superior footprint (WRS 26/27);
TM 7-2-88, 3-6-84
MSS 5-26-86, 9-26-90
Iron Range footprint (WRS 27/27);
TM 7-31-90, 5-12-90, 9-20-91, 2-25-92
St. Croix footprint (WRS 27/28);
TM 6-5-87, 2-25-92
MSS 5-4-87, 10-14-88 |
| Source Scale Denominator
|
24000 |
Section 3 |
Spatial Data Organization Information - - - - - -
Go back to top |
| Native Data Set Environment
|
These data --arrow83b.img
220megs-- were generated using ERDAS Imagine 8.3 on a SUN Sparc Station
5. |
| Geographic Reference
for Tabular Data |
|
| Spatial Object Type
|
Raster |
| Vendor Specific Object
Types |
cell |
| Tiling Scheme
|
none |
Section 4 |
Spatial Reference Information - - - - - -
Go back to top |
| Horizontal Coordinate
Scheme |
UTM |
| Ellipsoid |
GRS80 |
| Horizontal Datum
|
NAD83 |
| Horizontal Units
|
Meters |
| Distance Resolution
|
28.5 |
| Altitude Datum
|
Not applicable |
| Altitude Units
|
Not applicable |
| Depth Datum |
Not applicable |
| Depth Units |
Not applicable |
| Cell Width |
28.5 |
| Cell Height |
28.5 |
| UTM Zone Number
|
15 |
Section 5 |
Entity and Attribute Information - - - - - -
Go back to top |
| Entity and Attribute
Overview |
These vegetation cover
type data for northeastern Minnesota were derived from Landsat TM
and MSS data sources. Information regarding forest cover is based
on the dominant species or dominant association (e.g., northern hardwoods,
spruce-fir or aspen-birch). No tree density, volumn or specific age
class information is contained within these data. |
| Entity and Attribute
Detailed Citation |
|
Section 6 |
Distribution Information - - - - - - Go back
to top |
| Publisher |
Peter T. Wolter, Natural
Resources Research Institute - UMD |
| Publication Date
|
Periodically revised |
| Online Linkage
|
http://gisdata.nrri.umn.edu/coastalgisdata/Lake_Superior_Basin_shapefiles/arrow95.zip
|
| Contact Person Information
|
Gerald Sjerven
GIS Specialist
NRRI-CWE
5013 Miller Trunk Highway
Duluth, MN 55811-1442
Phone: 218-720-4388
FAX: 218-720-4328
E-mail: gsjerven@nrri.umn.edu |
| Distributor's Data
Set Identifier |
|
| Distribution Liability
|
|
| Transfer Format Name
|
|
| Transfer Format Version
Number |
|
| Transfer Size
|
|
| Ordering Instructions
|
|
Section 7 |
Metadata Reference Information - - - - - -
Go back to top |
| Metadata Date
|
|
| Contact Person Information
|
Gerald Sjerven
GIS Specialist
NRRI-CWE
5013 Miller Trunk Highway
Duluth, MN 55811-1442
Phone: 218-720-4388
FAX: 218-720-4328
E-mail: gsjerven@nrri.umn.edu |
| Metadata Standard Name
|
|
| Metadata Standard Version
|
|
| Metadata Standard Online
Linkage |
|
| |
|