Accuracy Assessment FAQ Metadata

Metadata also available as - [Outline] - [Text] - [Adobe Acrobat pdf file]

Frequently-anticipated questions:

• What does this data set describe?
  1. How should this data set be cited?
  2. What geographic area does the data set cover?
  3. What does it look like?
  4. Does the data set describe conditions during a particular time period?
  5. What is the general form of this data set?
  6. How does the data set represent geographic features?
  7. How does the data set describe geographic features?
  8. What biological taxa does this data set concern?
• Who produced the data set?
  1. Who are the originators of the data set?
  2. Who also contributed to the data set?
  3. To whom should users address questions about the data?
• Why was the data set created?
• How was the data set created?
  1. What methods were used to collect the data?
  2. From what previous works were the data drawn?
  3. How were the data generated, processed, and modified?
  4. What similar or related data should the user be aware of?
• How reliable are the data; what problems remain in the data set?
  1. How well have the observations been checked?
  2. How accurate are the geographic locations?
  3. How accurate are the heights or depths?
  4. Where are the gaps in the data? What is missing?
  5. How consistent are the relationships among the data, including topology?
• How can someone get a copy of the data set?
  1. Are there legal restrictions on access or use of the data?
  2. Who distributes the data?
  3. What's the catalog number I need to order this data set?
  4. What legal disclaimers am I supposed to read?
  5. How can I download or order the data?
• Who wrote the metadata?

What does this data set describe?

1. How should this data set be cited?

   Remote Sensing and GIS Group, Technical Service Center, US Bureau of Reclamation, Mail Code D-8260, POB 25007, Denver CO 80225, 1999, Badlands National Park Accuracy Assessment Data: USGS-NPS Vegetation Mapping Program Badlands National Park, USGS-CSS, Denver, CO.

   Online Links:

   • <http://biology.usgs.gov/npsveg/ftp/vegmapping/badl/badlaa.zip>

   Other_Citation_Details: Created under contract to the USGS-CSS-CSAS

2. What geographic area does the data set cover?

   West_Bounding_Coordinate: -102.943

   East_Bounding_Coordinate: -101.817

   North_Bounding_Coordinate: 44

   South_Bounding_Coordinate: 43.432

   Description_of_Geographic_Extent:

   Badlands National Park, SD including approx 5 mile buffer around park which includes private lands, portions of Buffalo Gap National Grassland, and Pine Ridge Indian Reservation.

3. What does it look like?

   <http://biology.usgs.gov/npsveg/badl/images/badlaa.pdf> (PDF)

   450 Kbyte

4. Does the data set describe conditions during a particular time period?

   Calendar_Date: Aug-1998

   Currentness_Reference:

   From the USGS-NPS Vegetation Mapping Program Badlands National Park, South Dakota Procedure Report, November 19, 1999. See: <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>

5. What is the general form of this data set?

   Geospatial_Data_Presentation_Form: Table

6. How does the data set represent geographic features?
   1. How are geographic features stored in the data set?

      This is a Point data set.

   2. What coordinate system is used to represent geographic features?

      Grid_Coordinate_System_Name: Universal Transverse Mercator

      Universal_Transverse_Mercator:

      UTM_Zone_Number: 13

      Transverse_Mercator:

      Longitude_of_Central_Meridian: -105

      Latitude_of_Projection_Origin: 0

      False_Easting: 500000

      False_Northing: 0

      Scale_Factor_at_Central_Meridian: .9996

      Planar coordinates are encoded using coordinate pair
      Abscissae (x-coordinates) are specified to the nearest 1
      Ordinates (y-coordinates) are specified to the nearest 1
      Planar coordinates are specified in meters

      The horizontal datum used is North American Datum of 1983.
      The ellipsoid used is Geodedic Reference System 80.
      The semi-major axis of the ellipsoid used is 6378137.
      The flattening of the ellipsoid used is 1/298.257.
      

7. How does the data set describe geographic features?

   Entity_and_Attribute_Overview:

   Information collected for Accuracy Assessment included:

   Plot Number Park Code Date Observer(s) Datum Accuracy UTM Coordinates: Easting Northing UTM Zone 9. Offset from Point: Easting (in meters) Northing (in meters) Topographic Description Elevation Aspect Veg Assoc. at Site Veg Assoc. 2 within 50m of Site Veg Assoc. 3 within 50m of Site Major Species Present (by strata) Canopy Closure of Top Layer Rationale for Classification Comments

   Entity_and_Attribute_Detail_Citation:

   See:

   Accuracy Assessment Field Form: <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>

   Total area (meters 2 / 4046.9 acres/m 2 / 2.471 acres/hectare) and number of polygons per mapping unit. <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>

   Summary of AA Results for Badlands National Park, by map class. <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>

   Contingency table (error matrix) for BADL vegetation mapping accuracy assessment. <http://biology.usgs.gov/npsveg/ftp/vegmapping/badl/aa_matrix.xls>

   The map codes as described in Appendix 14: <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>

8. What biological taxa does this data set concern?

   Taxonomy:

   Keywords/Taxon:

   Taxonomic_Keyword_Thesaurus: None

   Taxonomic_Keywords: Plants

   Taxonomic_Keywords: Vegetation

   Taxonomic_Keywords: National Vegetation Classification System

   Taxonomic_Classification:

   Taxon_Rank_Name: Kingdom

   Taxon_Rank_Value: Plantae

   Applicable_Common_Name: Plant

   Taxonomic_Classification:

   Taxon_Rank_Name: Division-Phylum

   Taxon_Rank_Value: Tracheophyta

   Taxonomic_Classification:

   Taxon_Rank_Name: Class

   Taxon_Rank_Value: Angiospermai

   Taxonomic_Classification:

   Taxon_Rank_Name: Class

   Taxon_Rank_Value: Gymnospermae

Who produced the data set?

1. Who are the originators of the data set? (may include formal authors, digital compilers, and editors)
   • Remote Sensing and GIS Group, Technical Service Center, US Bureau of Reclamation, Mail Code D-8260, POB 25007, Denver CO 80225

2. Who also contributed to the data set?

   Dan Cogan, Doug Crawford, Trudy Meyer, Jean Pennell & Jim Von Loh with RSGIS Group of USBR; Jim Drake of TNC; Bruce Bessken and Glenn Plumb of Badlands NP, NPS

3. To whom should users address questions about the data?
   USGS/CSS/Core Science Analytics and Synthesis
   c/o USGS-NPS Vegetation Mapping Program Coordinator
   Denver Federal Center, Building 810, Room 8000, MS 302
   Denver, Colorado 80225
   

   (303) 202-4220 (voice)
   303-202-4219 (FAX)
   gs-b-npsveg@usgs.gov
   

Why was the data set created?

The purposes of the mapping effort are varied and include the following: Provides support for NPS Resources Management; Promotes vegetation-related research for both NPS and USGS/CSS; Provides support for NPS Planning and Compliance; Adds to the information base for NPS Interpretation; and Assists in NPS Operations. The NPS I&M goals are, among others, to map the vegetation of all national parks and monuments and provide a baseline inventory of vegetation.

How was the data set created?

1. What methods were used to collect the data?

   Method 0 of 1

   Type: Field

   The accuracy assessment (AA) for the BADL vegetation mapping project consisted of preliminary planning and discussion, logistical planning, fieldwork, analysis of fieldwork, and computation of final results. Preliminary planning involved BOR/RSGIG and Dr. Jack Butler (a plant ecologist contracted to collect the field AA data). Following detailed discussion, a modified accuracy assessment procedure dubbed "front-loading" (Owens 1998) was selected using protocols outlined in the Accuracy Assessment Procedures (TNC 1994).

   The following guidelines for the AA procedure were adopted at this time:

   Observations of vegetation classes were to be ground-based.

   Ground sampling techniques were to be similar to the Observation Points collected during initial classification.

   The number of samples per plant association/map class would vary depending on abundance of the class upon the landscape.

   No maximum number of points was assigned so that the Park could be sampled as completely as possible.

   Logistical planning for the AA revolved around coordination of work schedules and finding reasonable work areas within the Park. Collection of AA points was confined to within Park boundaries but not limited to the previously described gradsects. Instead, AA's were based on availability of access and to a lesser extent, time constraints. The actual assessment was begun prior to completion of preliminary vegetation maps for the Park, thus the need for the front-loading of the sample points. Selecting random AA sampling sites beforehand was deemed unnecessary due to familiarity of the principal researcher with vegetation types and distribution at BADL. The final point chosen for assessment was selected to be as representative as possible of the vegetation in the immediate area, well away from stand boundaries, and in a stand larger than the minimum mapping unit (exceptions were made for wetland and some shrub classes which rarely exceeded 0.5 hectares in size). Field ecologists were supplied with a vegetation key, to be used to determine plant associations/map classes entered on field forms.

   AA data, including limited habitat and vegetation data, was recorded on field forms to document the classification decision made by the investigator. This form was modified and expanded from previous forms to include fields for additional community types found within 50 meters of the actual assessment point. Modifications were made to help accommodate several types of difficult situations, such as AA points located in small inclusions, heterogeneous polygons/stands, and GPS PLGR error.

   465 AA data points were collected during August 1998. The weather at this time was unusually warm and vegetation readily identifiable unless heavily grazed. In the Park's North Unit, an extremely heavy growth of yellow sweetclover was present over most grassland and shrub communities. While noted during the AA sampling, very little yellow sweetclover was present during the previous year (1997), the year of aerial photography acquisition and interpretation. AA points were collected in proportion to the size of the plant association/ map class within the Park; e.g. more points were collected within extensive types.

   Reference:

   From what previous works were the data drawn?

   How were the data generated, processed, and modified?

   Date: Sep-1999 (process 1 of 1)

   Accuracy assessment of the BADL project area was conducted in September 1999. This involved entering all accuracy data points into a digital coverage and overlaying these electronically onto final vegetation maps (by DOQQ). AA identification numbers plotted alongside each point allowed for comparison with accuracy assessment data forms. A contingency table was set up to record the reference data (collected in the field) versus the sample data (vegetation map) for each map unit.

   Errors of commission (i.e. user's errors) for each class were calculated by dividing the number of correctly classified samples by the total number of samples that were classified as belonging to that map class. Errors of omission (i.e. producer's errors) for each class were calculated by dividing the number of samples that were classified correctly by the total number of reference samples in that class. Confidence intervals for each map class were calculated using one of the methods shown in Table 2 (USGS-NPS Vegetation Mapping, Badlands National Park, South Dakota, Methodology Report) depending on the normality and size of the data.

   Overall total accuracy for BADL was calculated across all sampled map classes by dividing the number of correctly classified accuracy points by the total number of accuracy points. Confidence intervals for overall total accuracy were calculated using the equation for normally distributed data (see above). A Kappa Index (Foody 1992) was used to help account for any correct classification due to chance.

   A total of 458 accuracy assessment points were used to assess the accuracy of the BADL vegetation map by: using AA points collected during the summer of 1998 (front-loading method); entering AA point coordinates into an electronic format to overlay on the vegetation map; comparing map vegetation classification (transferred from photo interpretation) with field assessment of vegetation type to determine errors of omission and commission; resolving questions by referring to original data forms, so that Dr. Butler could make the final determination; recording all information on the AA matrix.

   Overall, initial accuracy of the vegetation map is 80.6% for all vegetation classes and the Kappa Index is 78.2%. Results for each vegetation class are discussed here, and recommendations are made relative to creating a more accurate vegetation map, as desired.

   The specific results are presented in Tables 5 and 6 below. In general, the percentage of the Park that an individual map class covered is reflected in the number of AA points collected for that type. For example, map class 16-19 (Western Wheatgrass Alliance / Western Wheatgrass - Green Needlegrass Grassland) covers approximately 38% of the Park, and is represented by 29% of the AA points, and map Class 1 (Prairie Dog Town Complex) occupies approximately 2% of the Park and is represented by 3% of the AA points. An exception for this is map class 2 (Badlands Sparse Vegetation Complex), which covers approximately 46% of the Park but is represented by only 14% of the AA points.

   Two rare shrub classes were either not assessed or lightly assessed, due to their lack of abundance within the Park. These include map class 33 (Rabbitbrush Shrubland), which was not assessed and map class 38 (Sandbar Willow Temporarily Flooded Shrubland), which had one AA point but the polygon was attributed with a land use type; map class 56 (intermittent stream).

   What similar or related data should the user be aware of?

How reliable are the data; what problems remain in the data set?

1. How well have the observations been checked?

   Overall, initial accuracy of the vegetation map is 80.6% for all vegetation classes and the Kappa Index is 78.2%. Results for each vegetation class are discussed here, and recommendations are made relative to creating a more accurate vegetation map, as desired.

   The specific results are presented in Tables 5 and 6 in the "USGS-NPS Vegetation Mapping Program, Badlands National Park, South Dakota Methodology Report" <http://biology.usgs.gov/npsveg/badl/badlrpt.pdf>. In general, the percentage of the Park that an individual map class covered is reflected in the number of AA points collected for that type. For example, map class 16-19 (Western Wheatgrass Alliance / Western Wheatgrass - Green Needlegrass Grassland) covers approximately 38% of the Park, and is represented by 29% of the AA points, and map Class 1 (Prairie Dog Town Complex) occupies approximately 2% of the Park and is represented by 3% of the AA points. An exception for this is map class 2 (Badlands Sparse Vegetation Complex), which covers approximately 46% of the Park but is represented by only 14% of the AA points.

   Two rare shrub classes were either not assessed or lightly assessed, due to their lack of abundance within the Park. These include map class 33 (Rabbitbrush Shrubland), which was not assessed and map class 38 (Sandbar Willow Temporarily Flooded Shrubland), which had one AA point but the polygon was attributed with a land use type; map class 56 (intermittent stream).

2. How accurate are the geographic locations?

   The UTM coordinates and elevation of all plots were logged using a hand-held Precision Lightweight Global Positioning System (GPS) Receiver (PLGR) unit.

3. How accurate are the heights or depths?

4. Where are the gaps in the data? What is missing?

   AA data, including limited habitat and vegetation data, was recorded on field forms to document the classification decision made by the investigator. This form was modified and expanded from previous forms to include fields for additional community types found within 50 meters of the actual assessment point. Modifications were made to help accommodate several types of difficult situations, such as AA points located in small inclusions, heterogeneous polygons/stands, and GPS PLGR error.

   465 AA data points were collected during August 1998. The weather at this time was unusually warm and vegetation readily identifiable unless heavily grazed. In the Park's North Unit, an extremely heavy growth of yellow sweetclover was present over most grassland and shrub communities. While noted during the AA sampling, very little yellow sweetclover was present during the previous year (1997), the year of aerial photography acquisition and interpretation. AA points were collected in proportion to the size of the plant association/ map class within the Park; e.g. more points were collected within extensive types.

   All accuracy data points were entered into a digital coverage and overlayed electronically onto final vegetation maps (by DOQQ).

5. How consistent are the relationships among the observations, including topology?

   Unknown

How can someone get a copy of the data set?

Are there legal restrictions on access or use of the data?

Access_Constraints: None

Use_Constraints:

Acknowledgment of the USGS/CSS, National Park Service, and the USBR/RSGIS Group would be appreciated in products derived from these data. Any person using the information presented here should fully understand the data collection and compilation procedures, as described in the metadata, before beginning analysis. The burden for determining fitness for use lies entirely with the user.

1. Who distributes the data set? (Distributor 1 of 1)
   USGS/CSS/Core Science Analytics and Synthesis
   c/o USGS-NPS Vegetation Mapping Program Coordinator
   Denver Federal Center, Building 810, Room 8000, MS 302
   Denver, Colorado 80225
   

   (303) 202-4220 (voice)
   303-202-4219 (FAX)
   gs-b-npsveg@usgs.gov
   

2. What's the catalog number I need to order this data set?

   Badlands National Accuracy Assessment Data

3. What legal disclaimers am I supposed to read?

   The U.S. Geological Survey and the National Park Service shall not be held liable for improper or incorrect use of the data described and/or contained herein. These data and related graphics (if available) are not legal documents and are not intended to be used as such.

   The information contained in these data is dynamic and may change over time. The data are not better than the original sources from which they were derived. It is the responsibility of the data user to use the data appropriately and consistent within the limitations of geospatial data in general and these data in particular. Any related graphics (if available) are intended to aid the data user in acquiring relevant data; it is not approriate to use the related graphics as data.

   The U.S. Geological Survey and the National Park Service gives no warranty, expressed or implied, as to the accuracy, reliability, or completeness of these data. It is strongly recommended that these data are directly acquired from an U.S. Geological Survey and National Park Service servers and not indirectly through other sources which may have changed the data in some way. Although these data have been processed successfully on a computer system at the U.S. Geological Survey, no warranty expressed or implied is made regarding the utility of the data on another system or for general or scientific purposes, nor shall the act of distribution constitute any such warranty. This disclaimer applies both to individual use of the data and aggregate use with other data. It is also strongly recommended that careful attention be paid to the contents of the metadata file associated with these data. Mention of trade names or commercial products in this metadata report does not constitute endorsement or recommendation for use by the U. S. Department of the Interior, U. S. Geological Survey and National Park Service.

4. How can I download or order the data?
   • Availability in digital form:

     Data format:	HTML
     Network links:	<http://biology.usgs.gov/npsveg/badl/index.html>

   • Cost to order the data: None

Who wrote the metadata?

Dates:

Last modified: Nov-2001
Last Reviewed: 14-May-2010

Metadata author:

USGS/CSS/Core Science Analytics and Synthesis
c/o USGS-NPS Vegetation Mapping Program Coordinator
Denver Federal Center, Building 810, Room 8000, MS 302
Denver, Colorado 80225

(303) 202-4220 (voice)
303-202-4219 (FAX)
gs-b-npsveg@usgs.gov

Metadata standard:

FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata (FGDC-STD-001.1-1999)

Metadata extensions used:

• <http://www.nbii.gov/portal/server.pt/community/fgdc_metadata/255>