C-CAP United States Virgin Islands, St. John 2005-2007-Era Land Cover Change Analysis

Create custom data files by choosing area, map projection, file format, etc.

Date that the source FGDC record was last modified.

Converted from FGDC (Federal Geographic Data Committee) CSDGM (Content Standard for Digital Geospatial Metadata) (version FGDC-STD-001-1998) using 'fgdc_to_inport_xml.pl' script. Contact Tyler Christensen (NOS) for details.

Partial upload of Positional Accuracy fields only.

Partial upload to move data access links to Distribution Info.

NOAA OCM | Source Geospatial Form: remote-sensing image | Type of Source Media: CD-ROM

Process_Description

This dataset was created by Sanborn.

This dataset contains the classifications of 2005-era imagery and 2007-

era imagery. It is used for change analysis between these years. The

attributes include a from-to change description category. There are

481 classes corresponding to the 481 possible from-to change combinations.

In this section of the metadata, the 2005-era classification

procedure is described, then the 2007-era change detection classification procedure is

described. Finally in the post-processing section, the process of combining

the maps to generate a spatial matrix of change is discussed.

Early Date (2005-era) Classification:

This section outlines the classification procedure for the U.S. Virgin Islands C-CAP product.

The IKONOS image tiles utilized for this project were provided by GeoEye as orthorectifed,

georeferenced products and were consistent with the quoted 1:12000 spatial accuracy. In conjunction with the

IKONOS, Sanborn used LiDAR collected in 2007 for the island. Training data was collected in the

field between May 31th and June 5th, 2009. Land cover observations were recorded by analysts. The

referenced dataset went through a QA/QC procedure to ensure it met accuracy standards. Image segmentation

was completed at multiple scales using the multispectral (2.4 m) imagery in order to group like spectral

and textural objects within the imagery. For consistency, the associated impervious data set was incorporated

into the segmentation layer as a boundary delimiter. Segments can share boundaries of an impervious surface,

but can never overlap an impervious surface. Image segments are classified using a decision tree classifier

in See5 software and an iterative nearest neighbor classifier in Definiens software. Nearest neighbor

training data is collected through field observations, and analyst photo interpretation/selection.

Models are built to refine or reclassify land cover areas by utilizing the wealth of attribute information

linked to each segment within Definiens. As with any automated or semi-automated land cover classification

there are often inconsistencies in the land cover map. The final step before map finalization was to remove

inaccuracies through manual segment labeling as interpreted by an analyst. Sanborn used independent

reviewer's comments to further refine the land cover map.

Attributes for this product are as follows:

0 Background

1 Unclassified

2 Impervious

3

4

5 Developed, Open Space

6 Cultivated Crops

7 Pasture/Hay

8 Grassland/Herbaceous

9 Deciduous Forest

10 Evergreen Forest

11 Mixed Forest

12 Scrub/Shrub

13 Palustrine Forested Wetland

14 Palustrine Scrub/Shrub Wetland

15 Palustrine Emergent Wetland

16 Estuarine Forested Wetland

17 Estuarine Scrub/Shrub Wetland

18 Estuarine Emergent Wetland

19 Unconsolidated Shore

20 Bare Land

21 Open Water

22 Palustrine Aquatic Bed

23 Estuarine Aquatic Bed

24 Tundra

25 Snow/Ice

Ancillary Datasets:

Non-IKONOS image datasets used are LiDAR (2 foot point spacing density), DCM (Digital Canopy Model derived from LiDAR),

slope, aspect, NWI, NLCD, field-collected points, photo-interpreted points.

QA/QC Process:

There were several QA/QC steps involved in the creation of this product. There was an internal QA/QC.

This was done by viewing the classification frame-by-frame along with the IKONOS imagery, the

classification, and high resolution reference imagery. NOAA staff completed a similar review and

provided both general and point comments.

Late Date (2007-era) Classification:

Pre-processing steps:

The ADS40 was provided to Sanborn as a geo-referenced product. Sanborn perfomed a quality

control on the imagery to check for pixel dropouts and other image issues. Individual image tiles were

combined to create a single image for the entire island. The early date (2005-era) classification was

created first using satellite data because at the time it was the most recent, cloud free imagery

available. The late date (2007-era) ADS40 became available shortly after the early date classification

was completed. The ADS40 imagery was found to have ortho-rectification accuracy. The 2005-era classifications

and imagery were geo-rectified to co-register with the ADS40 imagery.

Change Detection:

Once the images were aligned, the next step was to

determine the areas of change between the 2005-era and 2007-era data. Since the

2005-era classification was already complete, it is only necessary to classify

the areas of change to create a 2007-era classification of the same specifications.

The 2007 ADS40 imagery was segmented with the 2005 land cover classification as a hard boundary

delimter. Logical rulesets and thresholding based on the spectral properties of the imagery

and the early date classification were used to generate a mask of potential change. The change

mask was incorporated into the early date classification to create the draft late date classification.

This map was further revised with rulesets and thresholding. The final step before map finalization was

to remove inaccuracies through manual segment labeling as interpreted by an analyst. Sanborn

used independent reviewer's comments to further refine the land cover map.

Post-Processing Steps:

A GIS matrix algorithm was run with 2005 and 2007 datasets as inputs. The algorithm generates

a from-to file where each combination is uniquely labeled from 1 to 481.

Metadata imported