Division 2 2003 Imagery Acquisition

• Title: Division 2 - Acquisition and Preprocessing of Imagery for the 2003 Arkansas River Basin Update
• Description: Final Documentation of Tasks 2-4
• Decision Support - Doc Type: Task Memorandum
• Date: 6/25/2004
• DSS Category: GIS
• DSS: Arkansas River
• Basin: Arkansas
• Contract/PO #: OE PDA 04-74
• Prepared By: Riverside

D IVISION 2 MEMORANDUM
Table 7.0 Ground Control Points (GCPs) and Root Mean Square Errors (RMS)
Obtained in Geometric Correction of the Landsat Imagery for the 2003 Arkansas River
Basin Update
Path/Row
Acquisition Date
Sensor
Number of GCPs
RMS Error
31/33
16- April -2003
Landsat 7
25
0.64
32/33
26- May -2003
Landsat 5
25
0.59
33/33
27- June -2003
Landsat 5
26
0.65
34/33
14- August -2003
Landsat 5
29
0.60
15- September -2003
Landsat 5
27
0.65
Table 8.0 Ground Control Points (GCPs) and Root Mean Square Errors (RMS) Obtained
in Geometric Correction of Additional Landsat Imagery to Cover the Entire Area of
Division 2
Path/Row
Acquisition Date
Sensor
Number of GCPs
RMS Error
31/33
12- May -2003
Landsat 5
22
0.66
32/33
6 -July -2003
Landsat 5
23
0.61
33/34
27- June -2003
Landsat 5
26
0.68
34/33
4 -July -2003
Landsat 5
28
0.65
Once the best RMS errors were obtained for each Landsat scene, the rectified output file was created. In
order to make the grid of pixels in the uncorrected image match the grid for the reference image, the
pixels in the uncorrected image were resampled using the nearest neighbor interpolation. This technique
transfers original data values without averaging them as other resampling methods do (e.g., cubic
convolution, bilinear interpolation); therefore, the extremes and subtleties of the data values are not lost.
This is an important consideration when discriminating between vegetation types, locating an edge
associated with a lineament, or determining different levels of turbidity or temperatures in a lake (Jensen
1996). To determine an output pixel's nearest neighbor, the rectified coordinates of that pixel are
converted back to the source coordinate system using the inverse of the transformation. The pixel that is
closest to the retransformed coordinates is the nearest neighbor. The data file values for that pixel
become the data file values of the pixel in the output image (ERDAS Imagine 8.7 Online Documentation,
2003).
Page 14 of 23 ( Riverside Technology, inc.
Water Resources Engineering and Consulting