| | Image Processing
All images were processed using Intergraph Corporation ERDAS IMAGINE® 2014 software. Prior to image classification, each Landsat Level 1 data product was stacked into a composite multi-band image file and then converted to at-sensor radiance and top-of-atmosphere (TOA) reflectance; finally, adjacent scenes (Worldwide Reference System 2 [WRS-2] path 14 row 33 and path 14 row 34) were mosaicked to a single image and clipped to the study area extent. The six reflective bands (visible blue, green, and red; near infrared; and two short-wave infrared bands) for each sensor were stacked and used in subsequent processing and analyses (table 2). Radiometric processing reduces scene-to-scene variability due to gain changes between scenes or differences in solar zenith angles, exoatmospheric solar irradiance, and the earth-sun distance resulting from different acquisition dates and times (Chavez, 1988, 1989, 1996; Chander and others, 2009).
Table 2. Comparison of Landsat 5 and equivalent Landsat 8 spectral bands used in image analyses (U.S. Geological Survey, 2012). [Abbreviation: μm, micron]
Spectral Range |
Landsat 5 Band |
Wavelength (μm) |
Landsat 8 Band |
Wavelength (μm) |
Visible blue | Band 1 | 0.45-0.52 | Band 2 | 0.45-0.51 |
Visible green | Band 2 | 0.52-0.60 | Band 3 | 0.53-0.59 |
Visible red | Band 3 | 0.63-0.69 | Band 4 | 0.64-0.67 |
Near infrared | Band 4 | 0.77-0.90 | Band 5 | 0.85-0.88 |
Short-wave infrared | Band 5 | 1.55-1.75 | Band 6 | 1.57-1.65 |
Short-wave infrared | Band 7 | 2.09-2.35 | Band 7 | 2.11-2.29 |
|