Landsat+Southern+Brazil


 * Landsat Atmospheric Correction **
 * Lab 7: **


 * Available Landsat 5 Images for Path 220 Row 76 with Less than 20% Cloud Cover **


 * ID || LANDSAT || PATH || ROW || YEAR || DAY || CLOUD COVER % || DATE || QUALITY ||
 * LT52200761986209AAA03 || 5 || 220 || 76 || 1986 || 209 || 10 || 7/28/1986 || 9 ||
 * LT52200761986257XXX03 || 5 || 220 || 76 || 1986 || 257 || 0 || 9/14/1986 || 7 ||
 * LT52200761988183CUB00 || 5 || 220 || 76 || 1988 || 183 || 0 || 7/1/1988 || 7 ||
 * LT52200762003288CUB00 || 5 || 220 || 76 || 2003 || 288 || 5 || 10/15/2003 || 9 ||
 * LT52200762004099CUB00 || 5 || 220 || 76 || 2004 || 99 || 0 || 4/8/2004 || 9 ||
 * LT52200762004243CUB00 || 5 || 220 || 76 || 2004 || 243 || 0 || 8/30/2004 || 9 ||
 * LT52200762005101CUB00 || 5 || 220 || 76 || 2005 || 101 || 0 || 4/11/2005 || 7 ||
 * LT52200762005149CUB00 || 5 || 220 || 76 || 2005 || 149 || 0 || 5/29/2005 || 9* ||
 * LT52200762005165CUB00 || 5 || 220 || 76 || 2005 || 165 || 3 || 6/14/2005 || 7 ||
 * LT52200762005213CUB00 || 5 || 220 || 76 || 2005 || 213 || 0 || 8/1/2005 || 7 ||
 * LT52200762005325CUB00 || 5 || 220 || 76 || 2005 || 325 || 1 || 11/21/2005 || 9 ||
 * LT52200762006104CUB00 || 5 || 220 || 76 || 2006 || 104 || 1 || 4/14/2006 || 7 ||
 * LT52200762006120CUB00 || 5 || 220 || 76 || 2006 || 120 || 16 || 4/30/2006 || 9 ||
 * LT52200762006136CUB00 || 5 || 220 || 76 || 2006 || 136 || 8 || 5/16/2006 || 9 ||
 * LT52200762006248CUB00 || 5 || 220 || 76 || 2006 || 248 || 12 || 9/5/2006 || 9 ||
 * LT52200762006296CUB00 || 5 || 220 || 76 || 2006 || 296 || 2 || 10/23/2006 || 9 ||
 * LT52200762007107CUB00 || 5 || 220 || 76 || 2007 || 107 || 7 || 4/17/2007 || 9 ||
 * LT52200762007155CUB00 || 5 || 220 || 76 || 2007 || 155 || 1 || 6/4/2007 || 9 ||
 * LT52200762007171CUB00 || 5 || 220 || 76 || 2007 || 171 || 0 || 6/20/2007 || 9 ||
 * LT52200762007251CUB00 || 5 || 220 || 76 || 2007 || 251 || 0 || 9/8/2007 || 9 ||
 * LT52200762008190CUB00 || 5 || 220 || 76 || 2008 || 190 || 0 || 7/8/2008 || 9 ||
 * LT52200762008254CUB00 || 5 || 220 || 76 || 2008 || 254 || 1 || 9/10/2008 || 9 ||
 * LT52200762009112CUB00 || 5 || 220 || 76 || 2009 || 112 || 0 || 4/22/2009 || 9 ||
 * LT52200762009144CUB00 || 5 || 220 || 76 || 2009 || 144 || 0 || 5/24/2009 || 9* ||
 * LT52200762009240CUB00 || 5 || 220 || 76 || 2009 || 240 || 0 || 8/28/2009 || 9 ||
 * LT52200762009256CUB00 || 5 || 220 || 76 || 2009 || 256 || 5 || 9/13/2009 || 9 ||
 * LT52200762010035CUB00 || 5 || 220 || 76 || 2010 || 35 || 1 || 2/4/2010 || 9 ||
 * LT52200762011086CUB00 || 5 || 220 || 76 || 2011 || 86 || 12 || 3/27/2011 || 9 ||
 * LT52200762008254CUB00 || 5 || 220 || 76 || 2008 || 254 || 1 || 9/10/2008 || 9 ||
 * LT52200762009112CUB00 || 5 || 220 || 76 || 2009 || 112 || 0 || 4/22/2009 || 9 ||
 * LT52200762009144CUB00 || 5 || 220 || 76 || 2009 || 144 || 0 || 5/24/2009 || 9* ||
 * LT52200762009240CUB00 || 5 || 220 || 76 || 2009 || 240 || 0 || 8/28/2009 || 9 ||
 * LT52200762009256CUB00 || 5 || 220 || 76 || 2009 || 256 || 5 || 9/13/2009 || 9 ||
 * LT52200762010035CUB00 || 5 || 220 || 76 || 2010 || 35 || 1 || 2/4/2010 || 9 ||
 * LT52200762011086CUB00 || 5 || 220 || 76 || 2011 || 86 || 12 || 3/27/2011 || 9 ||
 * LT52200762010035CUB00 || 5 || 220 || 76 || 2010 || 35 || 1 || 2/4/2010 || 9 ||
 * LT52200762011086CUB00 || 5 || 220 || 76 || 2011 || 86 || 12 || 3/27/2011 || 9 ||
 * LT52200762011086CUB00 || 5 || 220 || 76 || 2011 || 86 || 12 || 3/27/2011 || 9 ||

Based on cloud condition and quality, I use the Anniversary date of 5/29/2005 and 5/24/2009 for my analysis. Those with a *.


 * Lab 8: **

The two Landsat 5 images chosen, as stated above, are provided below atmospherically corrected with spectral profiles for 1 pixel for each image.



**May 29, 2005 Spectral Profile and Image**

**May 24, 2009 Spectral profile and Image**




 * Spatial Enhancement, Multispectral Transformation, and Change Detection below in labs 9-10. **
 * Lab 9: **
 * Original Subset Image of May 29, 2005 **


 * Smooth Filters **


 * Median 40% 3x3 and Median 40% 9x9 **

For the first Image above I used a Median filter with a kernel size of 3 at 40%. As you can see for yourself the filter smooths out the pixels that are very visible in the original subset image. That no longer is the case here. In fact the image itself seems less busy and is softer on the eyes, making the image easier to see. For the second image, the applied filter is the same as the one above, Median filter; however it has a kernel size of 9, also at 40%. The pixels are smooth out even further to where perhaps is hard to see definition between the land cover.

For this image I used Roberts filter at 60% to sharpen the edges of the image, in order to see more distinction between the lines.
 * Edge Filter**
 * Reoberts 60%**



In the PCs images below it is evident that as the PC number bands progress the images become noisy and less visible. The first PC image is very similar to that of the original subset image however from there definition is harder to tell.
 * True color original subset and Six Principal Components**


 * Statistics**

As you can see in the correlation table there are many bands with negative correlations, the highest being Band 2 and 4 with a negative correlation of -.97859. I’m not quite sure what the negative correlations mean, however, there are some bands with high positive correlations. Those being band 4 and 6 at with .78 and bands 6 and 1 with a positive correlation of .83.

For the PC Variability Percentage table, over 98% of the variability is explained by the first 3 PC bands. 71% is explained by the first band itself and 25% by the following band.


 * Inverse PC**

As instructed I took the top PC images and created an inverse PC rotation image. There seem to be little difference from the original images. It seems to be clearer but less bright. The colors that are brighter in the original image are now more grayish in the inverse pc image. It is a bit less noisy from the smooth out pixels but still is hard to see some definition is some areas.


 * Lab 10: **
 * Two Color Multiview for band 4 **
 * PC band 2 **
 * Change in NDVI **

Lab 10 Report




 * Lab 11& 12:**

__** May 29,2005 **__

Training Set

Validation









__**May 24, 2009**__

Training Sets

Validation Sets









__**Change Detection Statistics**__









Lab 11&12 Report