Implica onsof Normalized Difference Using Landsat 7...

Stephanie Galaitsis TuŌs University Remote Sensing Final Project Spring 2012 Implica Ɵons of Normalized Dierence Vegeta Ɵon Index Around the Dead Sea Using Landsat 7 Satellite Imagery to Examine VegetaƟon Over Six Months March 2001 June 2001 September2001 Combined NDVI NDVI With ElevaƟon ObjecƟve: To examine whether dispariƟes of water in the region of Israel and the West Bank are eecƟng the land cover visible in satellite imagery. Normalized Dierence VegetaƟon Index (NDVI) is applied to explore the vegetaƟon vigor in the area. Conclusion: Many factors can account for the dier ence in vegetaƟon that corresponds to the geopoliƟcal boundary between the West Bank and Israel in 2001. However, without a perfect match for elevaƟon, it becomes more likely that the dierence is based in poliƟcal policies that impose constraints dictaƟng water allocaƟon between the two countries. Further work could reveal other condiƟons contribuƟng to the vegetaƟon Disparity between the two regions. Methodology: The images were downloaded from Landsat 7. March and September required preprocessing due to cloud cover in the north. The images were stacked and classied using ENVI’s Spectral Angle Mapper funcƟon to mask for all clouds and cloud shadows. NDVI is a band math operaƟon applied to the red and near infrared bands to accentuate the dierence (called the “red edge”) between these bands for healthy vegetaƟon. The resulƟng images, shown in black and white to the leŌ, have bright pixels in areas of healthy vegetaƟon and dark pixels in areas of no vegetaƟon—in this region the dark areas are deserts. Density slicing of the NDVI images shows vegetaƟon vigor changes over the six month Ɵme period of the three images. Using the same slicing for each image, with magenta as strong vegetaƟon and yellow as minimal vegetaƟon, the below images were produced. March 2001 June 2001 September 2001 Results: The NDVI images were combined, seen in the upper right, to further examine this process. Darker pixels indicate areas of less strong vegetaƟon growth, and the combined image shows the geopoliƟcal boundaries of the West Bank, shown in the upper right, with the darker pixels acƟng as vegetaƟon markers. Using ASTER data a 3dimensional view of the area, to the right, was examined to see if the demarcated boundaries corresponded to other regional terrain features. However, certain bright regions of Israel proved to be at the same elevaƟon as the West Bank, but were s Ɵll characterisƟc of Israeli vegetaƟon vigor, sugges Ɵng that elevaƟon is not a determinis Ɵc factor for the NDVI results. The images show that vegetaƟon in the west tends to be more vigorous throughout the three months examined

Transcript of Implica onsof Normalized Difference Using Landsat 7...

Page 1: Implica onsof Normalized Difference Using Landsat 7 · Normalized Difference Vegeta Indexon (NDVI) is applied to explore



Stephanie Galaitsis 

Tu s University 

Remote Sensing 

Final Project 

Spring 2012 


Implica ons of Normalized Difference  

Vegeta on Index Around the Dead Sea 

Using Landsat 7 Satellite Imagery to Examine 

Vegeta on Over Six Months 


March 2001 

June 2001 


Combined NDVI 

NDVI With Eleva on 

Objec ve:

To examine whether dispari es of water in the  

region of Israel and the West Bank are effec ng the land cover visible in satellite imagery.  

Normalized Difference Vegeta on Index (NDVI) is applied to explore the vegeta on vigor in the  



Many factors can account for the differ‐ence in vegeta on that corresponds to the geopoli cal boundary between the West Bank and Israel in 2001. 


However, without a perfect match for  

eleva on, it becomes more likely that the difference is based in poli cal policies that impose constraints dicta ng water al‐loca on between the two countries. 


Further work could reveal other condi‐ons contribu ng to the vegeta on  

Disparity between the two regions. 


The images were downloaded from Landsat 7. March and  

September required preprocessing due to cloud cover in the north. The images were stacked and classified using ENVI’s Spectral Angle Mapper func on to mask for all clouds and cloud shadows. 


NDVI  is a band math opera on applied to the red and near ‐ 

infrared bands to accentuate the difference (called the “red edge”) between these bands for healthy vegeta on. The resul ng images, shown in black and white to the le , have bright pixels in areas of healthy vegeta on and dark pixels in areas of no vegeta on—in this region the dark areas are deserts. 


Density slicing of the NDVI images shows vegeta on vigor changes over the six month  me period of the three images. Using the same slicing for each image,  with magenta as strong vegeta on and yel‐low as minimal vegeta on, the below images were produced. 

March 2001 

June 2001 

September 2001 



The NDVI images were combined, seen in the upper right, to further examine this process. Darker pixels indicate areas of less strong  

vegeta on growth, and the combined image shows the  

geopoli cal boundaries of the West Bank, shown in the upper right, with the darker pixels ac ng as vegeta on markers. 


Using ASTER data a 3‐dimensional view of the area, to the right, was examined to see if the demarcated boundaries corresponded to other regional terrain features. However, certain bright regions of Israel proved to be at the same eleva on as the West Bank, but were s ll characteris c of Israeli vegeta on vigor, sugges ng that  

eleva on is not a determinis c factor for the NDVI results. 

The images show that vegeta on in the west tends to be more vigorous  

throughout the three months examined