Satellite imagery from the Landsat 5 Thematic Mapper sensor and the Landsat 8 Operational Land Imager were used to investigate changes in overall evergreen vegetation occurring between the 1986-1989 and 2013-2014 time periods. Two path/rows of imagery, from the spring, summer, and fall seasons for each time period were mosaicked together. The imagery was then subset to remove the presence of clouds from the datasets. Images were further subset using the impervious data from the 2011 version of the National Land Cover Database. Unsupervised classification was used to spate each time period image into two classes, evergreen vegetation and everything else. Each subsequent time period was subjected to successive unsupervised classification routines until confused classes were minimized. The resultant classifications were then used to model a raster data layer that included four classes of change information, Change from Evergreen, Change to Evergreen, No Change – Evergreen, and No Change – Not Evergreen.
The following image data were used in the analysis. Images are listed by World Reference System II path/row designations acquisition date.
2735_011287.img
2735_012214.img
2735_031114.img
2735_031486.img
2735_032289.img
2735_062813.img
2735_072086.img
2735_072387.img
2735_110986.img
2735_111287.img
2736_011287.img
2736_012214.img
2736_031114.img
2736_031486.img
2736_032289.img
2736_062813.img
2736_072086.img
2736_072387.img
2736_110986.img
2736_111287.img
The entirety of each scene may not have been used as the aim was to cover the largest areal extent while minimizing the impact of cloud cover. Therefore, portions of scenes used to fill in where cloud cover existed in other image dates. Even though this process was used, complete coverage of the 27/35-27/36 was not possible due the limited number of scenes available within the prescribed time periods.