Methodology

For the purposes of this study, mangrove forests within the Niger River Delta are studied ranging from latitude 4°10' to 6°20' N to longitude 2°45' to 8°35' E. Mangrove forests are threatened worldwide, and within the world's largest delta it is not hard to see the vast amount of mangroves currently under threat in Nigeria (Nwilo/Badejo). Mangroves in the Niger River Delta are under threat from a multitude of sources including their main threat, that of humans. Currently, humans have a large stake-hold within the delta as it is an untapped resource of crude oil and natural resources. Oil companies have thus made Nigeria their "home" and continue to deplete the land of its resources. With oil production comes environmental degradation and within the delta the mangroves are suffering. The area of mangroves within the delta is decreasing and with the use of remote sensing we can see the rate and extent as to which the mangroves are being destroyed.

In using remote sensing for this project, Landsat TM and ETM+ images were used from the Landsat 7 and Landsat 5 satellites. The project portrays a time series of the delta from images from December 19, 1986, December 17, 2000 and January 3, 2007. Landsat images were obtained and downloaded from the Global Land Cover Facility (GLCF) at the University of Maryland. In total 3 Landsat images were used.

These images were then loaded using ENVI 4.8, a software that is used for remote sensing manipulation.   Bands 1 (blue), 2 (green), and 3 (red)  represent the image in the visible electromagnetic spectrum.

1986 Visible

2000 Visible

2007 Visible

In order to accurately view the Landsat 7 ETM+ image from 2007 bad values had to be replaced in order to "erase" the striping in the image. See below.

2007 ETM+ with bad values (stripes)

Next the images were loaded using the layering of the red, near-infrared and mid-infrared bands (3,4,5). Layering the bands in a specific fashion of 4, 5, 3 RGB (Red-4, Green-5, Blue-3) helps produce an image that allows for the easy viewing of mangrove vegetation (dark salmon color). This is what is known as an unsupervised classification which helps clearly define the vegetation, i.e. mangroves in these images. Unsupervised classifications group together pixels of similar spectral properties into distinct classes. The images below show the images in an unsupervised classification of RGB 4,5,3.  Mangrove forests turn a salmon color and are thus easily distinguished from land and other types of vegetation. In the images themselves blue represents water, yellowish/orange represents land and other types of vegetation and the pinkish/salmon color represents the mangroves.

1986 RGB 4,5,3

2000 RGB 4,5,3

2007 RGB 4,5,3

 Furthermore, by conducting a NDVI or a Normalized Difference Vegetation Index we are provided with a greeness index where we can see the vegetation in the study area. Negative NDVI values that approach negative 1 represent water, values close to zero represent barren areas of rock, sand or snow and low positive values represent vegetation. Below is an example of an NDVI for the 1986 image. Values represented in white are vegetation whereas the black refers to that of water and land forms.

1986 NDVI

After conducting a NDVI then we can use density slice overlays which use color theory in showing the different types of vegetation present in the image. Color theory is extremely helpful and useful when analyzing these images because it allows the researcher to see the extent of vegetation as well as different types of vegetation present. In addition to vegetation, density slices also show the type of water present in the delta whether they be fresh, salt or polluted water sources. Density Slice images can be found under the "Results" section of this blog.

Lastly, in interpreting the loss of mangrove forest within the Niger Delta change detection analyses were conducted. Change detection analyses were conducted from 1986-2000, 2000-2007 and then overall from 1986-2007. Change detection helps show the change in the extent of vegetation over time. In the images blue represents a decrease in greenness whereas red represents an increase in greenness. Different shades of the colors, particularly darker shades intensify the increase or loss of greenness. However, it must be noted that in these images the red that is located in the rivers and tributaries represent an increase in polluted water. Also, due to cloud images in the 2000 images the intensified color is an anomaly. Change detection analyses images can be seen in the "Results" section of the blog as well.

Images using the thermal bands were also used to help detect the location of gas flares. These images can be found in the "Results" section as well.