Lab 8:

Goal:

The main goal of this lab was to introduce us to the ability to use reflectance output as a means of analyzing the earth surface. We would accomplish this by gaining a better understanding spectral reflectance signatures of various earth surfaces including water, asphalt, vegetation, rocks, and soils. We need to gain a better understanding for this process if we plan to advance our skills in the remote sensing world for if we ever need to classify structure we first need to be able to read the spectral reflectance output.

Methods:

To obtain a spectral reflectance output we first needed to digitize a small area of known structure. In the first case we navigated to Altoona lake and digitized a small portion of the lake. Once we acquired our digitized area we then could run it through a raster signature editor. The editor would generate a print our of the spectral reflectance and the proportion of each band that was showing (figure 1).




Figure 1: spectral reflection signature for standing water of lake Altoona.

We then repeated this process for the following surfaces

standing water

moving water

vegetation

riparian vegetation

crops

urban grass

dry soil

moist soil

rock

asphalt highway

Airport highway

concrete surface (parking lot)

Once collected we could place all of the signatures onto the same graph and infer the following. A) what reflectance bands would be best suited for collection of these surfaces ( in this case band 2-5). B) why some of the same material had different spectral reflectance (may be caused by variation in moisture). Last why would some areas not show similar results to what we expected ( could be due to high cloud cover).

Figure 2: All spectral signatures in one graph

Figure 3: NDVI model for Eau Claire county

Section two of this lab was to introduce us to a widely used application in remote sensing. A tool used to monitor the health of vegetation. It is known as a NDVI or a normalized difference vegetation index. This index can show you the areas that are represented by high loads of healthy vegetation and does so by changing all of the band number to print out as a one or a zero. Any area that prints out as a one will be very bright white and represents healthy vegetation, while the dark greys and blacks are non-healthy vegetation or urban areas. The out put for the eau Claire area can be seen in figure 3.

The second model is of similar properties but generates an output to locate highest ferrous material. Material that is critical for mining soils. This model similar to the NDVI will calculate all of the exposed soil and place one values in all areas of high ferrous material. The output can show any user high valued areas for mining. Although one needs to be careful with this output because as shown in figure 4. The output can not analyze any areas that are covered and it can be skewed based on farmland areas that have placed ferrous material back into the soil when it may not originally be there.

Figure 4: Shows the eau Claire county ferrous material output

Results:

The results of this lab were very straight forward. We were able to better understand the use and necessity of a spectral reflectance signature. These are very useful for any user that needs to classify areas within an image.

We also where able to generate both a NDV and a ferrous model which enlightened us not only on the tools needed to run these models but the areas around our homes that have healthy vegetation and areas with high ferrous. As you can not comparing the two images there is a direct correlation to areas of low vegetation and or unhealthy vegetation and areas of high ferrous this could have been caused by lack of exposure to bare soil in the models.

Sources:

Satellite image is from Earth Resources Observation and Science Center, United States Geological

Survey.