1.
I began the
module, by reading the introduction provided in the map assignment.
2.
I clicked on the
earth explorer website link, where I saw that my account from the previous
semester was already logged in.
3.
In the search
criteria window, I changed my criteria to decimal degrees, and clicked on the
blue Add Coordinates box where I copied the degrees provided. I saw that a
marker appeared on the map, somewhere within Central America.
4.
I clicked the
“Data Sets” button.
5.
In the data sets
window, I navigated to “Landsat” and saw a fairly extensive set of sub-menus. I
attempted to locate the mentioned ‘Landsat 7 ETM + SLC-on (1999-2003)’ option, but it was not there. Instead, I selected
Landsat 7 ETM + C1 both Level 1 and 2. It was the closest name to the set
mentioned in the lab.
6.
I viewed both
sets of the data, and was not able to locate the specific aerial image for the
date mentioned in the assignment. Because all of the Landsat 7 level 2 data
required payment, I downloaded the most recent image from September 5th
of 2018. I thought this would be the most appropriate, as the assignment
mentions that the most recent data ought to be used in most cases.
7.
I downloaded the
highest quality GeoTIFF file, and saved it to my raster folder. I named the
file Landsat1Level1.
8.
The file
downloaded in a strange .gz format. I had to unzip it twice, before I was able
to see all the .tiff files contained within.
9.
I opened a new
MXD file, and named it AZajacMaya_PyramidsWk1.mxd as the directions required. I
saved the MXD to my MXD folder within Module 1.
10.
I added the Imagery basemap to my blank MXD file, from the drop down Add Data
menu.
11.
I extended Arc Catalog, and connected to my Module 1 folder. There, I right
clicked on my Shape sub-folder and selected Create new Personal geodatabase. I
named the Geodatabase Maya_Pyramids according to the directions.
12.In
the ArcMap main menu, I clicked on Windows and selected Image Analysis. A new
window appeared, which I have tagged to the right side of my screen.
13.
In the main menu, I clicked on the Customize Menu and selected Toolbars, then
Image Classification. I placed the toolbar in a free space at the top.
14.I
clicked the Add Data Button and navigated to the raster images that I
downloaded beforehand. Because my Landsat 7 image was different, I obviously
could not open the very same image that the instructions suggest. I did however
select the 8th band and opened it. When asked if I wanted to build
pyramids, I clicked yes.
15.I
took a bit of time to examine the Landsat image, and to compare it to the
Imagery basemap.
16.I
tried it on 1:30,000, and 1:8,000 and tried to see the contrast. It was
difficult however, to see detail on the black and white landsat image.
17.
I went into the layer properties symbology tab, and clicked “Display Background
Value 0” as no color. I have also noticed, that my Landsat band is visibly
damaged, and contains missing patched throughout.
18.
I added the Mirador point shapefile, and zoomed to it. I made it appear as a
large yellow dot, and I took time to inspect what lies underneath. The mound
was not very easy to see on the landsat aerial, but was very visible over the
imagery basemap.
19.
After inspecting the image, I followed the directions provided in the lab to
pull up the Composite Bands tool in the Data Management folder. For the input
files, and I navigated to my Raster folder and selected the first three bands,
in the appropriate order. I ran the tool, and the newly created raster was
added to my MXD. (the result was saved to my geodatabase prior.)
20.
I entered the layer properties, and went into the display tab. There, I changed
the band numbers next to the designated colors, (Red: 3 Green: 2 and Blue: 1)
as they were indeed backwards. I have also selected to display the value of
zero as no value at all. I pressed apply, then OK to display the image with the
new the new settings.
21.
I tried to pan-sharpen the image, but
was not able to do it. I made sure that my settings looked exactly as they did
in the example provided in the lab and I was still unable to check the box. The
problem was not that the option was not available, it seemed to be (the text
next to the box appeared in black, as if it was clickable.) But it was simply
not clickable. I have no idea what is going wrong here, but I was not able to
pan-sharpen the image.
22.
I zoomed my map to the Mirador point file.
23.
I right clicked on my Nat_Clr layer in the Image Analysis window, and selected
the Accelerate option.
24.
I right clicked on the Nat_Clr layer in the Image Analysis window once more, and
opened the options by clicking on the icon in the top left. I went into the
NDVI tab, and made sure that red is set to 3 and infrared band is set to 4. It
was.
25.
I clicked over onto the Pan-Sharpening tab, and inspected the settings. They
seemed to correspond to what was demonstrated in the lab.
26.
I moved on to the orthorectify tab, and took a look at the settings. Without
changing anything, I moved on according to the lab directions.
27.
I did the same with the Hillshade tab, and after ensuring that the options are
the same as shown in the example I moved on.
28.
I moved on to the Stretch tab. I examined it, and moved on leaving the default
settings.
29.
I moved on to the Display settings. I tried to move the shade around to 45,
then to -45. I shifted the Gamma settings back and forth and reset them to
default at the end. I proceeded to inspect all the tools in the Image Analysis
window as outlined in the directions.
30.
I right clicked on my table of contents, and selected to add a group layer. I
renamed it “Landsat”. I right clicked on the new layer, and added the 9 layer
bands to it. The program asked me if I wanted to build pyramids each time it
recognized a new band. I clicked ‘Yes’ each time, and eventually all black and
white Landsat images appeared under the new group layer. I minimized all of
them in order to be able to see the table of contents better.
31.
I examined the table provided in the third part of the exercise, and pondered
the wavelengths of each band.
32.
In the image analysis toolbar, I selected bands 2, 3 and 4 by holding the shift
key. Then, I pressed the layer compression button which became highlighted.
33.
A new temporary layer was added to my table of contents, but the grass appeared
blue in color. When I went into the layer properties, each band that appeared
was named band one. Either way, I selected the third “band 1” for my red color,
and the first for my blue color. It seemed to have worked just as well, and my
raster gained the near-infrared appearance.
34.
I renamed the new layer False_Color, and exported it to my database. The new
exported layer was added to my map. I had to re-set the bands once more, but
this time they were appropriately named Band 1, band 2 and band 3.
35.
I went back to my Image Analysis toolbar, and attempted to shift the settings
around to improve the image quality. After resetting everything back to zero, I
clicked the DRA button to normalize the image at current extent. The option
seemed not to help much.
36.
I feel that I understand the processes explained in the lab, but I found
difficulties in the simple facts that 1: I was unable to acquire the actual
imagery needed for this lab, and had to work with a seemingly damaged raster
that did not provide a full view of the landscape, and 2: My pan-sharpening
option would not work, and I do not know why. Because of this, the pixels
remained large and the image was very vague when zoomed in.
37.For
next weeks lab, I intend to track down better imagery to work with and possibly
make more compatible raster files.
38.
I took time to create an aesthetically sound JPG, and exported the file.
At the very end, a similar layout was generated on Angkor Wat in Cambodia.
I created a new MXD file, and accessed Earth Explorer to
download Landsat data for the general Angkor Wat region. While I was able to
use the image analysis window in ArcMap to combine all the appropriate bands, I
found that the general look at the area did not yielded a whole lot of
information to me. After zooming in closer, I saw that I was able to make out
the canals associated with the site. This is where the False Color, and the
SWIR layer came truly in handy.
It is difficult to estimate the size of the entire “island”,
but it seems that the stone temples take up only about 20% of the site. Most of
the area is still covered by jungle, and surrounded by water. Though there is
quite a bit of anthropogenic features surrounding the entire island, much of it
seems quite untouched.
