Project 2-GIS & Landscape Design/Management

Part 2-Analysis

This phase of the project utilized a number of the Spatial Analyst tools. The biggest challenges came with the  processing phases.  Initially the image files (photos) were evaluated using the Isocluster tool, followed by the software using the Maximum Likelihood evaluator.  This was to classify the images into 50 classes. Initially this process failed many times.  I was able to get it to process using 25 classes.  Later the instructor provided a 50 class raster using the mosaic tool.  I then changed my strategy from using the 7 images I had converted to using the provided mosaic file. The next steps involved identifying the pixel values to determine the types of land classes.  The goal being to classify into 3 classes, either Trees, Grass or Impermeable surfaces. I then used the Reclass tool to classify the 50 classes into three classes.  This process ultimately failed over 15 times, however through a consultation with my GIS internship supervisor, I found some work-arounds using the Extract by Attributes.  After this Extract by Attributes the software had no issues reclassifying the raster to 3 classes. Another part of this whole process involved using the Extract by Mask tool to make the project relevant to just the neighborhoods of the study area. Finally the tables were manipulated using the field calculator to compute percentages of trees, carbon storage and carbon sequestration, followed by constructing the graphs to add to the maps.
Despite some processing frustrations, I learned a number of things about the Spatial Analyst tool set and classifying and reclassifying rasters

Project 2-GIS & Landscape Design/Management

Part 1-Base Map of Marin Study Area

Base Map of Marin, California study area.

Project 1-Asthma-San Francisco Bay Area

Analysis of Asthma Hospitalization and Correlation with Air Quality and Race

This project was iniated for a presentation to a county board, hospital or other government officials.  It is to show asthma hospitalizations for the San Francisco Bay Area.  It seeks to find which counties, or ethnic groups may be more susceptible to asthma hospitalizations and therefore should receive more government or private funding. Factors shown for this determination include, total number of hospitalizations, county populations, air quality, race distribution and uninsured/unemployment rates.

Observations & Conclusions

Alameda and Santa Clara Counties have the largest populations. The largest ethnic group is Hispanic. The group with the most hospitalizations for Asthma is the African-American group. The most African Americans are concentrated in the Oakland, California area of Alameda County. The closest hospitals to this area are Oakland East and Oakland West hospitals.

It is this observers point of view that if funds were to be allocated. Alameda & Solano Counties should receive the dollars as they have large African American populations. These counties are also relatively uninsured and have higher unemployment. The air quality in these counties also are the worst in the Bay Area and which possibly points toward a correlation to the higher asthma hospitalization rates in these counties.

Some Issues:  My hospital files seems to show varying amounts of hospitals.  I think it has doubled up in some cases.  I also had problems computing and showing "proximites" to hospitals in the Oakland, CA area due, I believe to the tainted hospital files.



Project 1-Prepare

Air Pollution, Asthma and Race in the San Francisco Bay Area

Links to Metadata files as created:


Link 1
Census Table PCT-6


Link 2
Asthma Hospitalization Rates


Link 3
Ozone Levels Bay Area 2004


Link 4
Particulate Matter Bay Area

These were the only files I created Metadata for as the other files in my ArcMap drawing already had Metadata created. A learning experience navigating around the Metadata Editor...............I'll be quicker next time!

Module 5-LIDAR

LIDAR-Interpretation-Pensacola Beach, FL

This weeks Module is on LIDAR.  The challenge assignment was to take the raw LIDAR data and produce a map using the IDW tool in ArcMap.  After manipulating the image, the assignment was to identify a road, sand dune and water feature.  This seemed straight forward as one could key on the relative elevations and the somewhat straight line for the road.  I'm sure there is more to LIDAR interpretation, but this was a basic start.

Module 4-Supervised Classification

Supervised Classification of Germantown, Maryland

This week's Module consisted of classifying land cover types in Germantown, Maryland.  This was done through a supervised classification method utilizing ERDAS software.  The project was mostly straightforward but I did have to work through the signature classes a few times to make sure I didn't have any dissimilar land cover types.  Overall this seems like a really a very necessary set of skills to know for a Remote Sensing specialist.  The biggest frustrations came with the system crashing multiple times in the exercise and assignment, but in the end I believe it came together.

Link to XPS drawing:

XPS Drawing

Module 3-Orthorectification

Rectify Image of Pensacola Bay

This lab was educational and it was relatively easy to understand the concepts as the lab instructions were quite helpful.  The items that took the most time were the continual crashing of the system, playing with the RMS values for the Ground Control Points (GCP's) and trying to get the resolution of the original image better.  I never was able to get the image to display better and it was a little tricky to get my first 3 control points placed with a low RMS.  I was able to identify some nice features in the reference and input images but the resolution was so poor it took a few wild guesses to get it close.  My overall RMS was 0.658. 

The ERDAS software is real quirky and it seems like we are missing some tutorials or other key information that may help?

The links to my XPS Drawing and Excel file is below:

XPS Map

Excel file