Week 10: Protecting Critical Infrastructure

This follows on from last week. This time we had to use the elevation data and create lines of sight from the NORAD tunnel entrance at Cheyenne Mountain, Colorado for people protecting the entrance. It was quite interesting to see this work, although my 4th deliverable, supposed to have been created in ArcScene (the 3D view), only would work in 2D as a jpeg and I'm not even sure I got it to work correctly.

However, in 2D, here are a couple of images: the grey one with the red-and-green lines shows an aerial view of the NORAD tunnel entrance with lines of sight (green = visible, red = impeded). The other one is a conventional map view with a 3-mile buffer round the NORAD site and a 500-foot buffer round the nearest airport, or rather, heliport.

Week 9: Data preparation for Homeland Security

This week we had to prepare a series of data layers for next week's Homeland Security lab. In fact much of the work was done for us, but there were still some challenges. I didn't have trouble unzipping data or finding enough space on my H drive (although I may soon run out of space), but the challenging part was more about working backward from the end of the instructions (within a given section) when I couldn't make complete sense of what the end of the section should be. The road section directions were a problem for me so I ended up doing much of the original directions to get to the same endpoint. The lab was useful for getting an better understanding of how to group layers, and I am delighted that I was able to rectify the projection of one raster that initially was a long way away from the site in Colorado. And finally, I am glad that we are learning about how to organize the data and the environments within ArcGIS.
One map is just the screenshot which is intended to display the group layers that we created (or manipulated). The other, which is a "real" map, i.e. containing layout items, is a larger-scale map showing the location of the NORAD site in relation to the elevation. I didn't want to change the labels of the group layers in case I didn't change them all back correctly, so they are labeled with my initials as per the layers.

Week 8: Crime in Washington D.C.

This week we did proximity analysis to examine, first, various categories of crime in August 2009 in Washington D.C. and the relationship between the crimes and the location of police stations, and second (shown here) where juvenile crimes were committed in 2009 in D.C. and how that related to the 1000-foot-radius "drug-free" zones around the schools that Washington has mandated. In fact I am uncertain about whether the drug-free zones are effective as drug-free zones, since the vast majority of the offenses listed are non-drug-related. Moreover, we don't have data that describes the levels and distribution of crime before the drug-free zones were established. Given that 1304 offenses are listed as having been committed within the 1000-foot buffers round the schools, they don't seem to be fully effective as "crime-free" zones. That is an average of 10 offenses per school across the year, or a bit less than one per month - not a high number, but not zero either.
In the map I thought it would be interesting to see whether there are higher rates of crimes around high schools than elementary or middle schools, which is why I symbolized the schools differently. However there does not seem to be much relationship: The highest numbers of offenses are near two high schools and one elementary school. Several other high schools have low rates of crime, and the rest of the elementary and middle schools have either low or moderate levels of crime.

Week 6: Location Decisions

For much of this lab I felt very frustrated - suffering from jet lag and working with a Really Slow laptop. Also I didn't discover a vital starting point for Project 2 until I had completed almost all of Project 1 (in an effort to discover the starting point for Project 2) - and I have run into more than the usual roadblocks. However I did get to the end in time. So...

This lab was about a hypothetical couple making a decision to relocate to Alachua County, Florida to be closer to family, with other factors also included. The lab involved calculating distances, reclassifying the distance calculations, and producing weighted overlays that would come up with the "best" census tracts in the county for this particular couple.

The base map (not shown ) is of various features in Alachua County that the hypothetical grandparents (with grandkids who live in the selected-in-red census tract) might want to be close to. This was a fairly basic map.

The second map (see top right) was probably the most complicated one I have done so far. I still need to learn how to work with map templates, as the arrangements were complicated for me and I removed all the four data layer displays in order to make the succeeding map. This second map (one map with five layouts, or five layouts with one map, don't know which) shows the distance from various features that the grandparents might be interested in - North Florida Regional Medical Center, the census tract containing their grandkids, bus routes, community centers, and the University of Florida.

I eventually got the models to run correctly - or at least to run and to display and to acknowledge that they existed, which initially they refused to do - I briefly posted a screenshot on this blog that said ArcMap said the model had run successfully, as proof in the absence of any other! I am troubled by the models though because when I reversed the scales, the bottom number (the 9) wouldn't reverse. But I couldn't get the model to run properly when I tried fixing those numbers in the model prior to running it, so eventually I left the numbers alone and just reversed them with one click. The map at top left shows the final weighted overlays from the models.

I am still a bit confused about whether I have done the reclassification correctly. Not all the distance calculations yielded the same distance per category, so I ended up with some smaller numbers of categories. I'd like to understand this better but after spending about 20 hours on this+project 1 (I blame this partly on the laptop; I didn't have a chance to print out all of the pdfs before I left town and was unsuccessful in trying to connect remotely on other people's computers so that I could print the pdfs elsewhere, so for this lab I have been reading the project1&2 pdfs on screen, which is dismayingly slow to advance each part of a page. Hence I didn't see the note about which census tract to use for project 2 until I had worked my way almost all the way through project 1 looking for the answer), I didn't feel I had the time to ask more questions about it.

On the plus side, I learned a lot in this lab and I feel good about managing to get a lot of information in fairly clear format on to one page. I am grateful for the information about setting environments and organizing data within ArcMap. (Now, about those map templates...)

Week 5: Impact assessments

This week was more relaxed, fortunately. The readings were about urban planning and impact assessment. The two maps shown are from an ESRI lab. The first shows location quotients for agriculture/forestry/fishing in local government authorities within the planning area of mythical Pewter City; the other shows the occupancy rate of university students, also in Pewter City. Location quotients are familiar to me because I did a project comparing the demographics of areas around greenway trails to those in towns where the greenways were found - essentially calculating greenway location quotients. So you would think this was easy, though I initially found it confusing because of all the acronyms, especially labels like AG_FRS_Ref for "agriculture, forestry and fishing in the reference area". Eventually my brain kicked in and I was able to complete this one. ESRI labs are generally very straightforward and this one was too, although they don't make you think nearly as hard as the regular labs. But that's a good thing for me this week.

Week 4 Oil Spill - GIS response summary, and Animation

The animation was part of our participation exercise and produced a six-frame animation of the extent of the oil spill, from 29th April 2010 to 26th May 2010.
The link to the animation is here: http://students.uwf.edu/db27/OilSpillExtents.avi

The hard part for me was getting the projections to line up. I finally did that by clearing the projection information for each oil extent shapefile individually, redefining the geographic coordinate system to what it said it was before, and then projecting it to the same projected coordinate system that it said it had before. Don't quite know why this worked, but I was thrilled to see that it did. The background files were unprojected, and when I projected them to the same coordinate system (that the oil spill files said they were originally, NAD UTM 16N), it all worked together. But not until I had tried every other unsuccessful combination of projection-on-the-fly, adding the (first projected) background files and then adding the (apparently projected) oil spill files, etc. etc. Once it worked though it was great.
An improvement on this animation would be to have the date of each oil spill extent display. I tried labeling in ArcMap, for the states and a label of the range of dates, but in animation all these labels continually flashed on and off, so I deleted the labels. There must be a better way!

On the role of GIS in disaster response, specifically in regard to the Deepwater Horizon oil spill in the Gulf of Mexico:
GIS has an essential role in disaster response, and its role is immediately visible to us in the oil spill maps in the New York Times and in this week's lab assignment. Below is a list of GIS involvement in disaster response, in general, and in the case of BP oil spill. After a disaster we can expect GIS to play an integral role in:
1) Showing the location of the event and the spread of its impace. In this case, GIS has been used to map the oil spill, both as oil on the water and as oil plumes.
2) Showing the location of vulnerable populations and their habitat. This could be people in homes at risk from a flood or an earthquake (and should have been prepared before an actual disaster). In the BP oil spill, the vulnerable populations include marine life, birds, shore animals. It also will include information about fishing grounds and other places (see below) at risk since the livelihood of fishermen and others whose livelihood depends on the ecology of the Gulf is also threatened.
3) Showing locations where there is damage or harm, to people and/or habitat. After an earthquake this could be where buildings are damaged; since the oil spill GIS has probably been used to map where oil-soaked birds or marine mammals have been rescued or found. Such mapping also gives more specific and recent information about vulnerable population whereabouts.
4) Showing the location of measures taken. In the case of the oil spill, examples might include where oil booms are, or are recommended (and whether they are one or the other) and information about cleanup (see below)
4) Showing where assistance centers are, from hospitals to FEMA offices to animal shelters. In the case of the oil spill, it might include FEMA offices, wildlife rescue sites, places for volunteers to go, other government cleanup staging areas, etc.
5) Educating the media in all sorts of ways: about where the damage is, the extent of the cleanup, etc. etc. An oil spill example is the GIS map of the extent of the closed-to-fishing boundary, an updated version of which has been shown daily in the New York Times.
6) Helping to direct the cleanup, in terms of recording where what is needed and what it will take in materials and people and costs, as well as recording locations where workers have gone to clean up and the status of those efforts. In the case of the oil spill, this would include a lot of coastal information about the booms, the wildlife cleaning and marsh protection efforts, and more.

Week 4: Oil Spill in Gulf of Mexico

This week we had two tasks. The first was to display in Google Earth the area closed to fishing on two dates (25th May, 1st June) after BP's oil spill in the Gulf of Mexico, and to calculate the difference in square miles between the two areas. The second, some maps from which are shown here, was to create maps of several environmental sensitivity and environmental impact/effort indices for a chosen section of the coastal panhandle of Florida. I chose an area that encompassed parts of Gulf and Franklin counties.

The first map (or rather the one at left) shows types of shoreline, from sand to wetlands, coded according to their sensitivity to an oil spill. A second map (top left)shows some of the creatures at risk from an oil spill, and the areas where they are likely to be found. Another map (top right) shows federal and state jurisdictions, as well as boom operations (already staged and only proposed: only one boom of 23 was coded as staged). Shoreline on only one of the two counties is under federal or state jurisdiction, which might create conflicts at a time of crisis management like this.

My biggest technical problems were in downloading an elevation file (not shown, though it did eventually appear) and in creating a graph of the different shoreline types and lengths. I did produce the graph, but I couldn't get it to display colors that matched those in the map, apparently because ArcMap will only do this if the table from which the map is produced is spatially defined. I couldn't figure my way out of this one.

Parts of the lab were frustrating but it was tremendously interesting to be dealing with such a current and urgent topic. It was also enlightening to see how much preparation has been done to be aware of the aspects of the environment that are at risk and to be ready to prevent or at least mitigate that risk.

O