GIS: The What and Why

I've found that the most common response to the news that I'll be studying GIS this fall is "What's GIS?" It's a frustrating question for me, since I usually stumble my way through the explanation -- and likely frustrating for the listener too who has to suffer through the stuttering. The problem isn't a poor grasp of the subject matter, but having to decide what definition of a complex topic to choose, how to tailor it to the level of understanding of whoever is asking, how much time I actually want to spend explaining it and how thorough an explanation they really want to hear. On this occasion, no one is explicitly asking the question so I can take the chance to define it and its importance on my own terms, as I understand it.

Almost all data can include an element of location -- address, postal code, geographic coordinates. A geographic information system (GIS) allows us to integrate, manage, analyze and display this geographically referenced information. Depending on the knowledge level and objectives of the user, a GIS can take on many definitions:

• a container of digital maps
• a computerized tool for solving spatial problems
• a digital inventory of geographically distributed facilities and objects, in the case of utilities management, transportation, or resource management
• a tool for visualizing geographic information to reveal patterns or relationships that are otherwise invisible
• a means of performing operations on spatial data that would be too complicated or tedious if performed by hand

What Can It Do?

Geography plays a role in nearly every decision we make: choosing sites for facilities or services, targeting market segments, planning distribution networks, evaluating emergency response routes and providing monitoring during and after an emergency, redrawing political boundaries. All of these problems involve questions of geography, as do countless others. So, as it has been said: "The application of GIS technology is limited only by the imagination of those who use it."

Some concrete examples pulled from this site:

• showing distribution of food banks and shelters in relation to low-income populations to determine whether more services are needed and where
• identifying where recommendations for environmental improvements have been made within a watershed to determine gaps
• development of efficient delivery routes for transportation firms
• plotting climate and land characteristics in order to identify possible locations for lost settlements of the norsemen in North America
• plotting contaminated underwater streams to see if they affect water wells for residents
  
• evaluating emergency response routes and providing ongoing monitoring during and after an emergency (used extensively for recovery effort in Florida during and after Hurricane Andrew in August 1992)

In general, as I see it the importance of GIS is its power to inform better decisions. The ability to integrate vast amounts of different types of data (with location as the unifying element), then visualize, query, and perform analyses on the data, allows the user to see patterns that might otherwise be invisible and predict the outcome of a given decision. This capability -- simulation of a particular decision to assess its effects before actually implementing it -- is and will become more important as we address environmental problems and increasing resource scarcity. Things must be done right the first time around, in the most efficient and least wasteful/harmful manner possible -- be it urban planning, construction, manufacturing, etc. Although counterproductive in a monetary based economy, where it's in the firm's best interest to design a product or service that will soon become obsolete to ensure continued business, we must soon get in the habit of using resources responsibly, doing things perfectly the first time around so they don't need to be redone a few months or years in the future.

Songo Songo Island Map

Inspired in part by the work of Peter Minton at EVS-Islands, earlier this week I put together a map of Songo Songo Island, a tiny island in the Indian Ocean off the coast of Tanzania that I had the good fortune to visit last summer. Administratively, the Songo Songo ward is part of the Kilwa District, Lindi Region of Tanzania. The Songo Songo group of islands is best known for the significant underlying natural gas field. The main island is the center of the Songo Songo Gas-to-Electricity Project, which includes on-island gas processing facilities and a pipeline network to transport the gas to Dar es Salaam where it powers five gas turbine electricity generators.

Despite the infrastructure present, walking along its white sand, palm tree lined beaches, the island seems remote and untouched, as one might expect of a small island in the Indian Ocean. The Songas facilities are confined to a small corner of the island and are essentially invisible unless you purposefully seek them out. Children from the settlement on Songo Songo initially shy away at the sight of visitors, but curiosity and playfulness gradually win out over shyness. One afternoon sitting on the beach we found ourselves literally encircled by a dozen smiling, laughing faces. We couldn't convey anything of importance to them with our limited Swahili, but they seemed amused nonetheless to be in our company, just as we were to be in theirs.

Click here for a larger image

Working with Quantum GIS, I started with a Landsat7 ETM+ image of the island from which I could digitize all the features that would eventually make up the map. I created shapefiles of the shoreline, sand, vegetation, built areas, roads and paths, and reefs. It would have been nice to find a shapefile of contour lines for the island, or digitize contour lines myself from a DEM. But there is very little change in elevation on the island, so it's not the end of the world. I did try to include a scale bar in kilometers, but the software wasn't doing exactly as I wanted it to -- still figuring things out. The island is about 4km long and 1.5km wide at its widest.

Walkability Map of Listowel, Ontario: Cartastrophe or a Good Start?

A walkable community is one where all services and amenities are close at hand to residential areas (or integrated into residential areas in mixed use development), eliminating the need to use or ideally to even own a vehicle. The benefits of such walkable neighbourhood arrangements to our health, the environment and our communities are obvious:

• It's been shown that residents of pedestrian-friendly neighbourhoods are healthier and weigh less than their sprawling neighbourhood counterparts, and because of driving less suffer few car accidents.
• The environmental problems associated with vehicle emissions are no secret nowadays: poor air quality, greenhouse gases and global warming. Designing lower emission vehicles, despite all the hype and marketing from car companies, is not the solution. (Is smoking light cigarettes a good idea? No, kick the habit!) Well designed towns and cities that eliminate our reliance on vehicles and fossil fuels are what's desperately needed. Walking and cycling are zero-pollution means of transportation.
• Compactness and density are essential conditions for a walkable neighbourhood. A higher population density makes public transportation and bicycle infrastructure feasible. Bus and rail routes and bike paths/lanes can serve a greater number of people. Dense, walkable neighborhoods also provide local businesses with increased foot traffic, allowing them to thrive. Not having to drive between destinations allows pedestrians to stop at more stores/restaurants/theaters/etc.
• A walkable neighbourhood promotes face-to-face interaction and community involvement. Instead of being couped up in a vehicle en route, isolated from the world, the morning commute could be a bike ride alongside a neighbour or a trip to the grocery store a short walk in the fresh air, encountering friends along the way.

Now, on to the map. It's quite simplistic and possibly full of cartographic faux pas', being a first attempt, but it hopefully conveys the message with as little confusion as possible. I used Quantum GIS to put together the map, a free to download desktop GIS application. The whole process was quite instructive, having never done it before. I searched for and downloaded the datasets, chose an appropriate (hopefully) projection for my area of interest (UTM, Zone 17N), imported the layers into QGIS and decided on color schemes, transparency, and other design concerns. I also made use of the "buffer" function to create an area around the town's bike paths to be displayed -- a linear park essentially -- and to create 500m radius "walkability" circles around the supermarkets and schools, two of the most frequented locations in a daily routine. Rather than using buffer circles which don't account for the actual layout of roads, a walking time cost analysis showing along-street displacement times from the location in question would have been more appropriate. Unfortunately that's still over my head at this point, and I'm not certain that QGIS has that capability. Well, here it is...

Click here for a larger image.
Comments and criticisms welcome.

This Weekend's Bike Trip in Google Maps

Resting and relaxing after this weekend's bike tour to Kincardine to visit a couple friends -- lots of sun, sand (no surf unfortunately), good company, and a drink or two. Despite a headwind on the way there, mother nature served up perfect weather for making the round trip from home to Lake Huron -- 211 km total. Though not the most direct options possible, my choice of there and back routes kept me on pavement but off the busy main highways of the area. The southern route is slightly shorter and very flat most of the way, making for a low degree of difficult. The northern return route adds 10kms and passes through hillier areas, but is more scenic. Since both routes only pass through only a handful of small villages, finding food or beverages on a Sunday or at odd hours could be a problem. Plan accordingly if you're passing this way. Here's a Google map of the trip:
View Atwood - Kincardine - Atwood (Ontario, Canada) in a larger map

Last Chance for the Tour de France

Even if you object to pro cycling on the grounds of alleged widespread doping, the spectacle of the Tour is still hard to ignore: a pack of 200 cyclists winding through French countryside for 3 weeks in July, ascending mountains in the Alps and Pyrenees, then flying down them at speeds approaching 100km/h, sprinting to the finish of a day's stage after 200kms+ and 5+ hours in the saddle.

Only 2 days/2 stages remain in this year's Tour, but they promise to be 2 of the most interesting. Tomorrow's stage ends with a climb up Mont Ventoux, the so called "Giant of Provence". While Alberto Contador's advantage over the rest of the field is almost unassailable at 4 minutes 11 seconds, the other 2 podium spots are still undecided -- expect some drama and aggressive attacking and drama tomorrow. Sunday the riders arrive at Paris and finish the Tour, as always, on the Champs-Élysées.

Thanks to hardcore fans of the sport who have put in some extra effort, nowadays there are lots of ways to interact with the Tour, the route, the individual stages. At the paris.thover.com blog there is a KML file of the full route of this year's Tour, which opens in Google Earth allowing you to see every kilometer of the race (in 3D if you have terrain enabled) and virtually fly over the stages. For live online coverage of the Tour, go to ATDHE.net between roughly 8:30am and 11:30am EST and find the first Tour de France link for English coverage. The official race website is also useful for news, videos, and all the information you could want or need about the standings, riders, and course.

Enjoy!

Adobe Acrobat Webinar: Geospatial and Mapping

Attending this webinar today on facilitating "data maintenance workflow for individuals in the field by using Acrobat 9 with professional desktop GIS software" was well worth the price of admission for me (it was free!). As a newcomer to the field, it was an opportunity to see an example real world application of a GIS and how markup from the field is processed by an in-office GIS technician. The technology of GeoPDFs -- a spatially intelligent PDF map complete with attributes, labels and layers -- is also impressive for a few reasons: unlike hard copy maps which become obsolete and need to be physically replaced, GeoPDFs can be quickly updated with whatever new data is available in the GIS database; a GeoPDF map of an area can be quickly created upon request unlike a hard copy map; updated mapping data markup on a GeoPDF by individuals in the field can be automatically located and easily interpreted by a GIS technician in the office.

I was disappointed upon not winning either of the draw prizes at the conclusion of the webinar though -- a copy of Acrobat 9 Pro Extended and some guide to Acrobat 9 Pro. I guess the knowledge imparted was reward enough.

It looks like the webinar will be available for download on demand from the Acrobat Users website soon. Keep an eye out for it.

Penn State Online Lidar Technology and Applications Course

Penn State is offering an online course in Lidar Technology and Applications this fall, 10-12 hours a week for 10 weeks. There are no official prerequisites but they do recommend that one have an understanding of GIS, remote sensing and cartographic mapping. On the website they offer an online 20 question quiz for self assessment of one's preparedness, a mark of 16 or higher indicating that you should be well prepared for the course.

I gave it a whirl and scored 16 (80%) -- good news for me!