Massachusetts Institute of Technology
Department of Urban Studies and Planning

11.520: A Workshop on Geographic Information Systems
11.188: Urban Planning and Social Science Laboratory

Lecture One: Overview of Course, GIS Principles, Elements of Maps, ArcGIS Basics

September 8, 2010

(Joe Ferreira with contributions from Visting Prof. Zhong-Rhen Peng who taught the Fall 2003 class)

Course Overview & Logistics

· Discuss: Syllabus/Lectures/Labs/Homeworks/Project/books
more info on class website: http://mit.edu/11.520

· Times: Wed. lecture (14E-310, 2-3:30), Mon. lab (37-312, 2-3:30, then supervised lab until 5:00)

· Extra lab: This Friday 11-2 and Sunday 4-7 in computer lab (37-312: for those *not* in MCP orientation "GIS Intro' last week
[See Introduction to GeoProcessing Presentation (Lecture notes by Lisa Sweeney, MIT GIS Services, during MCP Orientation, Sept. 1, 2010)

· Stellar website only for lab exercises, homework, and test

for 11.188: http://stellar.mit.edu/S/course/11/fa10/11.188/index.html
for 11.520: http://stellar.mit.edu/S/course/11/fa10/11.520/index.html
[NOTE: MCP1 students use GIS Intro on Stellar for the first exercise]

· Other GIS courses: 11.205, 11.534, 11.521, Fall & IAP miniclasses (field data collection, digital elevation models,Harvard, BU,...

· Student background

· Next Monday: Go directly to the Room 37-312 lab at 2 PM

Examples to view and discuss:
... which are GIS? what to learn? how to add your own data/analyses?

Spatial analysis using commercial GIS software

In Lisa Sweeney's Introduction to GIS and GeoProcessing: :Introduction to GeoProcessing Presentation

In Lab #1: ArcMap (part of ESRI's ArcGIS) to do thematic mapping of Cambridge block groups and overlay on Google Earth.

Lab #1 does little spatial analysis (only simple thematic mapping) and mostly visualization of geographic patterns

Spatial analysis example: which Cambridge housing sales are in low-income neighborhoods, in flood plains, near major roads...

Private sector mapping services - try these on your own

Mapquest or Google-Maps to find a location and generate a street map: www.mapquest.com, maps.google.com

Google-Earth or Virtual Earth to navigate and 'fly' over the earth: earth.google.com, local.live.com

Note overlays of vector (roads), raster (imagery), 3D (buildings) and mashups of user-generated data

Explore user interface (managing 3D view) and virtual earth's "bird's eye" view (using pictometry data)

Explore use of SketchUp to add your own 3D models

SmartPhone navigation and GPS-based tracking (e.g., various iPhone mapping applications)

Web services using open-source (LAMP) tools

Compare ways of accessing data on MassGIS website: http://www.mass.gov/mgis

via ordinary browser, Oliver (java) applet, and OGC Web Mapping Services protocols

Examine MIT Library GeoWeb service (using GeoServer): http://web.mit.edu/geoweb/

Evolution of GIS: A Timeline from 1970s to now

GISytems, GIScience, GIStudies, and GIServices
- GISystems – A computerized tool that helps solve geographic problems.
- GIScience – The development of data models, algorithms, and methods for representing geography and spatial relationships in order to support spatial analysis and location-based computing
- GIA - Geographic Information and Analysis
- GIStudies – the systematic study of society’s use of geographic information, including institutional, organizational and procedural issues.
- GIServices – The business of providing GIS data and analysis tools to GIS users (often by chaining interoperable components in lego-block fashion).

Example GIS Applications - think about useful data models for each example

Resource inventory (what is available where?)
Thematic maps (spatial pattern of population density, income, household size, etc.)
Network Analysis (How to get somewhere by road in the shortest amount of time?)
Location Analysis (Where is the 'best' place to locate a shopping mall?)
Terrain Analysis (What is the danger zone for a natural disaster? What can be seen from here [visibility analysis])
Spatio-Temporal Analysis (Land use: what has changed over the last twenty years, and why?)
Runoff models (If it rains 1 inch in 3 hours, how big is the storm surge at the bottom of the hill?)

Data Models -- How to encode Geometry and Geography

Model spatial objects by describing their boundaries (vector data models)
- Use coordinates that locate the spatial objects on the Earth
- Point--a single XY location is often sufficient
  - MBTA Stops
  - Is Boston a point?--At different scales or for different purposes, Boston could be a point or polygon.
- Line--only one dimension needs to be represented
  - Street centerline, MBTA Railroad track, ridgeline, bux route
  - How does it matter if street is modeled as centerline or as the void between blocks?
- Polygon--2D planar surfaces
  - Cambridge border, central square boundary, census tract, parcel, ...
  - What about river boundary, edge of ocean (at high tide?)
- Beyond planar surfaces - terrain models, 3D CAD models, ...
Model the space that contains spatial objects (raster data models)
- Divide space into 'cells' and encode whether a spatial object is in each cell
- 30m x 30m grid cells for Landsat image - classified based on predominate land use within each cell
- 6 inch pixels for color orthophotos developed from aerial photography
- 3 km x 3km x 1 km (height) volumes for meterological modeling

What does it mean to be "doing GIS"?

using the tools of Geographic Information Systems to solve a problem

such as those in the previous examples
a GIS project might have the following stages:

define the problem
acquire the software (and the hardware? or network resources?)
acquire the data
clean the database
perform the analysis
interpret and present the results
institutionalize the process so that the analysis can be routinely redone

GIS tools and techniques are becoming more integrated with other tools and/or embedded in web services
- Do spatial analysis without being 'in' a GIS package
- Include location, spatial indexing, etc. in general purpose database management tools

Elements of the Map

Scale (Distance on the map compared with distance on the earth)
Symbolization
Projection (Euclidan coordinate system instead of Lat/Lon)
... Ignore for today and jump into ArcMap to get a feel for key features of a typical GIS package
- We'll pay more attention to maps, data models, and ArcGIS basics in other lectures or in lab

4. ArcGIS Basics-- preview of Lab Exercise 1

Mapping Cambridge home sales and household income (with ArcMap onscreen with data from exercise #1)

Things to think about

Setting Up a Work Environment

Starting ArcGIS

Setting up workspace, scratchspace, database, etc.

Getting Data Into ArcMap

Data Frame Properties: Name, Units (Map, Display)

Layer Property

Tool in/out

Attribute Data

Basic Map Making

Simple Symbolization

Thematic Symbolization

Saving Your Work and Printing Output

For now

Explore cross-referencing of map and data tables

Discuss data model for geometry and attributes

Compare uses of ArcMap with those of Mapquest and Google Earth

Overview of ArcGIS software (with Lab #1 exercise on-screen)

What are the various parts of ArcGIS and what do they do
How is ArcMap organized
Can you utilize ArcGIS Help effectively
How can a problem be split into operations that utilize the underlying 'vector' and 'raster' data models
- E.G., which Cambridge housing sales occurred in low-income neighborhoods

'Vector' data models for geospatial location
- Geometry model:
  - boundary representation 'vector' model of spatial features
  - assign spatial feature ID to each spatial object within each map 'layer'
  - points (sales), lines (streets), and polygons (block groups)
    - Separate X,Y data for every part of every feature
    - Shared X,Y data for shared points and boundaries
- Attribute data model
  - relational tables linked to spatial features via feature-ID
  - graphical interface to utilize geometry/attribute links
    - highlight map features by selecting attribute table rows
    - highlight attribute table rows by selecting map features
- More than one way to represent the geometry: shape detail, shared boundaries, ...
  - ArcGIS shapefiles are the simplest and most common vector model
    - Only one feature type per shapefile
    - Several (4+) files on disk for each shapefile: cambbgrp.xxx
  - ArcGIS coverages use a more complicated data model
    - Data model includes topology, shared boundaries, ... (illustrate difference)
    - Geometry stored in sub-directory: cambtigr and sales89
    - Attributes stored in local database: in shared info sub-directory
- Complications
  - islands, lakes, overpasses
  - shared edges?, do attached links move when you move points?
  - ambiguity: summer/winter wetland boundaries
  - scale, generalization, conflation, slivers
  - Coordinate systems and projections
  - One-to-many relationships among spatial features and events
- Thematic mapping - tip of iceberg regarding GIS applications
  - Symbology
    - many options
    - review 'symbology' page of layer properties
    - review ArcGIS help files for symbology
  - Different classification schemes (show help page):
    - Equal Interval
    - Natural Breaks
    - Quantile
    - Standard Deviation
  - Normalization: people or population density - Why do we care? (show examples)

For MONDAY - go directly to the lab (Room 37-312) at 2 PM

From Syllabus - do introductory readings ("What is GIS" intro from NCGIA curriculum, or Ormsby, chapter 1-2 or O'Sullivan and Unvin chapter 1, and Zeiler chapter 1)
Need combination lock plus MIT card to access the lab and MIT (athena/email) ID to login.
ArcGIS software is also available for installation on personal computers
- Fill out request form at: http://ist.mit.edu/services/software/vsls/esri
- but, will only work when networked to MIT with a valid MIT ID
- requires other software (Kerberos and VPN to connect from off campus, AFS to access data)
- Deatailed instructions available on websites at CRON (http://cron.mit.edu/p.lasso?t=8:5:5) and MIT Library (http://libraries.mit.edu/gis/)

Created by Prof. Zhong-Reng Peng, September 2nd, 2003.
Modified by Jinhua Zhao, September 7th, 2004.

Last Modified by Joseph Ferreira, September 8, 2009.

Back to the 11.520 Home Page.
Back to the CRN Home Page.

For information about this page please send e-mail to 11.520staff@MIT.EDU

Massachusetts Institute of Technology Department of Urban Studies and Planning