Midterm Review Flashcards
Scale
Ratio of map unts to GIS units
Vector Data Model
- Use discrete elements such as points, lines, and polygons to represent the geometry of real world entities3. Features are stored as a series of x-y coordinate system in geometry types: points, lines, or polygons
- Each feature has attributes or info
- Feature class is a collection of similar objects with the same attributes
Rastor Data Model
- Uses grid cells arranged in rows and columns with a value assigned to each grid• Uniform, regular cells
• Each cell, attributes are constant
Precision of raster model
• Half the cell size is how precise you can be
• Finer grid size=more precise also larger file (more data)
Spaghetti Vector Model
- Each line is stored separately in the database
- Lines and polygons can overlap
- Shared polygon boundaries are stored twice
- Since lines stored twice, storage is inefficient
Topological Vector Model
- Shares lines (state of Illinois will share with Missouri, used for parcel maps, very powerful)
- Complex data set up
- Geometric properties that do not change when bent or stretched
- Bending polygon doesn’t change neighbor because they bend also
- Planar topology (2D, no overlapping lines)
- Exhaustive (no gaps in polygon)
- Three tables
- Polygon (contains all polygons)
- Node (all nodes)
- Link (begin and end of nodes)
- Cons (cost more, clean data entry)
- Pros (Analysis, efficiency)
Types of Vector Files
ESRI Coverage
Shapefile
Feature Class/Geo Data Base
ESRI Coverage
• Kind of a pain, cease up when there are spaces in saving
• Three basic relations
o Connectivity (arcs connect at nodes)
o Area definition (are is series of connected arcs)
o Contiguity (arcs have left and right direction)
• Really old school
• Lines stored once
• Pro (land boundaries, move one line and other follows is critical)
• More complex than shapefiles due to topology
• Important to move files carefully so don’t miss key component (e00)
• They concentrate on storing topology. So you will see that the emphasis is in storing the geometry elements first, that is the nodes, edges that make up all the geometries. You will then see a separate set of tables that relate those geometries to the attributes
• Nodes at every node intersection
• Not have two or more nodes on top of eachother
• Have sense of direction
• Work well for topological relations (parcel boundaries)
• Doesn’t work well for 3D like a bridge
Shapefile
Spaghetti vector (lines stored twice)
• .shp file
• Pro (smaller file, quick retrieval, thematic)
• Is an open specification that was simple to implement
• Has a large library
• Each geometry worried about itself and not other geometries that it intersected
• No complicated math to insure correct
• Can have multiple geometries cross, and two points on top of eachother
Feature Class/Geo Data Base
Newest one
• Supports geographic concepts
• Contains multiple points, polygon, and polyline layers
Large Scale vs. Small Scale
- Only dealt with it in the annoying preset layers, preset zoom on the scale
- GIS terminology, large and small are reversed
- Map units/real units….1:24000 is a small fraction hence small scale, a 1:2 scale will be considered a large scale
Four types of data Quality
Geometric Accuracy
Thematic Accuracy
Temporal
Precision
Geometric Accuracy
Does it reflect actual geometry of the map (is it shifted?)
Thematic Accuracy
Did we pick the right them for the map
Temporal
When was data created? (shorline is good example)
Precision
How precise measuring depends how precise the layer was made, example length of rods, depends •
Metadata
- State has really strict codes
- Has notes with the map to help show how accurate, who made it, how it was made, etc.
- ArcCatalog has these notes
Geo Databases
Way arc prefers to store stuff, shape file has the .shp on the end of file, also looks like cylinder
Map Document vs. Layers
- Separation of info allows reality to be seen as a series of overlapping layers of info
- Each layer can be treated separately, but also combined
- Ex. Soils, floodplains, zoning, topography
Moving Files in ArcMap
• Understand the shape files aren’t just one file and there are 4-5 pieces attached with it
Types of Coordinate systems
Unprojected
Projected
Geoid/Ellipsoids
Unprojected
• Based on spherical globe coordinates, degrees of lat and long
Projected
- Converts spherical coordinates to planar using math equations
- Projects 3D to 2D map
- Every projection based on a GCS, ever GCS has datum, so every projection has a datum
Geoid/Ellipsoids
• Geoid is the lumpy shape that the earth actually is
• Projections are all about choosing an ellipsoid that represents the shape of the earth
• You can make an ellipsoid a sphere, but usually have one longer that represents earth a little better
• Defined with major and minor axis
• Geoid-earth is lumpy not perfect ellipse (theoretical surface defined by gravity measurements)
1. Complex so usually use ellipse (another source of error)
Datum
- Major and Minor access of the ellipsoid and where it is centered at on earth
- Minimize error from geoid and ellipse use a datum (shifts ellipse relative to geoid)
- Local and geocentric/world datum
- Word we say for ellipsoid
- It is both the technical specs of the ellipsoid and survey benchmarks
- Named after the names they are done for (North American)
- Must know and record datum to use data correctly later
Map Projection Process
- Land elevation defined to geoid (Mean sea level)
- Geoid to more regular ellipse
- Math equation convert spherical into planer (where distortion occurs)
- Exact mathematical formulas (one datum to another takes specialized fitting, not exact, have errors, only convert when necessary)
- Taking a 3D map and spreading it flat in 2D
- Converts degrees to meters or feet
- Taking a light and shining it through the ellipsoid and projecting on the wall is a good example
Scale Factor for Projection
- Scale only true on points and lines of tangency
- Factor=actual scale/principal scale(scale of generated globe)
- The scale you see on the map doesn’t hold for the entire map, because all maps are distorted
- On any map as you move around the map the actually scale varies but in general it is oriented around the actual scale
Distortion Properties of Projections
- Distance preserves distance or you can preserve, in powerpoints
- Or equal area which is often a conic, area preserved but shapes are off
- Compromise (distorts all four properties, Robinson)
- Large scale maps (cities, small states=virtually no distortion)
- Small Scale maps ( distortion inevitable, purpose drives choice
Map Projection Surfaces
Planer/Azimuthal (true direction Projection)
Conic
Cylindrical
Planer/Azimuthal
- Used on maps that show great circle routes from point to destination
- Can trace projection on the wall (projected onto flat plane)
- Looks like dart board, equal parallel and straight meridians
- On airport walls
Conic
- Like a cone put over the globe
* Distorts direction and shape but preserves distance and area
Cylindrical (Mercator)
- Like a globe traced around a cylinder
* Distorts distance and area but preserves direction and shape (Mercator)
Light Sources
Orthographic
Gnomonic
Stereographic
Orthographic
Light source infinity, how earth is viewed from far away planet
Gnomonic
Light source at center of globe
Stereographic
Light source on surface of globe
Tangent vs. Secant
Tangent • Cylinder/cone is tangent to the globe • Has single standard parallel • No distortion along parallel, increases with distance from it Secant • Cylinder/cone is secant to the globe • Has two standard parallels • No distortion along parallels, increases with distance from them
Aspect
- Location of point or line of tangency on the projection surface
- At equator=equatorial aspect
- Norma aspect (used most recent), planar (normal aspect polar), conic (normal aspect is apex of cone), cylindrical (normal aspect is equator)
Commonly used Projections
Geographic coordinate system (GCS)/Unprojected
Universal Transverse Mercator (UTM)
State Plane
GCS
- If data called in doesn’t have a projection, Measured in angular degrees of lat and long
- Primary use for data distribution by GIS
- User chooses projection for desired coordinate system
- Not used for maps or analysis due to distortion
- Kind of is projected but weird (distorted), treats everything like a rectangular grid
- Datum and ellipse adjustment is used for GCS
Universal Transverse Mercator (UTM)
- Divides world into 60 zones (minimal distortion for each zone)
- Wrapping around the globe, the UTM system is a secant projection which means it has two line, rotated around 60 times for the US
- Usually have north being the aspect
State Plane
- Divides the 60 into even more zones (large states=more zones, unique FIPS number)
- They are like states have all chosen what the best projection is for their mapping
- Larger states divide into multiple state planes (California has 5)
- They pick a projection that is good for that region, then set the lines of tangency to make since for that little square
- All based on projection
- Usually a UTM or conic projection
How projections are saved in ArcMAp
• Just saved in the .prj file
• Use define project tool to fix stuff
o Tells arc how to interpret the file that comes in
Choosing the right projection
- Cylindrical (preserves direction and shape, good for N-S oriented areas)
- Conic (preserves area and distance, good for E-W oriented areas)
- Azimuthal (preserves area and distance, good for satellite data and polar regions)
Trouble shooting Projections
• Oregon covers two UTM zones and two state plane zones
o Answer: state created its own Coordinate system
• Lousianna centered in UTM and two state plane
o Use UTM zone parameters but adjust to central meridian of that zone
• Kansas (oriented E-W so conic is better)
o Modify standard parallels so they approximate state better
• Turkey (Find central meridian, determine lat of parallels, choose reference latitude, and create new coordinate system)
• Most common problem is a mismatch between label and actual CS
o Dertemine if it exists
o Determine actual CS
o Define labe correctly
o Test to see if alignment problem is fixed
Types of Data
Nominal
Categorical
Ordinal
Interval/Ration
Nominal
• Single symbol maps (state names, parcel ID number, each figure has own value, labels usually portrayed on map)
Categorical
• Unique values maps (defined numbers or distinct categories, interstate, freeway, cover type, symbol is its own category)
Ordinal
• Unique values maps (ranks the categories, low, medium, high, best city to live in)
Interval/Ration
• Quantities maps (colors, symbol, density)
Thematic vs. Image Rasters
Thematic
o Land use or rainfall, may be continuous or discrete
Image
o Satellite or air photos, represents brightness, usually continuous
Classify Data
Natural Break
Equal Interval
Quantile
Standard Deviation
Natural Break
o (exploits natural gaps in data, good for unevenly distributed, default method, works well for most data)
Equal Interval
o (specify number of classes, equal spaced classes, best for uniformly distributed data)
Quantile
o Same number of features in each class, could get very unevenly spaced
o Depend on data distribution
Standard Deviation
o shows deviation from mean
o user chooses units, assumes data are normally distributed
Map Documents
- contains one or more data frames
- stores properties for each layer
- stores references to files
- Broken data links when document can’t find data due to pathname
- Absolute path (placed on central server to be used by all)
- Relative path (keep data with the map)
Layers
Held in memory and saved with map
Queries
• Extract certain records from a map or table, records meet certain criteria
Joins
- Join two tables based on a common spatial relationship
- Starting using attribute tables a lot, may not have a power point, just lab stuff
- A lot of summarizing
- Have Special Query Language (SQL)
Cardinality
number of elements in a set or group
Summarizing Fields
• Group cities according to the airport they served and summed the population
Relates Joins
o Fields same type and unique
o It doesn’t tell you the join is messed up until the results are messed up
Boolean Expression And vs. Or
And=both must be true
Or=than either may be true
Trying to find city that matches all of the criteria
Spatial Queries
Intersection (does road cross)
Contain (is aspen inside)
Proximity (within 200 m)
Simple Inside
Which country is each school in
Simple Distance
Which attraction is closest to each hotel
Summarized Inside
how many schools in each county
Summarized Distance
How many attractions are closest to each hotel
Vector pro
Good topology, good for graphics, good network analysis, homogeneous data
Vector con
Complex, slow data collection, sharp boundaries
Raster pro
Simple, easy overlay, large area
Raster con
Large data, low geo. accuracy, no topology
