Overview Flashcards
What is Geography?
the study of spatial (geographical):
- Components (heterogeneity)
- Relationships (structures/interactions/dependence/spatial autocorrelation)
- Process
- Long-term welfare
What is the purpose of applying geographic knowledge?
- Generating new knowledge
2. Solving problems
Russell Ackoff’s DIKW hierarchy
As the connectedness and understanding of data increases the hierarchy moves from dat–>information–>knowledge–>wisdom
Use of analytical GIS tools to:
- Describe
- Explain
- Predict
- Support decision-making
Description (first use of GIS tools)
- Qualitative descriptors: topological relationships (within, contain, overlap)
- 1-point (aspatial): mean, scatterplots, histograms
- 2-point (spatial): centroid (center of mass), point of minimum aggregate travel (MAT), dispersion, Moran’s I, semivariogram
- Multi-point (spatial): shape, size, patch fragmentation
- Location (geographic descriptor)
Explanation (second use of GIS tools)
- Queries and visualization
- Data transformation
a. buffering
b. point in polygon
c. overlay - Exploratory data mining
- Spatial inference/modeling
Prediction (third use of GIS tools)
- Spatial interpolation
a. inverse distance weighting
b. kriging
c. density estimation - Spatial modeling
a. spatial regression models
b. spatial process models
c. agent-based models (week 7)
Spatial Decision Support (fourth use of GIS tools)
- Map communication
- Spatial data integration
- Location-allocation
- Optimization
- Routing (shortest path, TSP)
Limitations/Considerations
- Spatial Heterogeneity
- Spatial autocorrelation
- Ecological fallacy
- MAUP - Scale and zonal effects
- Uncertainty and error
a. conception
b. measurement and representation
c. analysis
Spatial Heterogeneity
Uneven distribution of various concentrations of each species (or characteristics) within an area.
“patchy distribution”–this makes generalization highly fallable
Spatial autocorrelation
co-variation of properties within geographic space
statistical analysis is problematic because many common analyses (regression) assume independence among observations
Ecological fallacy
Inferences about the nature of individuals are deduced from inference for the group to which those individuals belong
“inferring individual characteristics from aggregate population data”
OR correlations observed at population level cannot be applied to individuals
Example: wealth is positively correlated with tendency to vote conservative, but wealthier states are predominantly democrat
MAUP - Scale and zonal effects
error that occurs when points are aggregated into districts, resulting values (totals, rates, proportions) are influenced by the choice of district boundaries (like census enumeration districts)
Uncertainty and error
degree to which the measured value is estimated to vary from the true value
Information Systems/Science
- hardware/software
- software
- (G)UI, Tools, DBMS, Data
- GIS data models and database management systems (DBMS)
- CAD, graphical, image
- raster
- vector
