GIS Spatial Statistics 2 — Questions and Answers
Question 1: Kernel Density Estimation (KDE) in GIS is primarily used to:
- Convert vector polygon data to raster format
- Create a smooth continuous surface estimating the density of features (Correct answer)
- Calculate Euclidean distances between point pairs
- Join attribute tables based on spatial location
Correct answer: Create a smooth continuous surface estimating the density of features
KDE places a kernel function over each point and sums the contributions to create a smooth density surface showing where features concentrate.
Question 2: Local Indicators of Spatial Association (LISA) are designed to:
- Produce a single global autocorrelation index for the entire dataset
- Identify local spatial clusters and spatial outliers at each feature location (Correct answer)
- Calculate geographic distances using network paths
- Generate random point patterns for simulation
Correct answer: Identify local spatial clusters and spatial outliers at each feature location
LISA statistics decompose global measures like Moran's I into location-specific values, identifying local clusters and outliers rather than a single summary statistic.
Question 3: In a Complete Spatial Randomness (CSR) test for point patterns, the null hypothesis states that:
- Point features are significantly clustered in space
- Point features follow a regular lattice-like distribution
- Point locations are distributed randomly with no spatial dependence (Correct answer)
- Point features exhibit strong positive spatial autocorrelation
Correct answer: Point locations are distributed randomly with no spatial dependence
The CSR null hypothesis assumes points are generated by a homogeneous Poisson process, meaning each location is equally likely to contain a point independently of other points.
Question 4: The Anselin Local Moran's I classifies locations into which output categories?
- Hot spots, cold spots, and statistically insignificant areas only
- High-High, Low-Low, High-Low, and Low-High spatial association types (Correct answer)
- Positive, negative, and neutral autocorrelation zones
- Dense, sparse, and uniform distribution classes
Correct answer: High-High, Low-Low, High-Low, and Low-High spatial association types
Local Moran's I produces four cluster/outlier types: High-High and Low-Low (spatial clusters) and High-Low and Low-High (spatial outliers).
Question 5: Ripley's K function is used in spatial statistics to:
- Analyze point pattern clustering or dispersion across multiple distance scales (Correct answer)
- Measure polygon shape compactness and regularity
- Assess spatial autocorrelation in raster cell values
- Calculate shortest paths in transportation networks
Correct answer: Analyze point pattern clustering or dispersion across multiple distance scales
Ripley's K function counts the number of neighboring points within increasing distances, allowing analysis of clustering or dispersion at multiple spatial scales.
Question 6: Kriging, as a spatial interpolation method, is used to:
- Classify raster pixels into land use categories
- Estimate attribute values at unsampled locations using geostatistical methods (Correct answer)
- Identify the shortest path between network nodes
- Delineate watershed boundaries from elevation data
Correct answer: Estimate attribute values at unsampled locations using geostatistical methods
Kriging uses the spatial autocorrelation structure modeled in a variogram to create optimal linear unbiased estimates at unsampled locations.
Question 7: In geostatistics, a variogram is used to:
- Visualize distortion in different map projections
- Quantify how spatial dissimilarity between point pairs changes with increasing distance (Correct answer)
- Calculate descriptive statistics for polygon areas
- Define the geographic coordinate reference system
Correct answer: Quantify how spatial dissimilarity between point pairs changes with increasing distance
A variogram plots semivariance (half the average squared difference between pairs) against separation distance, describing the spatial autocorrelation structure of the data.
GIS Spatial Statistics 2
Kernel Density Estimation (KDE) in GIS is primarily used to: