Spatial Analytics

Exploration: This is done by working with an existing spatial data file or through geocoding, where values such as postcodes and latitude/longitude points are translated into specific geographic coordinates and projections.

Modeling: Employing descriptive analytics to calculate both values and shapes, such as an isochrone mapping, which is used to visualize distances and travel times between points. It’s also possible to encode geographical entities such as lines and polygons to build 2D and 3D models of real-world objects.

Comparison: Processing various spatial shapes together makes it easier to calculate areas of overlap or boundary and therefore generate new spatial information in the form of calculated points, lines, or polygons.

Prediction: By reviewing how spatial analytics changes over time, analysts can detect patterns and present interactive maps with forecasted data.

Spatial Data Analysis: Data is collected, processed, and augmented to generate value according to location-based attributes, properties, or relationships. This allows access to details, such as location, position, and distance, that would otherwise be difficult to obtain.

Spatial Autocorrelation: Testing determines whether data points that are closely co-located are also similar when it comes to other attributes. For example, spatial autocorrelation can investigate whether a disease is isolated or present in clusters around an area.

Spatial Stratified Heterogeneity: The uneven distribution of features in a spatial region is measured to determine how patchy/heterogeneous a series of layers/strata is within the defined boundaries. Commonly used to determine coverage within a series of geospatial polygon zones as part of a larger descriptive analysis.

Spatial Interpolation: Location-based data points with known attributes are used to estimate the values at other unknown points. This type of interpolation is commonly used to estimate temperatures between weather station locations to create an interpolated statistical “surface” across the region of interest.

Spatial Regression: Models are built that consider spatial characteristics alongside traditional numeric features to infer numeric results such as salaries and birth rates.

Spatial Interaction: Insights are drawn from the interaction of different entities including points, lines, and polygons. For example, boundaries may touch, areas may overlap, or a spatial object can be completely contained by another.

Simulation and Modeling: An analysis and understanding of geospatial objects and their properties offers a measurement of their changes over periods of time following the experimental conditions.

Multiple-Point Geostatistics (MPS): A collection of algorithms that simulates spatial structures and patterns based on a statistical model. MPS often focuses on describing geospatial structures through probability distributions and is used for subsurface reservoir models.