GISP-E cheat sheet

Vector model stores features as points, lines, or polygons with explicit x,y coordinate pairs; raster model stores a grid of cells, each holding one attribute value, so a single raster layer represents exactly one theme.

Topology defines spatial relationships (adjacency, connectivity, containment) independent of coordinates; a planar enforcement build detects dangling nodes, undershoots, and overlaps that geometry alone cannot flag.

Vector overlay rules: union keeps all areas from both layers, intersect keeps only overlapping areas, identity keeps all of the first layer plus attributes of the second, and clip cuts geometry without merging attributes.

Raster resolution is the ground size of one cell; halving cell size quadruples cell count and storage, so resampling to coarser cells loses detail irreversibly while finer resampling invents no new information.

A spatial reference has two parts: a geographic datum (ellipsoid plus tie to Earth) and a projected coordinate system (map projection plus linear units); mismatched datums cause shifts of tens to hundreds of meters.

Map projections always distort shape, area, distance, or direction; conformal preserves local angles and shape, equal-area preserves area, equidistant preserves distance along specific lines only, and no projection preserves all four.

Universal Transverse Mercator divides Earth into 60 zones of 6 degrees longitude; scale factor is 0.9996 on the central meridian, so UTM distances are slightly too small away from it and zones are unsuitable beyond about 3 degrees from the meridian.

A state plane or other local projected system uses feet or meters with a defined false easting and northing; a common error is mixing US survey feet and international feet, producing a scale error near 2 parts per million.

Geodatabase feature classes can hold a single geometry type with a defined spatial reference; feature datasets group classes sharing one spatial reference and enable topologies, geometric networks, and relationship classes.

Buffer distance is measured in the units of the layer's coordinate system, so buffering a geographic (degree) layer with a distance meant as meters yields meaningless results; project first or use geodesic buffering.

Spatial join transfers attributes based on a spatial relationship (intersect, within, closest); the target layer keeps its geometry, and one-to-many matches require a defined match option such as one-to-one, one-to-many, or match closest.

Slope is the maximum rate of change between a cell and its neighbors in percent or degrees; aspect is the downslope direction in compass degrees, with flat cells conventionally assigned -1.

Hydrologic sequence: fill sinks, compute flow direction (D8 assigns each cell to one of eight neighbors), accumulate flow, then delineate watersheds by selecting cells above a flow-accumulation threshold.

Inverse distance weighting interpolates as a weighted average where weight equals distance to a negative power; it is exact at sample points, cannot extrapolate beyond the sample range, and produces bull's-eye artifacts near clustered points.

Kriging is a geostatistical interpolator using a fitted variogram (nugget, sill, range) to model spatial autocorrelation; it can produce a prediction standard error surface, unlike deterministic methods such as IDW or spline.

Moran's I tests global spatial autocorrelation: values near +1 indicate clustering of similar values, near -1 indicate dispersion, and near the expected value indicate randomness; a z-score and p-value determine significance.

Hot spot analysis (Getis-Ord Gi*) returns z-scores per feature; a high positive z means a feature is part of a cluster of high values, but significance requires the associated p-value, not the z-score alone.

Raster map algebra uses local (per-cell), focal (neighborhood), zonal (per zone), and global (entire raster) functions; the analysis window and the processing extent and cell size environment settings change results.

Positional accuracy is the closeness of coordinates to true ground position, while attribute accuracy is correctness of nonspatial values; a dataset can be strong in one and weak in the other, so each is documented separately.

Topological consistency errors include dangling nodes, undershoots and overshoots, unclosed polygons, duplicate lines, and sliver polygons; each is detected with specific rules rather than visual inspection alone.

Metadata standards (for example the FGDC Content Standard for Digital Geospatial Metadata and ISO 19115) record lineage, positional accuracy, attribute accuracy, logical consistency, completeness, and entity definitions.

Lineage documents sources, processing steps, and parameters; without it a derived layer cannot be reproduced or defended, and it is the first item requested in data-quality disputes.

Scale is the ratio of map distance to ground distance; a 1:24,000 map shows more detail than 1:100,000, and a layer digitized at 1:100,000 should not be presented as accurate at 1:12,000.

Minimum mapping unit sets the smallest feature retained at a given scale; features smaller than the MMU are dropped or merged, so feature counts differ between scales of the same source data.

Visual variables are size, shape, value (lightness), color hue, color saturation, orientation, and texture; hue suits nominal data, value and size suit ordinal or numeric data, and shape suits nominal categories only.

Classification choices change the message: equal interval emphasizes range, quantile gives equal counts per class, natural breaks (Jenks) minimizes within-class variance, and standard deviation highlights deviation from the mean.

Map layout essentials: scale bar, north arrow or graticule, legend, title, data source and date, projection note, and author; a map without projection and source information cannot be interpreted reliably.

Relational keys: a primary key uniquely identifies a row, a foreign key references another table's primary key, and a one-to-many relationship is implemented by placing the foreign key on the many side.

Normalization reduces redundancy: first normal form removes repeating groups, second removes partial dependencies on part of a composite key, and third removes transitive dependencies between nonkey attributes.

A geodatabase relationship class stores associations between object or feature classes with defined cardinality (one-to-one, one-to-many, many-to-many) and can be simple or composite, where composite implies lifetime dependency.

Versioning allows multiple editors to work on the same geodatabase; reconciling compares versions, and conflicts arise when the same feature or topology element is edited in both, requiring a defined conflict resolution rule.

File geodatabase storage limits: a single file geodatabase can hold up to 1 TB by default, and each feature class or table is limited to about 1 TB; enterprise geodatabases scale with the underlying database management system.

Data conversion trade-offs: shapefile limitations include 2 GB size, 10-character field names, no true curves, and one geometry type per file; GeoJSON uses WGS 84 longitude/latitude by default, and GeoPackage is an open OGC container.

Interoperability standards include OGC WMS (rendered map images), WMS is not queryable feature data, WFS (vector features), WCS (coverage or raster), and the OGC API family; choosing the service depends on whether the client needs pixels or features.

Licensing and ethics: data may carry redistribution restrictions, and professional codes require documenting uncertainty, protecting confidential locations, and disclosing limitations of derived products.

Backup and archiving: retain source data, processing scripts or models, and metadata together so results are reproducible; a model or script without its input data and parameters cannot be rerun.