Geo-visualization and Cartography
Open in the Padhaiko app7.1 Map elements, map design principles and visual variables
Cartography is the art, science, and technology of communicating geographic information effectively through maps, while geo-visualization extends cartography to interactive, exploratory, multi-dimensional visual interfaces.
Essential Map Elements
A complete, self-contained layout map balances eight standard elements:
- Map Body (Data Frame): The primary geographic display area.
- Title (and Subtitle): Conveys theme, geographic region, and reference year clearly.
- Legend: Decodes every symbol, color ramp, line weight, and pattern shown in the map body.
- Scale Bar / Representative Fraction: Shows ground-to-map distance ratio (bar scales stay accurate even when enlarged/reduced in print).
- North Arrow / Graticule: Indicates geographic orientation (grid lines of latitude/longitude or UTM ticks).
- Inset / Locator Map: Places the study area in wider regional context or enlarges dense urban clusters.
- Neatline & Frame: Outer border framing the composition and clipping map extent cleanly.
- Credits / Metadata Note: Documents data sources, coordinate reference system (CRS), projection, author, and date.
Map Design Principles
- Visual Hierarchy: Emphasize primary thematic data with bold contrast and saturated color while pushing base layers (administrative borders, minor roads) into subdued grays.
- Figure–Ground Contrast: Make the focal subject ("figure", e.g., Nepal) visually pop forward from surrounding context ("ground", e.g., neighboring countries or ocean) via drop shadows, crisp borders, or lightness contrast.
- Balance & Negative Space: Arrange map elements evenly around the optical center without crowding borders or leaving awkward empty corners.
- Cartographic Generalization: When reducing map scale, simplify geometry to keep maps legible via selection, simplification (Douglas–Peucker), smoothing, aggregation (merging buildings into urban blocks), amalgamation, and displacement (separating overlapping road/rail lines).
Bertin's Visual Variables
Jacques Bertin identified seven visual variables (plus modern transparency) that encode qualitative or quantitative attribute differences onto map symbols:
| Visual Variable | Description | Best Measurement Scale | Examples |
|---|---|---|---|
| Position | Location on the map plane | All (spatial base) | Point coordinates, polygon boundaries |
| Size | Area, length, or width of symbol | Ratio / Interval / Ordinal | Graduated circles for city population; line thickness for traffic volume |
| Color Value (Lightness) | Light-to-dark shade of one hue | Ordinal / Ratio (Rates) | Light yellow to dark brown for population density |
| Color Hue | Dominant spectral wavelength (red, blue, green) | Nominal | Green = Forest, Blue = Water, Yellow = Cropland |
| Color Saturation | Intensity/purity vs. grayness | Ordinal / Uncertainty | Faded hues for low-confidence estimates |
| Shape | Geometric or pictorial icon form | Nominal | Cross = Hospital, Airplane = Airport, Triangle = Peak |
| Orientation | Angular tilt of marks or hatching | Nominal | Wind-direction arrows, fault strike symbols |
| Texture (Pattern) | Spacing/density of hatch or dot fills | Nominal / Ordinal | Swamp stipple, diagonal hatching |
7.2 Thematic mapping and data classification methods
A thematic map portrays the spatial pattern of a specific attribute (e.g., literacy rate, rainfall, COVID-19 incidence) over a geographic reference frame.
Four Core Thematic Map Types
- Choropleth Map: Shades predefined enumeration polygons (districts, wards) according to aggregated statistical values.
- Golden rule: Always map normalized rates or densities (e.g., persons per , percentage literacy, income per capita), never raw totals on unequal-area polygons (otherwise large polygons visually dominate simply because of their area).
- Graduated / Proportional Symbol Map: Scales point symbols (usually circles or squares) placed at feature centroids proportionally to raw count/magnitude (e.g., total city population, reservoir storage volume). Proportional symbols use continuous scaling ( for circle area), whereas graduated symbols bin values into – discrete size classes.
- Dot Density Map: Places uniform dots inside polygons where each dot represents a fixed quantity (e.g., ). Communicates spatial concentration visually without distorting polygon areas, though dot placement inside polygons is typically random unless constrained by land cover.
- Isarithmic (Isoline / Contour) Map: Connects points of equal value along smooth lines for continuous phenomena—either isometric (measurable at points, e.g., elevation contours, isotherms, isobars, isohyets) or isopleth (derived ratios such as population potential).
Data Classification Methods
When grouping continuous values into classes (typically –) for choropleth or graduated maps, class boundaries strongly alter the reader's impression:
| Method | Break Formula / Criterion | Strengths | Weaknesses |
|---|---|---|---|
| Equal Interval | Constant class width: | Intuitive legend; easy cross-map comparison | Skewed data leaves some classes empty and jams most features into one class |
| Quantile | Equal number of features () in each class | Guarantees balanced visual representation across all colors | Features with nearly identical values may split across classes; wide range inside extreme bins |
| Natural Breaks (Jenks) | Minimizes within-class variance and maximizes between-class variance | Highlights real clusters and natural gaps in the distribution | Break thresholds are dataset-specific, making multi-year comparison harder |
| Standard Deviation | Class breaks placed at | Clearly emphasizes outliers above/below the mean using a diverging ramp | Assumes roughly normal distribution; hides raw data units unless labeled |
1
Eight districts report annual tuberculosis incidence rates (cases per residents) sorted in ascending order:
Classify these values into classes using:
- Equal Interval classification.
- Quantile classification.
- Briefly explain why Jenks Natural Breaks would differ on this specific dataset.
Equal Interval classification (): Compute the common interval width:
The four class ranges and counts are:
- Class 1 (): ( districts)
- Class 2 (): ( districts)
- Class 3 (): ( districts — empty class!)
- Class 4 (): ( districts)
Quantile classification (): Each class receives districts in sorted order:
- Class 1: (range –)
- Class 2: (range –)
- Class 3: (range –)
- Class 4: (range –)
Why Jenks Natural Breaks differs: Inspecting the data shows three natural clusters separated by wide gaps ( and ): , , and (plus splitting and if ). Unlike Equal Interval (which leaves Class 3 empty) and Quantile (which arbitrarily splits from ), Jenks aligns class borders directly with those natural gaps to minimize within-class variance.
- Equal Interval (): (4 items), (2 items), (0 items), (2 items).
- Quantile ( items/class): , , , .
- Jenks Natural Breaks places boundaries at the wide numerical gaps (– and –) rather than leaving empty bins or splitting tight clusters.
7.3 Visualization of time-series, 3D and multi-scale data; dashboards and story maps
Modern geo-visualization goes beyond static 2D paper sheets to represent time, elevation, scale, and narrative context interactively:
- Time-Series Visualization:
- Small Multiples: Side-by-side map panels arranged chronologically for direct visual comparison without animation memory load.
- Animated Time-Sliders: Step sequentially through timestamps (e.g., hourly rainfall radar or annual urban sprawl).
- Space-Time Cubes & Sparkline Maps: Combine 3D trajectories or embed mini line charts on map regions.
- 3D Visualization:
- 2.5D Extrusion & Draping: Drapes orthophotos or thematic layers over a DEM/TIN surface, or extrudes building footprints by floor height.
- True 3D Scenes: Renders volumetric point clouds, underground utility tunnels, and textured CityGML / 3D Tiles buildings with interactive camera fly-throughs and shadow/viewshed simulation.
- Multi-Scale Visualization:
- Dynamically adapts level of detail (LOD) as the user zooms—showing generalized country polygons at zoom , simplified highways at zoom , and building footprints plus street labels at zoom (scale-dependent visibility rules and progressive vector simplification).
- Geospatial Dashboards and Story Maps:
- Operational Dashboards: Combine an interactive map with synchronized live KPIs, bar charts, gauges, and filter widgets (cross-filtering: panning the map updates all charts in real time), widely used in emergency operations centers and utility monitoring.
- Story Maps: Pair guided interactive maps, narrative prose, photographs, and scroll-triggered map viewpoints into cohesive multimedia reports for policy makers and the public.
7.4 Web map tiles: raster and vector tiles, tiling schemes and zoom levels
Serving millions of raw vector vertices or gigabytes of un-tiled imagery to a browser on every pan/zoom is impossible. Modern web mapping solves this using pyramid tile caches partitioned into a regular grid of square tiles at discrete zoom levels .
Tiling Schemes and Zoom Levels
- World Extent & Zoom Pyramid: In the standard Web Mercator (
EPSG:3857) Slippy Map (XYZ) scheme:- At zoom level , the entire world ( latitude, longitude) fits into single tile (typically or ).
- Each increment splits every tile into child tiles (), doubling horizontal and vertical pixel resolution.
- Thus, at zoom level , the grid has columns by rows, totaling:
- Tile Addressing Conventions:
- XYZ (Google / OpenStreetMap / Slippy Map): Origin at top-left (north-west); increases east, increases south. URL pattern:
/{z}/{x}/{y}.png(or.pbf). - TMS (OSGeo Tile Map Service): Origin at bottom-left (south-west); .
- WMTS (OGC Web Map Tile Service): Standardized XML/KVP or RESTful tile matrix set supporting arbitrary projections beyond Web Mercator.
- XYZ (Google / OpenStreetMap / Slippy Map): Origin at top-left (north-west); increases east, increases south. URL pattern:
Raster Tiles vs. Vector Tiles
| Aspect | Raster Map Tiles | Vector Map Tiles (MVT) |
|---|---|---|
| Payload Content | Pre-rendered pixel image ( or ) | Compressed protocol-buffer binary geometry + attributes clipped to tile grid |
| Where Styling Happens | Fixed on the server at render time | Dynamic on the client GPU (WebGL via MapLibre GL / OpenLayers) |
| Zoom & HiDPI Sharpness | Blurs when over-zoomed or on Retina screens unless tiles served | Razor-sharp at any fractional zoom or device DPI |
| Rotation & Labels | Rotating the map turns baked-in text labels upside down | Labels stay upright and dynamically reposition on rotation/pitch |
| Payload Size & Interactivity | Larger for rich maps (–); no feature attributes without UTFGrid/WMS | Very compact (–); supports hover/click queries directly on features |
| Best Use Case | Satellite imagery, hillshade DEMs, legacy browsers | Interactive basemaps, multi-theme styling, offline mobile vector maps |
2
In a standard Web Mercator XYZ slippy map ( tiles, equatorial Earth circumference ):
- How many total tiles cover the globe at zoom level ?
- Convert XYZ tile coordinate to its TMS -index.
- Compute the ground pixel resolution (meters per pixel) at the equator for .
Total tiles at zoom level :
Convert XYZ -index to TMS -index: Because XYZ counts downward from the North Pole ( to ) while TMS counts upward from the South Pole:
So the TMS coordinate is .
Compute equatorial ground resolution at : Total equatorial pixel width at zoom is :
- Zoom contains tiles ().
- The equivalent TMS -coordinate is .
- The equatorial ground resolution at is .
added
List the eight standard map elements and explain how visual hierarchy and figure–ground contrast improve cartographic communication.
Eight standard map elements:
- Map body (data frame), 2. Title/subtitle, 3. Legend, 4. Scale bar/RF, 5. North arrow/graticule, 6. Inset/locator map, 7. Neatline/frame, 8. Credits/source/CRS metadata.
Role of visual hierarchy and figure–ground contrast:
- Visual hierarchy ranks features visually from most important (bold colors, thick lines, large type for thematic data and title) to least important (muted gray roads, thin neatline), guiding the reader's eye immediately to the map's message.
- Figure–ground contrast separates the study area ("figure") from surrounding regions or water ("ground") via lightness difference, drop shadows, or crisp boundary outlines so the reader never confuses land and sea or focus and context.
The eight elements are map body, title, legend, scale bar, north arrow/graticule, inset map, neatline, and source/CRS credits. Visual hierarchy orders information by visual prominence, while figure–ground contrast visually separates the focal study area from background context.
Using Bertin's visual variables, recommend appropriate symbolization for: (a) Land-cover classes, (b) Total city population, (c) Literacy percentage by district, and (d) Road hierarchy (Highway > Collector > Local). Justify each choice.
Match visual variables to attribute measurement scales:
- (a) Land-cover classes (Nominal): Use Color Hue (green for forest, blue for water, yellow for cropland) and/or Texture, because hue communicates qualitative category differences without implying order.
- (b) Total city population (Ratio raw count): Use Size (proportional or graduated circles at city centroids), because circle area naturally encodes quantitative magnitude without polygon-area distortion.
- (c) Literacy percentage by district (Ratio normalized rate): Use Color Value (Lightness) in a sequential monochrome ramp on a choropleth map, because lighter-to-darker shades communicate ordered intensity intuitively.
- (d) Road hierarchy (Ordinal): Use Size (line stroke width) combined with Color Value, making highways widest/darkest and local streets narrowest/lightest.
(a) Color Hue (nominal categories); (b) Symbol Size (raw count magnitudes via proportional circles); (c) Color Value/Lightness (choropleth sequential ramp for normalized rates); (d) Line Size + Value (ordinal ranking).
Why should a choropleth map display normalized rates (e.g., population density) rather than raw counts (e.g., total population)? Classify the six values into classes using Equal Interval and Quantile methods.
Why choropleth maps require normalization: Enumeration polygons (districts or wards) vary greatly in land area. Shading polygons by raw totals causes geographically huge rural polygons to appear artificially dominant compared to tiny dense urban polygons; normalizing by area or denominator population removes polygon-size bias.
Equal Interval () and Quantile () classification:
- Equal Interval: Width .
- Class 1 ():
- Class 2 ():
- Class 3 ():
- Quantile ( items per class):
- Class 1:
- Class 2:
- Class 3:
Choropleth maps must use normalized rates so large-area polygons do not visually exaggerate raw totals. Equal Interval () gives , , ; Quantile ( items/bin) gives , , .
Compare geospatial dashboards and story maps, and name two techniques each for visualizing time-series and 3D geographic data.
Dashboards vs. Story Maps:
- Operational Dashboards focus on real-time situational monitoring via linked interactive maps, live KPIs, gauges, and cross-filtering charts for analysts and decision-makers.
- Story Maps weave interactive maps with narrative text, photos, and guided scroll viewpoints to communicate a structured geographic story to broader audiences.
Time-series and 3D visualization techniques:
- Time-series: (1) Small multiples (chronological grid of static maps), (2) Interactive time-slider animations (or space-time cubes).
- 3D data: (1) DEM terrain draping & 2.5D building extrusion, (2) True 3D point-cloud / CityGML mesh scenes.
Dashboards deliver real-time linked KPIs and interactive filtering for monitoring, while story maps deliver guided narrative storytelling. Time-series data uses small multiples and time-slider animations; 3D data uses 2.5D extrusion/terrain draping and true 3D CityGML/point-cloud scenes.
Contrast raster map tiles and vector map tiles (MVT) across four criteria. For zoom level in an XYZ slippy map, compute the total number of tiles covering the world and convert XYZ tile to its TMS -index.
Raster vs. Vector tiles (four criteria):
- Format: Raster tiles send pre-rendered PNG/JPEG pixels; vector tiles send compressed
.mvt/.pbfgeometry + attributes. - Styling: Raster styling is fixed server-side; vector tiles are styled dynamically on the client GPU (WebGL).
- Zoom/Rotation: Raster blurs when over-zoomed and rotates baked labels upside down; vector tiles stay crisp at any DPI and keep labels upright.
- Best use: Raster suits satellite imagery/hillshades; vector suits interactive multi-theme basemaps and offline mobile apps.
Compute tile count and TMS -index at :
Vector tiles transfer compact geometry/attributes styled client-side with crisp labels and rotation support, whereas raster tiles transfer fixed server-rendered images ideal for satellite imagery. At , there are tiles, and XYZ converts to TMS .