Temporal Raster Plot¶
A nested-time grid places observations by interval and position within that interval, encoding each cell's value to compare recurring temporal structure.
Core Idea¶
A temporal raster plot folds a long time series into a two-dimensional grid: one coordinate selects a repeated interval such as a day or water year, and the other selects position within it. Color or another cell mark encodes the measured value. Aligning corresponding phases makes within-interval and across-interval differences inspectable, but apparent bands are visual clues, not proof of periodicity or causation.[ref-8a65355f9ed8][ref-d16832c13c9b]
Scope of Application¶
A building-energy carpet plot can compare measured and simulated performance, as the LBNL/OSTI abstract verifies, but that abstract does not specify axes or case results. A USGS Minnesota River raster seasonality plot, Figure 16, compares water years by within-year position using blue for larger flow, green for middle flow, tan for lower flow, and white for no data. The same structure works for other measures only with an explicit interval, time zone, binning and missingness policy.[ref-8a65355f9ed8][ref-d16832c13c9b]
Clarity¶
In an author-constructed 2×2 grid, Monday's 09:00 and 10:00 loads are 10 and 14 kW, while Tuesday's are 12 and 8 kW. Day columns and hour rows reveal that the morning comparison changes direction across the two hours; the invented numbers are not LBNL observations. In USGS Figure 16, the report describes larger Minnesota River flows concentrated roughly April–mid-July before 1960 but mid-March–September afterward, with more green/blue and fewer tan cells by the late 2000s. This is a plotted contrast, not causal proof.[ref-8a65355f9ed8][ref-d16832c13c9b]
Manages Complexity¶
The grid compresses many observations and makes corresponding temporal positions easy to compare, but splits adjacent dates across row boundaries. TimeElide Figure 1(f–g) shows that absolute and normalized within-slice timing align unequal intervals differently. Color bands can obscure exact values, and coloring no-data white cells as low-flow tan would invent drought. Important patterns need the original series as a check.[ref-d16832c13c9b][ref-b74011894485]
Abstract Reasoning¶
Choose an interval matched to the analytic question, map observations to interval and within-interval position, define aggregation and missingness, then inspect the resulting grid. Test important patterns against raw timestamps, alternative folds and honest color scales.[^ref-b74011894485]
Knowledge Transfer¶
The portable pattern is folding a sequence into repeated slices and comparing aligned positions. This named visualization requires two nested time coordinates and measured cells; a generic heat map or spatial raster does not automatically have that structure. Run Chart is a one-axis contrast, not a verified parent. A broader fold-and-align prime would need evidenced non-visual transfer.
[^ref-8a65355f9ed8]: LBNL/OSTI SEE IT carpet-plot report; accessible official abstract only. [^ref-d16832c13c9b]: Tara Williams-Sether and Chris Sanocki (2025), USGS Minnesota streamflow trends report, Figure 16, DOI 10.3133/sir20235064E; publisher-indexed text describes the figure, without supplying individual cell values here. [^ref-b74011894485]: Original TimeElide visualization research (2021).
Relationships to Other Abstractions¶
Current abstraction Temporal Raster Plot Domain-specific
Parents (1) — more general patterns this builds on
-
Temporal Raster Plot is a kind of Visualization (graphics) Domain-specific
A temporal raster plot specializes visualization by mapping repeated time intervals and within-interval positions to value-coded cells.
Hierarchy path (1) — routes to 1 parentless root
- Temporal Raster Plot → Visualization (graphics)
Neighborhood in Abstraction Space¶
Temporal Raster Plot sits in a sparse region of the domain-specific corpus (98th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (2551 abstractions)
Nearest neighbors
- Actogram — 0.79
- Ridgeline Plot — 0.76
- Bubble Chart — 0.75
- Elapsed-Time Memory Decay — 0.75
- Degree Day — 0.75
Computed from structural-signature embeddings · 2026-10-08