Skip to content
LITHOEN

Tectonic plate

Pacific Plate

The largest tectonic plate, dominantly oceanic, whose boundaries include the San Andreas transform system and the subduction zone marked by the Mariana Trench.

World map showing twenty-four plate outlines as thin dark lines over a pale coastline, with heavier red plate-boundary curves over them, subduction segments dashed, and thirteen hatched purple regions marking zones of distributed deformation.
Twenty-four tectonic plates as outlined in model PB2002 (Bird, 2003), with the boundary curves Lithoen holds for them and the thirteen regions the model calls orogens. Every line here is a model boundary rather than a mapped geological contact: the model is a compilation from published sources at varying scales, its author states that accuracy never equals precision, and boundary class changes along a boundary — so only the segments the source labels subduction are shown as such. Outlines are generalised for display at a recorded tolerance. Areas near the poles are severely exaggerated by the projection, which is stated on the figure.

Figure by Lithoen. Plate model PB2002 from Bird (2003), doi:10.1029/2001GC000252; GIS conversion by Hugo Ahlenius, Nordpil, ODC-By 1.0. Made with Natural Earth. Underlying data remain under their own licences. · LITHOEN-ORIGINAL · Derivative: Generated from published Lithoen spatial records by scripts/generate-maps.ts; no coordinate was edited by hand. Features spanning ±180° are split for rendering only; the stored geometry is unchanged.

Classification

plate class
major

How this is mapped

2 published representations of where Pacific Plate is. Each is one interpretation, by one identified source, at one scale — not the shape of the thing itself.

  • Tectonic model geometry

    MultiPolygon · 507 positionspreferred for publication

    Geometry that exists because a model defines it — a rigid-plate outline, a zone where a model declares itself inapplicable. It is a property of the model, and changes when the model is revised. It is present-day geometry from a present-day model; a palaeogeographic reconstruction is a different thing and Lithoen does not hold one.

    The Pacific Plate as outlined in model PB2002 (Bird, 2003). This is the model’s boundary, not a mapped geological contact: it is where a rigid-plate model puts the edge of a plate, and Bird’s own documentation states that coordinates are given to a precision of about 60 m while "accuracy never equals precision". Where deformation is distributed rather than confined to a line, the model marks the region as an orogen and the outline there is a convenience.

    Boundary certainty
    InterpretedThe boundary was placed by a geologist reading evidence: mapping, geophysics, geochronology, structure. It is reproducible from that evidence and it is an interpretation of it. Most geological boundaries are this.
    Status of this geometry
    modelledThe status of the geometry, which need not match the status of what the entity’s claims assert.
    Dataset
    PB2002 plates, boundaries and orogens as GeoJSON (fraxen/tectonicplates)Plate model PB2002 from Bird (2003), doi:10.1029/2001GC000252; GIS conversion by Hugo Ahlenius, Nordpil, ODC-By 1.0.

    What was done to this geometry

    • By Lithoen — Douglas–Peucker, 1190 positions to 507
    • By Lithoen — Coordinates rounded to 4 decimal places, about 11 m of longitude at the equator.
    • Before it reached Lithoen — Bird (2003) digitised the boundary curves from published sources at varying scales; the model is a compilation, not a single survey.
    • Before it reached Lithoen — The GIS conversion (Ahlenius, ODC-By 1.0) split segments spanning the -180/180 meridian and moved them, as its README states. Four plates therefore arrive as multi-part geometries.

  • Boundary

    MultiLineString · 30 positions

    A line delimiting or separating things — a plate boundary, a suture, a mapped contact. A boundary is not an extent: it says where the edge runs, not what lies on either side.

    Pacific–Philippine Sea boundary as drawn in model PB2002 (Bird, 2003), from source segments PS/PA. The Mariana subduction zone, where the Pacific Plate descends beneath the Philippine Sea Plate and the sea floor reaches its greatest known depth. The model labels these segments as subduction. Digitised by: by Peter Bird, 1999.

    Boundary certainty
    InterpretedThe boundary was placed by a geologist reading evidence: mapping, geophysics, geochronology, structure. It is reproducible from that evidence and it is an interpretation of it. Most geological boundaries are this.
    Status of this geometry
    modelledThe status of the geometry, which need not match the status of what the entity’s claims assert.
    Dataset
    PB2002 plates, boundaries and orogens as GeoJSON (fraxen/tectonicplates)Plate model PB2002 from Bird (2003), doi:10.1029/2001GC000252; GIS conversion by Hugo Ahlenius, Nordpil, ODC-By 1.0.

    What was done to this geometry

    • By Lithoen — 2 source features carrying the same identifier were combined into one multi-part geometry; no coordinate was altered by the combination.
    • By Lithoen — Coordinates rounded to 4 decimal places, about 11 m of longitude at the equator.
    • Before it reached Lithoen — Bird (2003) compiled these curves from published sources at varying scales, crediting each segment to its own source; those credits are carried in the caption of every record.
    • Before it reached Lithoen — The GIS conversion (Ahlenius, ODC-By 1.0) split segments spanning the -180/180 meridian and moved them, as its README states.

How this record is known

10
Settled consensus
Cited as the accepted position of the discipline.

10 of 10 claims on this page cite at least one registered source. Counts describe how each statement is known, not how likely it is to be true.

What Lithoen records

Each claim states how it is known and cites its evidence.

EstablishedODC-By 1.0: derivatives permitted, attribution required, and carried into every derived figure.

A property of a dataset or model, shown here for context — not an observation about this record.

Every plate geometry Lithoen publishes is derived from a conversion of model PB2002 made available under the Open Data Commons Attribution License 1.0, which permits derivative publication and requires attribution. The repository asks that Hugo Ahlenius, Nordpil and Peter Bird be credited, and Lithoen renders that credit wherever a derivative appears — including on maps it draws itself, because drawing a map from someone else’s data does not make the data Lithoen’s.

  • definition basis: The licence was read in LICENSE.md and in the repository README on the date recorded on dataset:pb2002-tectonicplates-geojson.
  • precision note: The original PB2002 files on the author’s own site carry no licence at all. Lithoen therefore derives geometry only from the explicitly licensed conversion, and cites the original model as the source of the science.
  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Ahlenius 2014 Peer-reviewed

EstablishedThe dateline split is the converter’s edit, disclosed and attributed, not Lithoen’s and not the Earth’s.

A property of a dataset or model, shown here for context — not an observation about this record.

Four plates arrive from the conversion already split at ±180°, because the converter states that the main edits it made "regarded segments spanning the -180/180 boundary, which had to be manually split and moved". Lithoen records that as an inherited transformation on every affected geometry rather than presenting the split as its own, and never reports it as evidence that a plate is in pieces.

  • precision note: A transformation nobody performed on Lithoen’s side is still a transformation between the Earth and the reader, and disclosing only one’s own edits would misrepresent the chain.
  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Ahlenius 2014 Peer-reviewed

EstablishedPrecision is a property of the number; accuracy is a property of the knowledge.

General principle, shown here for context — not an observation about this record.

A coordinate can be precise and wrong. Bird’s own documentation for model PB2002 gives every point to six significant digits, so that round-off error in position does not exceed 60 metres, and then states plainly: "accuracy never equals precision." The number of decimal places a boundary is stored to says nothing about how well anybody knows where it is.

  • precision note: The controlling claim of the Phase 4 spatial layer. Every geometry Lithoen publishes carries a boundary certainty and an epistemic status of its own precisely because the coordinates cannot carry them.
  • generality: GENERAL_PRINCIPLE

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

EstablishedNorthwesterly motion relative to the North American Plate along the San Andreas fault zone.

Along the San Andreas fault zone, land on the Pacific Plate moves in a northwesterly direction relative to land on the North American Plate.

  • reference frame: Relative to the North American Plate.
  • locality: Along the San Andreas fault zone.

Evidence support: Supported by the cited evidence one source

Evidence: Kious 1996 Primary authority

EstablishedSUB, plus OSR, OTF, OCB, CRB, CTF and CCB from setting and relative velocity.

A property of a dataset or model, shown here for context — not an observation about this record.

PB2002 recognises seven classes of plate boundary step. Subduction requires a convergent velocity component together with a Benioff zone of intermediate to deep earthquakes and/or a parallel volcanic arc with Quaternary activity — criteria which, the author notes, exclude the Alps, Zagros, Pamirs, Himalaya, Tien Shan and Taiwan. The other six combine oceanic or continental setting with divergent, strike-slip or convergent motion.

  • precision note: The exclusions matter more than the classes. Convergence across the Himalaya is not in doubt; what the criteria exclude is the label, because there is no Benioff zone and no Quaternary arc there.
  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

EstablishedBoundary class varies along a boundary, so a boundary has no single class.

A property of a dataset or model, shown here for context — not an observation about this record.

A single class cannot always be specified for an entire plate boundary curve, or even for an entire segment, because the class changes with local changes in the azimuth of the boundary; relative velocities change smoothly in magnitude and azimuth along every boundary.

  • precision note: Lithoen therefore labels a boundary geometry as subduction only where the source does, and classifies no further. Where the source declines to classify, so does the corpus.
  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

Established52 plates, 229 boundary segments, 13 orogens.

A property of a dataset or model, shown here for context — not an observation about this record.

Model PB2002 comprises plate boundary curves in 229 segments, 52 plates given as closed counterclockwise outlines, and 13 orogens given as outlines with spelled-out names, chosen — the author writes — to help prevent confusion between orogens and plates.

  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

EstablishedBoundary segments carry individual source credits within the model.

A property of a dataset or model, shown here for context — not an observation about this record.

The segments of PB2002 are individually credited: different parts of a single plate boundary may be attributed to different sources, and the model records those attributions in its segment titles.

  • precision note: Lithoen carries each segment’s credit into the caption of the geometry derived from it, so that merging segments for display never erases who drew them.
  • generality: DATASET_PROPERTY

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

EstablishedEvery plate boundary Lithoen shows belongs to a named, dated, citable model.

General principle, shown here for context — not an observation about this record.

A plate boundary is a line in a model. PB2002 is one such model, published in 2003, dividing Earth into 52 plates; other models divide it differently, and a boundary’s position, its class and even its existence depend on which model is being read. Lithoen names the model on every plate geometry it publishes and never draws an unattributed boundary.

  • model: PB2002, from Bird (2003), doi:10.1029/2001GC000252.
  • generality: GENERAL_PRINCIPLE

Evidence support: Supported by the cited evidence 2 sources, multiple sources, one origin

Evidence: Bird 2003 Primary authority; Bird 2003 Primary authority

EstablishedPlates are lithospheric, not geographic; 52 of them, and 13 regions where the model fails.

General principle, shown here for context — not an observation about this record.

A plate is not a continent. Model PB2002 divides Earth into 52 plates, most of which are dominantly oceanic, several of which carry parts of two continents, and several of which carry no continent at all. It also marks 13 regions where no rigid plate describes the motion.

  • generality: GENERAL_PRINCIPLE

Evidence support: Supported by the cited evidence one source

Evidence: Bird 2003 Primary authority

In the knowledge graph

Record last updated 2026-09-06. Corrections follow the published corrections policy; superseded claims are retained, never silently edited.