Skip to content
CARTOLOGIUMThe living atlas
Survey desk + sources & coordinates
WGS 84

WAYS OF MAPPING / INTO THE EARTH

How to map
the inside
of Earth.

The deepest maps begin with something that reaches the surface. Follow a signal into a world we cannot photograph.

I / FROM SURFACE TO CENTRE6,371 km
MantleOuter coreInner coreSURFACE COMPOSITEINTERIOR MODEL
One planet, two kinds of image. NASA’s surface composite meets an original section based on PREM’s broad radii. Interior colors are conventions; the surface rim is emphasized.
NASA’s Earth ObservatorySource & rights +

Blue Marble 2002 — western hemisphere. NASA Science.

NASA imagery credited to NASA’s Earth Observatory; no third-party copyright notice on this source record. Public domain in the United States; NASA identifiers and endorsement are excluded. The cutaway is Cartologium’s separate illustration.

Open source record ↗

A signal is not yet a picture.

On a street map, you can point to a building and then visit it. A map of the deep Earth asks for another kind of trust. Its colors stand for properties inferred from evidence: they are not colors seen through a window in the ground.

Even the familiar blue globe above is an assembled view. NASA’s 2002 Blue Marble combines satellite observations collected over several months. That composition describes the visible surface and clouds. Joining it to a cutaway does not turn the interior into a satellite image. The seam marks a change in how the image is known.

Earthquakes release seismic waves that instruments can record at the surface. A seismogram traces the instrument’s response through time. In the USGS example below, the labels identify an earlier P arrival and a later S arrival. P waves involve compression; S waves involve shear. The two types do not travel in the same way.

II / READ THE ARRIVALSUSGS EDUCATIONAL EXAMPLE
USGS educational waveform example with the P and S arrivals labelled

PCompression
The earlier labelled arrival

SShear
The later labelled arrival

An educational waveform example from USGS. The source does not identify an event, station or calibrated axes. Use it to recognize the labels, rather than calculate a distance or magnitude.
U.S. Geological SurveySource & rights +

Example of seismic waveform. USGS Earthquake Hazards Program.

The USGS media record explicitly identifies this image as Public Domain. Its event, station, time and amplitude calibration are not identified; it is reproduced as an educational example.

Open source record ↗

Look first at what this example actually provides: a trace and two labels. It does not supply a named earthquake, station, calibrated time axis or amplitude scale. Those missing particulars matter. A plausible-looking wiggle cannot, by itself, tell us where an earthquake happened or what the entire planet contains.

What was measured—and what had to be inferred?

The planet becomes a profile.

One useful simplification is to describe how properties change with depth, setting aside differences between one longitude and another. The Preliminary Reference Earth Model, or PREM, is a published global reference from 1981. Its tabulated seismic velocities make an interior readable as a profile. The figure below uses EarthScope’s version calculated for a period of one second.

III / A MODEL, IN NUMBERSPREM · 1981 · 1 s

DEPTH BELOW SURFACE / km

02,0004,0006,371051015

RADIAL WAVE SPEED / km s⁻¹

P · compressionS · shear
READ THE SELECTED LAYER

Outer core

2,891–5,149.5 km deep

The liquid shell of the core. The reference profile sets shear-wave speed to zero here, while compressional waves still propagate.

Model values describe an averaged Earth. The thin surface layers are difficult to resolve at this whole-planet scale.

Original plot of EarthScope’s PREM 1-second table: Vpv and Vsv, the radial P and S velocities. Lines join tabulated values, preserving both sides of discontinuities. They are model values, not earthquake recordings.
Read the model values and method

Depth = 6,371 km − radius. The plot retains every source row, including repeated depths at boundaries. PREM includes anisotropy in the upper mantle; this figure uses its radial components only. The source’s ocean layer also has zero S-wave speed.

EarthScope model record ↗ · Original numeric table ↗
PREM, period 1 second
Depth (km)P (km/s)S (km/s)
01.450
11.450
21.450
31.450
35.83.2
45.83.2
55.83.2
65.83.2
75.83.2
85.83.2
95.83.2
105.83.2
115.83.2
125.83.2
135.83.2
145.83.2
155.83.2
156.83.9
166.83.9
176.83.9
186.83.9
196.83.9
206.83.9
216.83.9
226.83.9
236.83.9
24.46.83.9
24.48.022064.39602
258.021384.39616
268.020244.39639
278.019114.39662
288.017984.39685
298.016844.39708
308.015714.39731
318.014584.39754
328.013454.39777
338.012314.398
348.011184.39823
358.010054.39846
408.004384.39962
457.998724.40077
507.993054.40192
607.981724.40422
707.970394.40653
807.959064.40883
807.959064.40883
907.947734.41113
1007.936414.41344
1107.925084.41574
1207.913754.41805
1307.902424.42035
1407.891094.42265
1507.879764.42496
1607.868434.42726
1707.85714.42957
1807.845774.43187
1907.834444.43417
2007.823114.43648
2107.811784.43878
2207.800454.44109
2208.558954.6439
2308.578194.6509
2408.597434.6579
2508.616664.6649
2608.63594.6719
2708.655144.6789
2808.674384.6859
2908.693624.6929
3008.712864.6999
3108.73214.7069
3208.751334.7139
3308.770574.7209
3408.789814.7279
3508.809054.7349
3608.828294.7419
3708.847534.7489
3808.866774.7559
3908.8864.7629
4008.905244.7699
4009.133924.93249
4109.185114.96166
4209.236314.99083
4309.28755.02
4409.33875.04918
4509.389895.07835
4609.441095.10752
4709.492285.13669
4809.543485.16586
4909.594685.19504
5009.645875.22421
5109.697075.25338
5209.748265.28255
5309.799465.31173
5409.850655.3409
5509.901855.37007
5609.953045.39924
57010.004245.42841
58010.055435.45759
59010.106635.48676
60010.157835.51593
60010.157765.51602
61010.173255.52376
62010.188745.5315
63010.204225.53924
64010.219715.54699
65010.23525.55473
66010.250695.56247
67010.266175.57021
67010.751325.94513
685.310.798975.98986
735.310.954596.13609
77111.06566.24054
77111.065596.24039
785.311.091796.25067
835.311.181926.28608
885.311.269846.32065
935.311.355636.35442
985.311.439366.38742
1035.311.52116.41967
1085.311.600956.4512
1135.311.678966.48204
1185.311.755236.5122
1235.311.829836.54173
1285.311.902836.57064
1335.311.974316.59897
1385.312.044356.62674
1435.312.113046.65398
1485.312.180436.68071
1535.312.246626.70696
1585.312.311676.73276
1635.312.375676.75813
1685.312.43876.78311
1735.312.500836.80771
1785.312.562136.83197
1835.312.622696.85592
1885.312.682586.87957
1935.312.741886.90296
1985.312.800676.92611
2035.312.859026.94906
2085.312.917026.97182
2135.312.974736.99442
2185.313.032247.0169
2235.313.089637.03927
2285.313.146967.06157
2335.313.204327.08382
2385.313.261797.10605
2435.313.319447.12828
2485.313.377357.15055
2535.313.43567.17287
2585.313.494267.19529
2635.313.553417.21781
2685.313.613137.24048
2735.313.67357.26331
274113.680427.26593
274113.680447.26597
2785.313.690987.26563
2835.313.703027.2652
2885.313.715227.26471
289113.716627.26465
28918.064790
29718.199430
30718.360240
31718.513020
32718.658090
33718.795770
34718.926360
35719.050190
36719.167560
37719.27880
38719.384220
39719.484130
40719.578850
4149.59.649770
41719.668690
42719.753970
43719.8350
44719.91210
45719.985580
467110.055760
477110.122950
487110.187470
497110.249630
507110.309750
5149.510.355720
5149.511.028263.50431
5153.911.029943.50549
517111.036423.51002
527111.072493.53522
537111.105413.55823
547111.13523.57905
557111.161863.59767
567111.185373.61411
577111.205763.62835
587111.2233.64041
597111.237113.65027
607111.248093.65794
617111.255933.66342
627111.260633.6667
637111.26223.6678

Select the outer core and follow the dashed S curve. Its zero value represents a liquid region that does not transmit shear waves. The P curve continues. At the inner-core boundary, the model gives shear speed a nonzero value again. This is a property of the modeled material, not a claim that an ordinary S wave can pass straight through the intervening liquid.

The large rocky mantle is predominantly solid, even though it can deform and flow over geological time. The core has liquid and solid regions. A cutaway that colors all three in warm tones can accidentally make the planet look like a bowl of molten rock. Here the colors separate regions; they do not measure temperature.

At this scale, another choice becomes visible: the crust almost disappears. Enlarging it would make it easier to discuss, but would change the scale of the drawing. The globe emphasizes its surface rim; the numeric profile retains the source’s thin surface layers. Neither view should be mistaken for the crustal thickness beneath a particular city.

From a layered Earth to a mapped Earth.

A radial profile makes the same description available at every surface location. To investigate variation from place to place, seismic tomography combines signals that have sampled different paths. A proposed velocity model predicts travel times; comparing predictions with recordings helps researchers revise it. The result is an inferred distribution of wave speeds.

IV / MANY PATHS, DIFFERENT CONSTRAINTSSCHEMATIC

A single path leaves many possible interiors consistent with it.

Crossing paths interrogate a region from several directions.

Original diagram of sampling geometry. These are illustrative paths, not computed seismic rays or a recovered tomogram. More paths do not automatically guarantee a correct model.

The arrangement of the measurements matters as much as the colors in the result. Sampling geometry, noise, assumptions and the methods used to stabilize an inversion affect what can be resolved. Research on geophysical tomograms shows why a sharply drawn boundary or a fine output grid does not automatically establish equally fine knowledge.

Before reading an interior map as a landscape, find its depth, the property being plotted, the units, the reference model and the evidence for resolution. A velocity difference is not itself a photograph of a hot plume, a mineral or a reservoir of melt. Such interpretations need further physical evidence.

Read one boundary, twice.

Return to the profile and select the outer core. Compare what happens to P and S at its upper boundary. Then return to the globe. The color change and the change in the plotted properties describe the same broad boundary through different visual languages.

Return to the model ↑

Mapping the inside out begins with this discipline: keep the observation, the model and the interpretation close enough to compare. It lets a reader move from a mark on a record to a structure inside the planet without losing the reasoning between them.

Sources and method.

The profile plots the published table’s depth, Vpv and Vsv columns in kilometres and kilometres per second. All rows are retained, including repeated depths at discontinuities. The broad cutaway uses a 6,371 km Earth radius, a 3,480 km core radius and a 1,221.5 km inner-core radius. It omits the model’s finer subdivisions.

The NASA composite and USGS example have individual source and rights controls. Cartologium’s cutaway, plot and sampling diagram are original visual explanations. No computed ray tracing, local tomographic model or new earthquake observation is presented.

Inspect the EarthScope model record · Read the source table. The table was retrieved on 11 October 2026. Its one-second period is a model parameter, not a record-update interval.

  1. NASA. (2018, April 4). Blue Marble 2002 [Satellite-data composite]. https://science.nasa.gov/resource/blue-marble-2002/

    Inspect the source
  2. Wald, L. (n.d.). The science of earthquakes. U.S. Geological Survey. https://www.usgs.gov/programs/earthquake-hazards/science-earthquakes

    Inspect the source
  3. U.S. Geological Survey. (n.d.). Inside the Earth. This dynamic Earth. https://pubs.usgs.gov/gip/dynamic/inside.html

    Inspect the source
  4. Dziewonski, A. M., & Anderson, D. L. (1981). Preliminary reference Earth model. Physics of the Earth and Planetary Interiors, 25(4), 297–356. https://doi.org/10.1016/0031-9201(81)90046-7

    Inspect the source
  5. EarthScope Consortium. (n.d.). Seismic tomography. https://www.earthscope.org/what-is/seismology/seismic-tomography/

    Inspect the source
  6. Day-Lewis, F. D., & Lane, J. W. (2004). Assessing the resolution-dependent utility of tomograms for geostatistics. Geophysical Research Letters, 31, L07503. https://doi.org/10.1029/2004GL019617

    Inspect the source
  7. U.S. Geological Survey. (n.d.). Example of seismic waveform [Educational illustration]. https://www.usgs.gov/media/images/example-seismic-waveform

    Inspect the image record

Editorial text and original diagrams © Cartologium. NASA and USGS imagery retain their individual source rights. Prepared 11 October 2026.