The Gold Standard of Soil Science | Shrink–Swell and the Reference Volume in Soil Measurements

The dry-mass numerator and the bulk-volume denominator answer different questions. A density can use oven-dry soil mass while referring to the volume the specimen occupied when moist. “Dry” in the numerator does not automatically mean the denominator is the specimen’s later dry volume.

For this article, let Mₛ be a constant dry-solids mass. Define ρᵦ(V) = Mₛ/V for a stated bulk volume V that includes the relevant pore space. We exclude coarse fragments and assume no solid material is lost, dissolved away or added.

This notation makes the reference explicit. The same specimen can have one density based on its initial volume and another based on its smaller current volume.

The USDA NRCS soil-properties handbook distinguishes bulk-density reference states. The lesson is to read the method and volume definition alongside the number, rather than treating every dry-mass density as the same measurement.

Contents · Next chapter

Our fictional specimen contains 120 grams of dry solids throughout. In its initial moist state, it contains 30 grams of water and occupies 100 cubic centimetres. At an intermediate state, water mass is 12 grams and bulk volume is 90 cubic centimetres. At the final idealised dry state, water mass is zero and volume is 80 cubic centimetres.

We stipulate that all mass lost is water and that water density is 1 gram per cubic centimetre for this arithmetic. Actual measurement methods require their own controls and definitions.

The table preserves both changes. Losing water mass does not tell us, by itself, how much the specimen’s external volume changes. Here that second quantity is supplied independently.

Contents · Previous chapter · Next chapter

Using each state’s current volume gives dry-mass densities of 120/100 = 1.20, 120/90 ≈ 1.333 and 120/80 = 1.50 grams per cubic centimetre.

The dry endpoint is 25% higher than the initial value: (1.50 − 1.20)/1.20 = 0.25. The specimen’s volume decreased by 20%: (100 − 80)/100 = 0.20. These percentages differ because density is inversely related to volume at fixed mass.

If all three states are instead reported per initial reference volume of 100 cubic centimetres, the dry-mass density remains 1.20 throughout. That reference-based figure does not describe the current compactness of the smaller clod; it preserves a common denominator for accounting.

Neither calculation cancels the physical shrinkage. They describe different quantities. Label the denominator before comparing results. A rise from 1.20 to 1.50 using current volumes is a real change in that ratio, but it does not establish that a roller, load or management intervention caused it.

Contents · Previous chapter · Next chapter

Gravimetric water content w is water mass divided by dry soil mass. Initially w = 30/120 = 0.25 grams per gram. At the intermediate state, w = 12/120 = 0.10.

Current-volume volumetric water content θ is water volume divided by current bulk volume. Initially θ = 30/100 = 0.30 cubic centimetre per cubic centimetre. At the intermediate state θ = 12/90 ≈ 0.1333.

The conversion θ = wρᵦ/ρwater works when ρᵦ uses the matching volume and material basis. At the intermediate state, 0.10×1.333/1 ≈ 0.1333.

Using the initial density instead gives 0.10×1.20 = 0.12. That is water volume per initial reference volume, not per current specimen volume. It can be useful, but it needs that label.

The mistake is not multiplication. It is combining a current-volume interpretation with a reference-volume denominator. Keeping the states visible makes the apparent disagreement disappear.

Contents · Previous chapter · Next chapter

Assume, purely for our invented specimen, that solids have particle density 2.50 grams per cubic centimetre and that solid volume stays fixed. The solids occupy 120/2.50 = 48 cubic centimetres.

Subtract that solid volume from each current bulk volume to find pore volume. The three pore volumes are 52, 42 and 32 cubic centimetres. Subtract water volume to find air volume.

Across the whole change, thirty cubic centimetres of water leave while bulk volume falls by twenty. The difference is balanced by air volume increasing by ten.

Thus shrinkage need not equal water loss. The pore arrangement and air contribution matter. These numbers demonstrate one internally consistent account; they are not a universal shrinkage curve for clay-rich soils.

Contents · Previous chapter · Next chapter

Porosity is pore volume divided by the chosen bulk volume. For the three current-volume states, it is 52%, about 46.7% and 40%. In this model, the pore fraction falls as the specimen shrinks.

Degree of saturation is water volume divided by pore volume. Initially it is 30/52 ≈ 57.7%; at the intermediate state it is 12/42 ≈ 28.6%; at the idealised dry endpoint it is zero.

A water content of 0.30 by volume is therefore not the same as 30% saturation. The denominators are total bulk volume and pore volume respectively.

As an arithmetic check, θ should equal porosity times degree of saturation when all quantities use the same current volume. Initially 0.52×(30/52) = 0.30. At the intermediate state (42/90)×(12/42) = 12/90.

These checks identify inconsistent bookkeeping. They do not establish pore connectivity, permeability or the ease with which a plant can obtain the remaining water. Those properties require additional evidence.

Contents · Previous chapter · Next chapter

Imagine the dry clod occupies 80 cubic centimetres inside the original 100-cubic-centimetre envelope. A new twenty-cubic-centimetre gap surrounds it. If that gap belongs inside the layer-scale boundary, total volume can remain 100 even while the clod itself shrinks.

Density based on the clod is 120/80 = 1.50. Density based on the whole envelope remains 120/100 = 1.20 grams per cubic centimetre. The two measurements include different void space.

With our fixed 48-cubic-centimetre solid volume, whole-envelope pore volume is still 52. It now consists of 32 inside the clod and twenty outside it. Calling only one of those spaces “the pores” without a measurement boundary invites confusion.

A soil layer can change through surface lowering, lateral movement and crack formation together. The clod model does not establish how much of each occurs. South Dakota State University Extension’s bulk-density guide explains why sampling method and moisture state matter in shrink–swell soils.

Contents · Previous chapter · Next chapter

For a separate layer model, assume fixed horizontal area, no cracks within the counted envelope and the same proportional volume changes as the specimen. The initial thickness is ten centimetres; intermediate thickness is nine centimetres.

Initially θ = 0.30, so stored water depth is 0.30×100 = 30 millimetres. At the intermediate state θ = 12/90, and thickness is 90 millimetres. Storage is therefore (12/90)×90 = 12 millimetres.

Alternatively, use the initial reference basis: water per initial volume is 0.12, multiplied by initial thickness 100 millimetres. This also gives twelve millimetres.

Mixing current θ with initial thickness gives (12/90)×100 ≈ 13.33 millimetres, overstating storage. Mixing reference θ with current thickness gives 0.12×90 = 10.8 millimetres, understating it.

Both valid routes preserve the same water volume per horizontal area. The root-zone water-depth guide develops the conversion; this changed-thickness example adds the moving-volume condition.

Contents · Previous chapter · Next chapter

Suppose a different idealisation treats the specimen as shrinking equally in all three directions. A volume ratio of 80/100 = 0.80 gives a linear-dimension ratio of the cube root of 0.80, approximately 0.9283.

Each dimension decreases by about 7.17% relative to its initial length, not 20%. Multiplying 0.9283 three times reproduces approximately 0.80.

Changing the reference changes the percentage again. The initial length exceeds the final length by about (1/0.9283 − 1)×100 ≈ 7.72% of the final length. Neither percentage should be substituted for the other.

This is a geometric exercise. It does not establish isotropic shrinkage in actual soil or assign a laboratory shrink–swell classification. The fixed-horizontal-area layer in the previous chapter used a different constraint and therefore had a different thickness change.

Rewetting is another path. These drying data do not establish that a specimen will recover its earlier volume at the same water content. An investigation of swelling needs observations along that path as well.

Before taking a cube root or scaling a depth, identify which dimensions are allowed to change. A geometric shortcut is valid only for its stated shape and deformation assumptions.

Contents · Previous chapter · Next chapter

For a supplied-data investigation, record dry-solids mass, water mass, volume method, moisture or reference state, included cracks and the spatial scale represented. Keep the initial and current quantities in separate columns.

To investigate a real density change, also record material identity, depth, sampling disturbance, loading and wetting–drying history. A pair of numbers without compatible methods may not answer whether compaction occurred.

The soil compaction and moisture–density guide examines mechanical densification and laboratory reference curves. The calculation here does not replace those tests or their project-specific interpretation.

Use existing teacher-approved data or authorised laboratory reports for practice. This article supplies no drying, chemical-coating or excavation procedure. Its practical task is to make the denominators inspectable and identify which missing observation would distinguish alternative explanations. A precise calculation becomes more useful when the sample boundary is as clear as its units.

Contents · Previous chapter · Next chapter

Check one: another reference volume

A specimen has 150 grams of dry solids, initial volume 125 cubic centimetres and later volume 100 cubic centimetres. Initial-reference density is 1.20; later-current density is 1.50 grams per cubic centimetre. Volume falls 20%, while current dry-mass density rises 25%.

Check two: add remaining water

If that later state contains fifteen grams of water, gravimetric content is 15/150 = 0.10. Current-volume θ is 15/100 = 0.15; initial-reference water fraction is 15/125 = 0.12. Both fractions describe the same water with different volumes.

Check three: change the layer boundary

For a fixed-area layer initially 200 millimetres thick with current thickness 160 millimetres, current θ = 0.15 gives 24 millimetres stored water. Initial-reference fraction 0.12 times 200 also gives 24. Explain why using 0.15 times 200 would answer a mismatched calculation.

The check is successful when you can name the volume represented by every density and fraction, then show that equivalent accounting routes conserve the same water.

Contents · Previous chapter · Next chapter

Does higher dry-mass density prove mechanical compaction?

No. Check the reference volume, moisture state, material, method and history. Drying-related shrinkage can raise density based on current volume without added solid mass.

Is the initial volume always the right denominator?

No. A current-state description and a reference-based storage account have different jobs. State which quantity you need, then keep the basis consistent.

Can a clod result describe the whole cracked layer?

Only with an appropriate account of included and excluded voids and spatial representation. The same mass can have different clod-scale and envelope-scale densities.

Which route comes next?

Start with The Gold Standard of Soil Science and the soil-system science guide. For water accounting, use gravimetric moisture and root-zone water depth and infiltration, storage and drainage.

For other representation limits, continue with coarse fragments and carbon-stock denominators and composite samples and hidden spatial variation. For mechanical densification, read soil compaction and moisture–density curves. Geotechnical engineering develops professional ground questions beyond these paper models.

Connect the habit of preserving measurement meaning through The Gold Standard of Education and How X Works.

The primary definition source is USDA NRCS National Soil Survey Handbook, Part 618, especially its bulk-density reference states. SDSU Extension’s bulk-density explanation provides measurement context. The specimen states, pore accounts, geometry and layer examples are original teaching constructions, not characteristic values or site recommendations.

Contents · Previous chapter · Return to Soil Science