How do you read an unsaturated soil hydraulic conductivity curve? First identify whether the graph plots hydraulic conductivity against volumetric water content, matric pressure head or suction. A K(θ) curve uses water content as its reference, while a K(h) curve uses pressure head; both describe how readily water can move through unsaturated pore networks under an imposed hydraulic gradient. This guide explains their axes, logarithmic scales, relative conductivity, model fits and uncertainty.
Readers searching for unsaturated hydraulic conductivity versus water content, soil K(h) curves, van Genuchten–Mualem conductivity or how to interpret a log hydraulic conductivity graph need one distinction immediately: water stored in soil is not the same as water transmitted through it. Conductivity often declines sharply as soil dries, even while appreciable water remains. A retention curve alone does not measure conductivity, and one time-dependent infiltration test cannot establish the full conductivity function.
Read every K(θ) or K(h) plot with its measurement or modelling method attached. Units, sign conventions, wetting history, temperature, pore connectivity and the observed moisture range determine what the line can legitimately mean. A successful explanation translates the curve into conditional transport behaviour without pretending that every fitted value has been directly measured.
What hydraulic conductivity actually means
Hydraulic conductivity describes the proportional relationship between water flux and hydraulic-head gradient for a given porous material and hydraulic state, within the assumptions of the governing flow law. For unsaturated soil, conductivity changes as pores empty or fill; it is not generally a single constant for the whole drying cycle. A steep gradient can drive appreciable water movement even when conductivity is low, while a small gradient can produce little movement in a material with greater conductivity.
This distinction matters for language. Conductivity is a capacity to transmit water under a specified state, not the amount of water present and not the amount already transported. Reports sometimes call high conductivity fast soil, but the actual flux also depends on gradient, direction and flow geometry. Translate K as a transport parameter and q as an actual flux when the scientific context requires it.
Why unsaturated conditions change the pathway
Under saturation, many water-filled pores participate in connected flow. As the soil becomes drier, larger pores tend to drain and smaller liquid-filled connections carry much of the remaining flow. Changes in water-film thickness, pore connectivity and the distribution of filled pores can reduce conductivity by orders of magnitude. The exact relation depends on structure, texture, wetting path and other properties.
The everyday intuition that a damp soil can still conduct water is correct but incomplete. A soil may contain a sizeable volume of tightly retained water while transmitting it very slowly. A fine-textured material can retain water at a suction where a coarse material is nearly dry, but that does not automatically mean it offers the greater conductivity under all conditions. Storage and mobility must be compared separately.
K(θ) means conductivity as a function of water content
In K(θ), the independent variable θ denotes volumetric water content. A plot may display conductivity in metres per second or centimetres per day against θ expressed as a fraction or percent by volume. The curve commonly rises steeply as water content increases because more pore connections become active. On a logarithmic conductivity axis, an apparently smooth increase may represent enormous multiplicative changes in transport capacity.
Check whether θ is total measured volumetric content, effective saturation or a normalised model coordinate. Effective saturation is often calculated using residual and saturated water-content parameters, which may be fitted rather than directly observable endpoints. Values from different models cannot always be compared merely because the axis labels appear similar. Explicit definitions prevent the conversion of model coordinates into misleading physical claims.
K(h) means conductivity as a function of pressure head
A K(h) plot relates conductivity to a pressure-head or matric-potential coordinate. In unsaturated soil, pressure head h is commonly negative relative to atmospheric pressure, while some figures present positive suction magnitude instead. Those conventions can reverse the apparent left-to-right direction of drying. When suction magnitude grows, conductivity commonly declines; when a negative pressure head approaches zero from below, conductivity commonly increases.
The reader should never conclude that a curve rises during drying simply from its visual slope. First find out whether the horizontal variable is signed head, absolute suction, pressure in kilopascals or the logarithm of suction. A caption that states conductivity fell as suction magnitude increased is more robust than saying the line bent leftward or rose toward the axis.
Logarithmic vertical axes change what spacing means
Conductivity frequently spans multiple orders of magnitude, making logarithmic vertical axes useful. Equal vertical distances can then represent tenfold changes rather than equal additive increments. Moving from 10^-8 to 10^-6 metres per second is a hundredfold rise, even though the written exponents differ by only two. A graph of log10 K may use negative values, but that does not mean physical conductivity is negative.
A reader should check whether the plotted variable is K on a logarithmic scale or the numerical quantity log K on a linear scale. Those displays can look similar, yet axis labelling and units differ. Never subtract plotted logarithmic values and call the result a conductivity difference in metres per second. Convert back to the physical scale if an absolute comparison is needed.
Relative conductivity is a ratio, not another flow law
Relative hydraulic conductivity Kr is commonly defined as K divided by a reference saturated conductivity Ks under the same conceptual model. It is dimensionless when the numerator and denominator share units. A value of 0.01 means conductivity is one hundredth of the reference conductivity, not that one percent of pore volume contains water. Effective saturation, Kr and volumetric water content therefore answer three different questions.
Modelled Kr values generally fall between zero and one for the intended convention, but empirical measurements and reference choices may complicate direct comparison. Always ask what Ks represents: a measured saturated value, a fitted parameter, a laboratory estimate or a field-scale effective property. Calling every relative curve a percentage of saturation erases the transport information it was designed to represent.
The Darcy–Buckingham flow relationship
For one-dimensional vertical flow, a useful conceptual form is q = -K(h) dH/dz, where H is total hydraulic head and z is the spatial coordinate according to the chosen convention. The negative sign denotes flow from higher to lower total hydraulic head. The expression introduces the core reason a conductivity curve cannot by itself predict flux: the hydraulic gradient must also be known. In more complex conditions, gradients and conductivity may vary through the profile.
Students should distinguish pressure head, elevation head and total head. Water can move upward under strong capillary gradients even though gravity acts downward. A translation that equates negative matric potential with upward flow is incomplete; direction is determined by the total head gradient. Always retain the coordinate convention when interpreting a sign.
The saturated endpoint and its limitations
A fitted curve often approaches a saturated conductivity Ks as water content reaches a specified saturated value or pressure head approaches the saturated boundary. That endpoint can anchor a model, but field saturation is not a single uncomplicated state. Entrapped air, macropore continuity, sample orientation and flow geometry can influence the measured value. Laboratory and field estimates may differ substantially.
The term saturated should be checked against the method. Did the experiment truly establish full saturation, or did it assume an endpoint while some air remained? Was the saturated parameter fixed independently or estimated simultaneously with curve-shape parameters? A numerical Ks used in a model may be an effective calibration parameter, not a direct observation. These distinctions become crucial when comparing transport predictions across studies.
Retention curve versus conductivity curve
A soil water retention curve maps water content to suction or matric potential. A conductivity curve describes transport capacity at a hydraulic state. One maps a state relationship; the other maps a flow property. Because pore-size information affects both, retention measurements can inform conductivity models, but they are not a direct measurement of water flux. Converting retention into conductivity requires structural assumptions and often a separate scale parameter.
Two soils can show similar retention curves while differing in connected flow channels. Likewise, a soil can store considerable water at strong suction while transmitting it slowly enough to limit root water supply. This is why readers should follow the companion soil water retention guide without treating that plot as a substitute for K(θ).
Linking K(θ) and K(h) without treating them as identical
If a retention relationship θ(h) is known, one can use it to move conceptually between a K versus θ description and a K versus h description for the same material, branch and conditions. This is a mapping through the retention function rather than a change in physical units. Because retention can be hysteretic, the same θ need not map to the same h on wetting and drying paths.
The transformation may also magnify uncertainties at flat parts of the retention relation. A small uncertainty in θ can produce a large uncertainty in inferred h, which then affects the conductivity interpretation. The existence of two plotted functions does not prove that one is measured and the other derived. Check the original method and state whether each panel is observational, predicted or computed from shared parameters.
Why modelled curves often look smoother than data
A model imposes a mathematical form on scattered measurements and can produce a smooth function across a wide range, including states never directly observed. That smoothness is a property of the fitted equation, not evidence that natural soil contains no abrupt changes or preferential pathways. Readers should ask where actual data points appear and where the line is extrapolated beyond them.
A useful scientific graph shows fitted values against observations and ideally uncertainty bands or residuals. Without those, a sophisticated-looking curve may conceal weak identification of its dry-end behaviour. In educational interpretation, distinguish an observed response, a fitted interpolation and an extrapolated prediction. The same numerical value carries different confidence depending on which category produced it.
The van Genuchten retention expression
The van Genuchten retention family commonly expresses effective saturation as a function of pressure head using parameters that control scale and curvature. An often-used form is Se = [1 + (α|h|)^n]^-m for unsaturated conditions, with model-specific choices such as m = 1 – 1/n. The equation maps suction to effective saturation; it is not itself a direct law for conductivity.
The fitting parameters must be interpreted in the units and conventions used for h. The scale parameter α is not a universal fixed pore radius, and n is a shape descriptor rather than a particle-size measurement. If authors choose a different constraint or an alternative fitting function, coefficients may no longer be directly interchangeable. Good translation respects the definition attached to the particular equation.
The Mualem–van Genuchten conductivity relation
Combining a retention model with a pore-connectivity framework such as Mualem’s yields a predictive function for relative conductivity. A familiar implementation is Kr(Se) = Se^l [1 – (1 – Se^(1/m))^m]^2, under the stated parameterisation. The exponent l is a pore-connectivity or tortuosity-related model parameter. It is often assigned a conventional value, although that choice is not a universally measured constant.
Because the formula can be highly sensitive near dry and wet endpoints, excellent retention fits do not guarantee excellent conductivity predictions. Comparing independent flow measurements is stronger evidence than reporting a high goodness-of-fit for retention alone. The model is valuable precisely when direct conductivity data are scarce, but readers should see the inferential step rather than mistake an equation output for measured transport.
What the parameters can and cannot prove
Parameters such as α, n, m, l, θr, θs and Ks arise within a chosen modelling framework. Some can be measured approximately, some are fitted, and some are constrained for convenience. They may compensate for each other when data cover a narrow moisture range. Two distinct parameter sets can produce similar plotted curves over the observations yet diverge strongly when extrapolated to drier states.
A responsible report states the fitting interval, objective function and which parameters were fixed. It avoids turning a single coefficient into a sweeping physical explanation. For example, a larger fitted α should not automatically be presented as direct proof of one unique pore-size distribution without considering structure and the equation’s definition. Model parameters are concise descriptions of an assumed representation, not independent witnesses to every microscopic mechanism.
Hysteresis and the wetting–drying pathway
Soil water retention can differ between wetting and drying because water fills and empties pore networks through path-dependent processes. Conductivity can also show path dependence under comparable states. The presence of hysteresis means one universal curve may fail to capture a soil undergoing alternating rainfall and drying. Whether the difference is important depends on the moisture range, material and prediction being made.
When two datasets disagree, check whether one came from a drying experiment and the other from rewetting. Do not attribute all discrepancies to instrument failure before comparing the histories. A fitted dry-path curve used for rapid wetting may be a convenient approximation, but the assumption should be acknowledged. Words such as reversible and unique require evidence, not just a smooth loop drawn in a software package.
Coarse and fine textures on common scales
A coarse-textured material can lose water-filled connectivity rapidly over a relatively narrow suction range, while a fine-textured material may retain liquid connectivity under stronger suction. At saturation, however, the coarse material may transmit more rapidly, depending on structure and packing. This makes simple rankings such as sand always conducts faster potentially wrong across all moisture states.
Compare materials at a shared, physically meaningful independent variable and state which one was held constant. Comparing at equal θ is not the same as comparing at equal h because the retention relationships differ. A graph can show crossing conductivity functions. The scientifically accurate interpretation says which material conducted more at the specified state and method, rather than assigning an unconditional winner.
Pore connectivity, tortuosity and anisotropy
Hydraulic transport depends not only on pore volume but also on how pores connect across the sample. Tortuous flow paths increase resistance, while continuous macropores may provide rapid routes under certain wetting conditions. Some soils exhibit anisotropy: vertical and horizontal conductivities differ because layering, root channels, cracks or compaction create directional structure. A single scalar curve may represent one measurement direction or an effective isotropic approximation.
Ask whether the reported conductivity corresponds to vertical columns, horizontal cores, field-scale infiltration or model calibration. Applying a vertical laboratory result to lateral hillslope flow without qualification can produce misleading conclusions. Anisotropy matters particularly in structured and layered profiles where water can redistribute sideways above restrictive horizons. Graph interpretation benefits from a statement about spatial direction even when the figure itself omits it.
Temperature, salinity and fluid properties
Hydraulic conductivity includes the properties of the flowing fluid as well as the pore medium. Water viscosity changes with temperature, and dissolved salts can influence clay swelling or dispersion in susceptible soils. A reported K(θ) function may therefore depend on water chemistry and measurement temperature. Not every experiment measures these effects, but ignoring them can make comparisons appear more precise than they are.
If a study corrects conductivity to a reference temperature, read the correction method and distinguish hydraulic conductivity from intrinsic permeability. Permeability is a medium property in a framework that separates fluid properties, whereas hydraulic conductivity reflects both. Translating them as exact synonyms can create quantitative errors. A careful report retains the fluid condition whenever transport coefficients are compared.
The evaporation laboratory method
Laboratory evaporation methods can estimate unsaturated hydraulic properties by monitoring changes in sample mass and pressure-head conditions as water evaporates from a controlled specimen. With appropriate geometry and analysis, these observations inform retention and conductivity across part of the drying range. The method does not automatically cover extremely dry conditions or the near-saturated endpoint with equal accuracy. Sensor range and sample preparation are central.
A graph produced from an evaporation experiment should identify the inversion or calculation procedure used to turn observable mass and head changes into K values. Those values are inferred through a model and assumptions about flow, not read directly from a conductivity meter. If measurements cease when tensiometers lose hydraulic contact or cavitate, the plotted dry tail may be extrapolated rather than empirically supported.
Tensiometers and head measurements
Tensiometers measure a pressure difference related to soil matric potential while maintaining hydraulic contact through a porous cup, within their practical operating range. When used in pairs at different depths, they can help estimate hydraulic gradients. But pressure head alone does not give water flux; conductivity at the relevant state and correct geometry is also required. Instrument installation, equilibration and contact affect the reliability of the signal.
A reader should distinguish sensor limitations from soil physics when a trace suddenly levels or drops out. Conventional tensiometers face limits as tension increases and the water column becomes unstable. A missing dry-end record does not indicate the soil stopped transmitting water. It may only mean that the instrument no longer measured the state. Method transparency is necessary for interpreting the resulting K(h) curve.
Steady-state measurements and boundary control
Steady-state approaches estimate conductivity after a system reaches approximately constant conditions under a controlled hydraulic gradient. Such tests can be valuable for specified moisture states, but establishing uniform conditions may be difficult in unsaturated soil. Evaporation, redistribution and boundary contact can cause gradients to change within a specimen even while a surface flux appears stable.
Check where the water-content and head values were measured relative to the flow path. A conductivity number inferred from one gradient may represent a spatial average rather than a local value at a precise θ. A plot assembling several steady states is also different from one continuous drying transient. Both can be informative; they simply answer their questions through different experimental approximations.
Transient inverse estimation and identifiability
Inverse modelling estimates hydraulic parameters by finding values that reproduce observations such as moisture profiles, pressure heads or outflow time series. This technique can be powerful because direct K measurements are difficult. Yet multiple parameter sets may reproduce the same observations, especially when the experiment does not excite a broad range of states. Parameter identifiability is therefore an evidence question, not merely an optimisation software output.
A good study reports the observations fitted, initial conditions, boundary assumptions, measurement errors and uncertainty. It may compare several models or use independent validation. If a graph contains only a simulated K curve without uncertainty, do not claim that every point has been verified. The underlying data may constrain a few integrated responses while leaving dry-end conductivity largely uncertain.
What a residual plot can reveal
A residual is the difference between an observation and the corresponding model estimate, under a stated sign convention. Residuals plotted against moisture, head or time can reveal systematic failure even when an overall fit metric looks impressive. For example, persistent underestimation near saturation and overestimation at moderate suction could indicate an inadequate curve family or a boundary-condition mismatch.
Look for clusters, trends and changes in spread. Random-looking scatter near the measurement precision provides different evidence from an orderly sequence of positive and negative errors. Also examine whether the fitting objective weighted logarithmic conductivity, linear conductivity or observations by uncertainty. Optimising squared errors in one representation can prioritise different parts of a multi-order-of-magnitude curve.
Worked example: reading orders of magnitude
Suppose a hypothetical K(θ) graph reports K = 10^-5 metres per second at θ = 0.40 and K = 10^-8 metres per second at θ = 0.22 for the same specimen and protocol. Conductivity decreased by a factor of one thousand across those states. It is incorrect to say that the water content decreased by a factor of one thousand: the reported θ values differ by 0.18 cubic metres per cubic metre.
This comparison also does not establish that observed field infiltration would differ by precisely the same factor, because surface entry depends on transient gradients and boundary conditions. The correct conclusion is narrower and more useful: under the represented model or measurements, the material’s conductivity is highly sensitive to drying. Keep the numerical ratio attached to K rather than transferring it to every other water variable.
Worked example: negative head versus positive suction
Imagine one source plots h from -100 to -1 centimetres and another plots suction magnitude from 1 to 100 centimetres. They may describe the same physical progression but display it in opposite horizontal directions. At h near -1 centimetre the soil is relatively close to saturation; at h near -100 centimetres it is under stronger suction, subject to the exact reference convention.
If you describe both as conductivity rising with increasing head, the phrase can become ambiguous because one head coordinate is signed and the other is a positive magnitude. Instead write conductivity increased as matric suction magnitude decreased, or conductivity rose as signed pressure head approached zero. This wording survives translation into other languages and protects the physical meaning when graph orientations differ.
Hydraulic conductivity and root uptake
Roots extract water locally, generating gradients and altering the water-content distribution around them. As the surrounding soil dries, declining conductivity can restrict replenishment toward the root surface even when the bulk soil still contains water. This helps explain why total stored water and immediate plant access are distinct. Root hydraulic properties and atmospheric demand also influence uptake, so the soil conductivity curve cannot alone determine transpiration.
For plant-side measurements, the leaf water potential guide and stomatal conductance guide provide complementary evidence. Their variables should not be confused with soil hydraulic conductivity. The same word conductance can appear in plant physiology and soil physics while referring to different parts of the transport pathway.
Conductivity and contaminant transport
Water movement influences the transport of dissolved substances, but solute spreading also depends on advection, dispersion, diffusion, sorption, transformations and preferential pathways. A high conductivity under one state does not prove a contaminant will move rapidly across a landscape; gradients and connectivity matter. Conversely, a low matrix conductivity may coexist with fast bypass flow through rare connected channels.
When translating an environmental report, distinguish the conductivity function from an observed breakthrough curve. The former describes a hydraulic property; the latter records the appearance of a tracer or substance after transport through a particular system. Connecting them requires a transport model and boundary data. This distinction protects against overstated conclusions about groundwater vulnerability based only on one soil property.
Comparing laboratory cores with field measurements
Laboratory cores allow controlled boundaries and detailed monitoring, but sampling can disturb cracks, roots and large pores. A small specimen may omit rare high-conductivity pathways that dominate field response. Field inverse estimates integrate a larger and less controlled volume. Disagreement is therefore not automatically a failure of either method; they may represent different spatial scales and structural features.
State the sample size, orientation, disturbance status and measurement technique. If the field is layered or heterogeneous, a single core-derived curve should not be silently applied to an entire catchment. Sensitivity analysis can reveal whether predictions are strongly affected by conductivity uncertainty. The purpose of a K curve is to improve understanding of water movement, not to conceal scale mismatch behind a neat mathematical line.
What happens near the dry limit
Near very low water contents, liquid water becomes less connected and film flow, vapour transport or other mechanisms may become comparatively important. Common single-domain liquid-flow models may perform poorly without additional processes. The dry tail of a model curve can fall below direct measurement capability, creating dramatic-looking log-scale values that are predictions rather than observations.
Readers should check whether the reported range is physically measured and relevant to the process of interest. An extrapolated value of 10^-15 metres per second may be mathematically precise within the model but not experimentally verified to that precision. Qualifiers such as model-predicted conductivity at high suction communicate more than an unqualified number. Likewise, the absence of measured dry-end data should be reported rather than filled with invented certainty.
What happens near saturation
Near saturation, large connected pores can dominate flow, and small changes in air occupancy may create sharp conductivity changes. Soil shrinkage or swelling can also alter pore geometry as wetness changes. A smooth fitted curve may fail to capture rapid structural transitions, particularly in cracked or highly aggregated soils. Some samples show a marked difference between saturated and near-saturated measurement states.
Interpret the last few wet-end points with their actual uncertainties. If the instrument never achieved complete saturation, do not describe the curve’s upper endpoint as a directly measured Ks. A fitted saturation boundary should be flagged as such. This is a recurring lesson in technical translation: the label chosen by software does not replace the measurement protocol described in the methods.
How to read uncertainty bands
An uncertainty band around a K curve may represent parameter uncertainty, statistical confidence, predictive variation or a combination of measurement and model errors. Those concepts should not be translated as interchangeable. A narrow confidence band on fitted mean conductivity does not imply that every local field point will fall inside it. Conversely, a wide prediction interval reflects additional sources of variability.
Find the stated probability level and construction method, especially on logarithmic plots. Intervals symmetric in log space are generally asymmetric when expressed as absolute conductivity values. A graph without an interval should not be assumed precise; ask whether replicate measurements, error propagation or model alternatives were considered. Reporting uncertainty helps learners interpret numerical transport claims with appropriate restraint.
A practical graph-reading sequence
First name the axes and units. Second identify whether the horizontal coordinate is θ, signed h, suction magnitude or effective saturation. Third establish whether conductivity is absolute or relative and whether a log scale is used. Fourth locate observed points, fitted line and extrapolated range. Fifth check drying or wetting history, method, temperature and the reference Ks. Sixth interpret the trend without equating stored water with transport.
A concise example reads: Across the sampled drying range, inferred unsaturated hydraulic conductivity fell by approximately three orders of magnitude while volumetric water content declined, consistent with progressive loss of connected water-filled pathways. The result applies to this specimen and fitted method; independent field flow observations would be needed to support site-scale predictions. This explanation is precise without claiming more than the data show.
Common errors when reading K curves
Do not treat negative logarithms as negative conductivities. Do not equate relative conductivity with percent water saturation. Do not confuse K(h) with retention θ(h), even though they share the variable h. Do not assume every line is directly measured. Do not extrapolate far into unobserved dryness without warning. Do not compare a signed head graph with a positive suction graph before checking direction.
A subtler error is to claim that conductivity equals water flux. The missing hydraulic gradient can reverse the conclusion, especially in vertically stratified soil. Another is to translate hydraulic conductivity as intrinsic permeability without including fluid properties. Finally, never claim that a high saturated endpoint guarantees strong unsaturated flow under drought. The state dependence is the central information in the curve.
Learning exercise: compare two moisture states
Give students a hypothetical graph showing K = 10^-6 metres per second at θ = 0.32 and K = 10^-9 metres per second at θ = 0.18. Ask for the conductivity ratio, the absolute change in θ and the meaning of a logarithmic vertical axis. A strong answer says conductivity is one thousand times lower at the drier state within the represented curve, while θ fell by 0.14 cubic metres per cubic metre.
Now add a second plot of retention versus suction. Ask whether the water-content change directly measures conductivity. The answer is no: the first graph depicts transport capacity and the second describes state. The paired exercise builds the habit of translating graph nouns before performing arithmetic, a transferable skill across scientific subjects.
Learning exercise: identify an unsupported claim
Present a fitted K(h) line extending beyond the driest measured point and ask whether a report may state that conductivity was measured throughout the entire line. Learners should separate observed points, model-based interpolation and extrapolation. Next show a laboratory core curve applied to a heterogeneous field and ask which assumptions need review. The obvious answer includes scale, macropores, direction and boundary differences.
Good scientific reading involves more than spotting the trend. It asks what the procedure established and what alternative interpretations remain plausible. When learners annotate a plot with observed, inferred and uncertain regions, they gain a practical way to evaluate technical claims even when the underlying mathematical model is unfamiliar.
Frequently asked: does dry soil always mean zero conductivity?
No. Unsaturated hydraulic conductivity commonly becomes much smaller as soil dries, but zero is an idealised limiting description rather than a universal experimentally resolved value at ordinary low water contents. Connected films and other transport modes may persist, and the relevant range depends on the physical model. A log plot cannot display zero, so model curves may approach tiny positive numbers.
If a graph ends before the very dry region, it has not shown that flow stopped. It shows only what was represented in the data range. Use language such as very low conductivity rather than exact zero unless the context provides a legitimate physical and numerical definition. Keeping the distinction prevents overinterpretation of model limits.
Frequently asked: which model is best?
No single conductivity model is best for every soil, flow regime or observational range. van Genuchten–Mualem functions are widely used because they connect retention and conductivity conveniently, while alternative models may better capture particular pore distributions, structural domains or dry-end behaviour. Model choice depends on independent evidence, predictive purpose, complexity and parameter identifiability.
The fairest comparison uses held-out observations or a documented validation strategy, not merely the smallest fitting error on the data used to tune the equation. A simple model that performs reliably for its intended question can be more valuable than a complex curve with poorly constrained parameters. Explicitly state the limits of the chosen representation.
Frequently asked: why do studies disagree?
Differences can arise from sampling disturbance, pore orientation, wetting and drying history, fluid chemistry, pressure range, measurement method, model constraints and optimisation objectives. Even the same soil at different bulk density or temperature may yield different reported conductivity. The first comparison should therefore reconstruct the experimental context rather than assume one result is wrong.
Check whether one study measured conductivity directly under controlled flow while another derived it from retention using a parameterised model. Those are different evidence routes. Both may be useful when accurately labelled. Agreement over a limited moisture range does not establish agreement at the very wet or very dry endpoints, where extrapolation is often most consequential.
Sources and connected learning routes
The USGS review on measuring and modelling unsaturated hydraulic conductivity discusses the practical difficulty of obtaining K–water-content relationships. The foundational van Genuchten 1980 paper describes the linked retention and conductivity approach; readers should consult the original equation definitions.
For the adjacent question of how water first enters a soil surface, read the infiltration rate and cumulative intake guide. For why these properties matter to living systems, connect them to the dendrometer water-deficit guide and plant water potential. Accurate interpretation builds a chain from terminology to measurement, from measurement to model and from model to a carefully bounded conclusion.