A carefully weighed soil core is still a different system from the soil around it
A small soil core loses 18 grams during a sunny interval. The balance may be precise, but what does the loss represent? The core has been separated from deeper soil and nearby roots. Its water supply and temperature can begin to differ from the field. A reliable evaporation estimate needs the mass calculation and an explanation of those changed boundaries.
Microlysimeters use small soil systems to investigate evaporation. This guide concentrates on the isolated-core method: how a measured mass change becomes an equivalent water depth, why the core’s age matters, and how several local observations can support a wider estimate. It extends the existing soil-drying guide without replacing its explanation of evaporation stages.
The numerical records are original fictional exercises. Use supplied data rather than extracting soil or operating equipment without an appropriate study plan. The objective is to recognise which assumptions make the observation informative, and which additional evidence would test transfer to surrounding soil.
Hydrology subject guide · Learning that transfers into life · How X Works hub
Choose your reading route
Route 1: Read mass and boundary · Chapters 1–3
Route 2: Explain disturbance · Chapters 4–5
Route 3: Follow age and events · Chapters 6–8
Route 4: Test wider inference · Chapters 9–10
Route 5: Transfer and report · Chapters 11–12
Full chapter index · Independent worked case · Questions and practical checks
Chapter index
Read mass and boundary · Chapters 1–3
Define an isolated core rather than an entire field
Convert mass loss to water depth with the exposed area
Reconstruct evaporation from the full interval balance
Explain disturbance · Chapters 4–5
Explain hydraulic disconnection as a changed supply pathway
Keep root exclusion distinct from soil evaporation
Follow age and events · Chapters 6–8
Treat the wall and base as thermal boundaries too
Use exposure age as a variable, not a footnote
Handle rain and replacement without hiding a new initial state
Test wider inference · Chapters 9–10
Weight local evaporation by the surface it represents
Compare independent methods at compatible boundaries
Transfer and report · Chapters 11–12
Transfer from a bare patch to a cropped or drying surface
Report measured loss and field inference separately
CHAPTER 1 OF 12
1. Define an isolated core rather than an entire field
In the isolated-core approach, a bounded soil sample is weighed over an exposure interval. Its surface experiences the local environment, while its sides and base restrict exchanges with the surrounding profile. The measurement is most directly a change in the mass of that bounded system. Calling it field evaporation adds a representativeness claim that must be checked.
The original method developed by Boast and Robertson explicitly investigated how limited core length and exposure time could cause departures from a deeper soil system. That question remains central: the method is useful because its small boundary makes weighing practical, but the same boundary can alter the process being inferred.
Draw a core with an exposed upper surface, walls and a base. Mark rain, irrigation, evaporation, possible leakage and handling. Then draw the surrounding field with deeper water movement and roots. The drawings should differ. If they look identical, the explanation has hidden the intervention rather than examined it.
A small instrument is not automatically a poor instrument. It can answer a carefully defined local question very well. The important limit is that precision within the core does not establish equivalence to every surface outside it. Keep measured mass, inferred core evaporation and spatially expanded field evaporation as separate stages.
CHAPTER 2 OF 12
2. Convert mass loss to water depth with the exposed area
For a water mass change, equivalent depth equals water volume divided by exposed soil area. Taking water density as 1 gram per millilitre for the teaching calculation, a 20 gram water loss is 0.020 litre. Over a 0.01 m² surface, that represents 2 mm, because one litre per square metre equals one millimetre.
Use the actual exposed soil area, not the exterior footprint of a thick wall or the base of a weighing tray. If an exercise supplies a circular internal diameter of 0.10 m, its area is π × 0.05², approximately 0.007854 m². A 15.7 gram water loss then corresponds to approximately 2.0 mm.
The same balance error matters more over a smaller area. A 1 gram error corresponds to 0.10 mm over 0.01 m² but 0.20 mm over 0.005 m². This is a dimensional sensitivity, not a universal recommendation for core size. Increasing size can change other aspects of disturbance and practicality.
If the balance reports total mass, first establish which part of the change is water. Soil crumbs removed during handling are not evaporated water. A correct density conversion cannot repair a misclassified mass loss. The physical identity of the lost material belongs before division by area.
CHAPTER 3 OF 12
3. Reconstruct evaporation from the full interval balance
Let ΔM be ending water mass minus starting water mass. Let P be caught precipitation, I irrigation retained in the core, E evaporation and D drainage or leakage, all as water masses over the same period. With no other routes, ΔM = P + I − E − D. Therefore E = P + I − D − ΔM.
A fictional core gains 8 grams while receiving 30 grams rain and no irrigation. With no drainage, evaporation is 30 − 8 = 22 grams. The rising mass does not imply that evaporation stopped. It means incoming water exceeded the outgoing amount during the interval.
Now include a separately measured 5 gram drainage loss. Evaporation becomes 30 − 5 − 8 = 17 grams. Treating every loss from storage as atmospheric evaporation would overestimate it. A sealed-base design may exclude drainage, but that exclusion must describe the actual apparatus rather than be assumed from the word microlysimeter.
Keep handling changes separate. If an observer removes a 2 gram soil fragment, a water-only equation no longer explains total measured mass without an extra term. The useful response is to document or qualify the affected observation. Adding enough imaginary rain to close the account would make the equation balance while misrepresenting the evidence.
CHAPTER 4 OF 12
4. Explain hydraulic disconnection as a changed supply pathway
Surrounding soil can exchange water with deeper layers and adjacent regions. An isolated core restricts those pathways. As evaporation proceeds, its remaining water can diverge from the moisture distribution in the intact profile, even when both surfaces began similarly.
For a fictional comparison, two surfaces each lose 2 mm in the first day. The intact profile then receives a supplied 1 mm upward contribution from below, while the isolated core receives none. By the next day, the stores available to support evaporation differ. Equal initial conditions did not guarantee equal later conditions.
The direction of the bias is not fixed by the word isolated. Removing a deeper supply can dry the core sooner; preventing drainage can leave it wetter after rain. Which change matters depends on conditions, design and interval. State the omitted pathway and trace its consequence instead of announcing that every core must overestimate or underestimate.
An independent profile observation can help test this divergence. Compare compatible water-content patterns or other evidence inside and outside the core over time. A single initial moisture match is not a validation of a week-long exposure. The relevant test follows the changing system through the period used for inference.
CHAPTER 5 OF 12
5. Keep root exclusion distinct from soil evaporation
A core designed to exclude active plant uptake helps focus on soil-surface evaporation, but the surrounding cropped soil may continue losing water to roots. That can change its moisture relative to the root-excluded core. Isolating one flux is useful, yet it can alter the state that controls that flux.
Suppose an invented field soil loses 1 mm to soil evaporation and 2 mm to root uptake during a day. A root-free core with otherwise identical initial water loses only the 1 mm atmospheric component. Its remaining storage is now 2 mm larger than the surrounding comparison volume, before other pathways are considered.
On a later day, evaporation may respond to the changed moisture conditions. The core is not necessarily a continuing control for the field merely because excluding roots helped identify the first day’s flux. Root exclusion and representativeness must be evaluated together.
Do not label a microlysimeter loss as total crop evapotranspiration. That wider quantity includes transpiration under the defined crop boundary. The USDA discussion of partitioning uses microlysimeters as one part of an evidence chain, not a replacement for measuring or estimating every plant-water pathway.
CHAPTER 6 OF 12
6. Treat the wall and base as thermal boundaries too
The container changes more than water flow. Wall and base materials can alter heat conduction and storage around a small core. Temperature differences can affect evaporation and the apparent relationship between the core and surrounding soil. A water-tight base is not necessarily a thermally neutral base.
USDA researchers investigated wall materials and capping effects and found design-dependent temperature and evaporation differences under their study conditions. The lesson is to examine the thermal boundary, not to copy one experimental design or lifetime as a universal field specification.
In a fictional sensitivity calculation, an extra supplied 0.245 MJ/m² of energy entirely used for evaporation corresponds to 0.10 mm water at a rounded latent heat of 2.45 MJ/kg. This does not diagnose a container’s thermal error. It shows that a modest energy perturbation can matter when the daily water signal itself is small.
A comparison should therefore include thermal context as well as scale precision. If two materials produce different losses, consider their temperatures, moisture and exposure before attributing the difference to soil treatment. A scientifically useful control changes as few relevant boundaries as possible and records the ones that cannot be held constant.
CHAPTER 7 OF 12
7. Use exposure age as a variable, not a footnote
Exposure age means how long the bounded core has evolved since extraction or the relevant reset. It is different from the duration of the latest weighing interval. A one-hour observation made on a week-old core is not equivalent to a one-hour observation on a freshly representative core.
Consider fictional daily core losses of 3, 2, 1 and 0.5 mm. The intact comparison loses 3, 2.5, 2 and 1.5 mm. Their first-day agreement is good, but cumulative losses are 6.5 and 9 mm over four days. The growing difference suggests that an initially acceptable comparison did not remain equivalent.
Do not infer a universal four-day failure threshold from these invented values. Published investigations show that useful exposure periods depend on soil, depth, initial wetness and environmental conditions. A replacement schedule should follow evidence for the design and study, not the calendar interval that happened to work elsewhere.
Preserve core identity and extraction time in the dataset. Replacing a core resets its history and can change local initial conditions. A continuous graph assembled from successive cores should show that it combines samples. Numerical continuity does not mean one undisturbed soil column supplied every point.
CHAPTER 8 OF 12
8. Handle rain and replacement without hiding a new initial state
Rain can affect intact soil and a bounded core differently because drainage and lower-boundary exchange differ. After a wetting event, reassess the storage account and representativeness. Reusing a pre-rain interpretation without checking the new state can create an apparently continuous but physically inconsistent series.
Imagine a fictional core receives 8 mm rain, evaporates 1 mm and cannot drain. Its store increases by 7 mm. The surrounding comparison receives the same rain and loses the same evaporation but drains 4 mm deeper, increasing storage by only 3 mm. The surfaces have experienced the same weather but not the same water account.
Replacing the core may help restore a suitable initial condition when the study design supports it, but it introduces a new sample and weighing baseline. Do not treat the mass difference between two different cores as one interval’s evaporation. Their dry-soil masses and initial water stores need not match.
For an audit, record which intervals share a core, which include wetting and which follow replacement. This simple classification makes it possible to test whether error grows with age or changes after events. Without it, instrument evolution may be mistaken for an environmental drying-stage transition.
CHAPTER 9 OF 12
9. Weight local evaporation by the surface it represents
Several cores can reveal local variation, especially beneath incomplete vegetation cover. But a simple average answers an equal-sample question. A field estimate needs weights representing the relevant areas and a sampling design that supports the class means.
Suppose an invented field is 30% exposed inter-row soil and 70% shaded soil. Representative class evaporation estimates are 3 and 1 mm. The field soil-evaporation estimate is 0.3 × 3 + 0.7 × 1 = 1.6 mm. Equal numbers of cores in the two classes would give a naive mean of 2 mm instead.
The 1.6 mm remains a soil-evaporation estimate, not total crop ET. Nor does it demonstrate that every point in a class lost exactly its mean. Spatial variability, changing shade and the compatibility of core ages remain part of the wider inference.
Adding ten adjacent cores to one small patch does not necessarily provide ten independent samples of the whole field. Preserve positions and grouping. Better knowledge of one patch can be valuable, but it should not be reported as reduced uncertainty across unsampled soils or canopy conditions without a defensible expansion.
CHAPTER 10 OF 12
10. Compare independent methods at compatible boundaries
A weighing lysimeter containing soil and plants may estimate combined evapotranspiration, while root-free microlysimeters focus on soil evaporation. Subtracting one from the other can estimate transpiration only when area, interval and representativeness are compatible and the remaining pathways are adequately treated.
For an illustrative matched day, total ET is supplied as 4.8 mm and representative soil evaporation as 1.6 mm. The residual is 3.2 mm. It is a derived partition, not a direct observation of water leaving every leaf. Errors in either input affect the residual.
With independent standard uncertainties of 0.4 and 0.5 mm, the residual’s standard uncertainty is approximately √(0.4² + 0.5²) = 0.64 mm. Dependence would change that calculation. A small residual difference between treatments may not be distinguishable even when both totals are reported with several decimal places.
A tower footprint can also change with wind and atmospheric conditions, whereas core locations remain fixed. Agreement or disagreement must be interpreted using represented surfaces, not only instruments’ nominal units. The connected eddy-covariance guide develops that broader spatial-support problem.
CHAPTER 11 OF 12
11. Transfer from a bare patch to a cropped or drying surface
Start with a bare soil immediately after a defined wetting event, then change to the same soil beneath an incomplete canopy. Shade, root uptake and spatial variation now matter. The mass-to-depth conversion remains valid, but field representativeness needs a new evaluation.
Next keep the site but compare a fresh core with an older one. An observed difference could reflect changed hydraulic connection, moisture or thermal state. Do not attribute it automatically to a different atmospheric demand when the cores shared the same interval.
Finally shorten the time step. Two weighings a day can estimate interval-average exchange; they do not identify every short evaporation pulse. Continuous weighing offers different information but has its own response and correction requirements. A finer output grid created by interpolation is not equivalent to finer observation.
The transferable habit is to hold one relationship stable while identifying what changed. Conservation and dimensional conversion survive these exercises. Core lifetime, surface representativeness and spatial expansion do not survive by assumption. They travel only with evidence matched to the new setting.
CHAPTER 12 OF 12
12. Report measured loss and field inference separately
A useful statement identifies the core boundary, exposed area, interval, age, water additions and removals, mass precision and quality flags. It then explains how the local result represents the intended soil surface. Keep these stages visible rather than presenting one field evaporation number with an unexplained method label.
For example: a 0.01 m² core lost 18 grams water over the stated rain-free interval, giving 1.8 mm core evaporation under the no-drainage and no-handling-loss assumptions. Applying it to a field requires the stated surface-class and age checks. The first calculation is narrower and better identified than the second.
If a wetting event or missing weighing invalidates an interval, preserve its status. A model may estimate the missing evaporation, but the substituted value should remain distinguishable from measured mass change. Do not erase the reason an apparently smooth series was incomplete.
For retrieval, explain why the mass can rise while evaporation continues, why an older isolated core may differ from intact soil, and why equal numbers of shaded and exposed samples need not give an area-representative mean. Those answers show that the instrument is connected to a physical system rather than treated as a tiny unquestionable field.
Independent worked case and evidence checks
Independent case: Jo receives a fictional core record with exposed area 0.008 m². Total mass falls by 0.012 kg during an interval receiving 0.008 kg caught rain. A separately measured 0.004 kg water drains, and no solid material is lost. First reconstruct evaporation as a water mass before converting to depth.
Here ΔM = −0.012 kg. Evaporation is 0.008 − 0.004 − (−0.012) = 0.016 kg, or 0.016 litre using the supplied teaching density. Dividing by 0.008 m² gives 2 mm. Reporting only the 0.012 kg mass fall would give 1.5 mm and omit rain and drainage.
Change one assumption: 0.003 kg of the recorded mass decrease was a documented soil fragment removed during handling. The water-only storage change is then −0.009 kg, so evaporation is 0.008 − 0.004 + 0.009 = 0.013 kg, corresponding to 1.625 mm. The original arithmetic was consistent with its original assumptions; better evidence changed the identity of part of the loss.
Now compare two fictional surface classes. A representative exposed class occupies 40% of the field and has corrected evaporation of 2.5 mm; a shaded class occupies 60% and has 1.0 mm. The area estimate is 1.6 mm. A total field ET observation of 4.0 mm would leave 2.4 mm as a compatible residual transpiration estimate, not as a microlysimeter observation.
For an age check, two fresh cores in one setting give 1.9 and 2.1 mm, while two older cores give 1.2 and 1.4 mm. Their group means are 2.0 and 1.3 mm. The difference motivates testing age-related divergence. It does not prove that every old core should receive a fixed 0.7 mm correction; initial moisture, positions and thermal conditions may also differ.
A stronger fictional comparison places fresh and older cores in matched settings across several relevant conditions and retains independent surrounding-soil observations. Record differences by age and wetting history before choosing a usable exposure period. Do not choose the period only because it makes a preferred field estimate agree.
Use a balance uncertainty example to locate the next improvement. If each independent weighing has standard uncertainty 0.001 kg, the difference has approximately 0.001414 kg standard uncertainty. Over 0.008 m², that is about 0.177 mm water-equivalent uncertainty before rain, drainage, area and representativeness uncertainties are included. Extra digits in the final 1.625 mm do not remove those limits.
If the objective is detecting a 0.05 mm treatment difference, this weighing uncertainty alone would be material. If the objective is distinguishing 1 mm from 4 mm loss, the same instrument may still provide useful evidence. Adequacy depends on the question and complete uncertainty, not a rule that every small mass-change instrument is either accurate or inaccurate for all purposes.
Finish with an evidence ledger: direct masses and area; documented water transfers and handling; inferred evaporation; and the separate assumptions for field expansion. Ask which observation would challenge each step. This turns a small weighing exercise into transferable scientific reasoning without pretending that every isolated soil core remains identical to the land around it.
An additional geometry check separates a smaller footprint from a smaller edge influence. For a circular exposed surface, perimeter divided by area is 2/r. A radius of 0.05 m gives 40 m⁻¹, while 0.02 m gives 100 m⁻¹. The smaller core has more boundary length per unit exposed area. This does not calculate an evaporation bias or establish an optimal diameter. It explains why an edge-related disturbance cannot automatically be assumed to shrink in proportion to the soil area.
A second uncertainty check distinguishes a fixed balance offset from drift. If the same fixed offset is added to both weighings, it cancels in their difference. If the offset changes between weighings, it does not cancel. Conversely, a common multiplicative scale error affects the mass difference. Treating all balance errors as two independent random numbers can therefore misdescribe the uncertainty. Preserve calibration and weighing history before applying the convenient root-sum-square calculation used in the earlier exercise.
For a final sampling check, suppose four neighbouring cores share the same shading and soil condition, while a distant unsampled area has different texture. A small spread among the four readings demonstrates local consistency; it does not measure variation in the distant area. The next useful observation may be a deliberately selected contrasting setting rather than another adjacent replicate. This distinction keeps instrument repeatability, local variability and field representativeness from collapsing into one misleading confidence statement.
Worked comparison
| Fictional surface class | Area share | Representative soil evaporation | Contribution |
|---|---|---|---|
| Exposed soil | 30% | 3 mm | 0.9 mm |
| Shaded soil | 70% | 1 mm | 0.7 mm |
| Field estimate | 100% | Area-weighted | 1.6 mm |
Questions and practical checks
Does every mass decrease measure evaporation?
No. Drainage, leakage or solid-material loss can contribute. Identify the material and pathways before converting mass to water depth.
Is a sealed base automatically ideal?
It simplifies some accounting but changes hydraulic and thermal boundaries. Whether the core represents surrounding soil depends on design and conditions.
How often must a core be replaced?
There is no universal interval supplied here. Exposure age, soil depth, wetness, roots and thermal behaviour need evidence for the particular study. Published tests should not be turned into an unconditional schedule.
Can I call the result crop ET?
Not for a root-free soil-evaporation core. Total crop evapotranspiration includes plant transpiration and requires a compatible broader observation or model.
Why keep core identity after replacement?
A new sample has a new mass and initial condition. The difference between two different cores is not one core’s evaporation. Preserve sample and baseline histories.
Continue with the connected subjects
Understand soil drying and broader fluxes
Soil evaporation curves, drying stages and cumulative loss
Eddy-covariance ET and the surface represented by a flux
Check the account and comparison
Budget closure and uncertainty in residual partitions
Atmometers and the difference between reference demand and actual loss
Compare soil and plant measurement boundaries
Ventilated evaporation chambers: vapour export, storage and surface disturbance offers a contrasting soil-evaporation method. Compare both methods’ altered boundaries, rather than assuming agreement proves an undisturbed surface.
Stem heat-balance sap flow: thermal storage and whole-plant water accounts follows the plant pathway. Combine soil and plant estimates only after their areas, intervals and water-storage terms are compatible.
Sources and scope
Boast and Robertson (1982): isolated-core method and exposure-time evaluation
USDA ARS: wall materials, capping and thermal disturbance
Use of microlysimeters to measure evaporation from sandy soils (1993)
Marek and colleagues (2019), USDA ARS: isolated-core interpretation and spatial sampling
The numerical cases in this guide are original fictional teaching examples, not measurements from a real site. This is an educational explanation, not a field safety instruction, engineering approval or an official certification. For a real project, use qualified practitioners, current product documentation and site-specific evidence. Sources were checked on 30 October 2026.