Salt changes the surface condition, not the rules of water and energy accounting
A freshwater pan and a salty lake stand under similar weather. It is tempting to apply the pan depth directly, or reduce it by the same percentage as the salt concentration. Neither shortcut follows automatically. Dissolved substances change the vapour pressure supported by the surface, while lake temperature, mixing and heat storage can change at the same time.
This guide follows saline-lake evaporation from water activity to a defined vapour-pressure difference, then connects changing salinity with water mass and lake volume. Original examples distinguish a deliberately simplified transfer model from a complete evaporation estimate. They also show why freshwater entering a salty lake can create a surface patch with different exchange.
We do not estimate a current real lake’s loss or predict its future level. The examples use supplied fictional activities, temperatures, coefficients and densities. Keep the physical relationships separate from those illustrative inputs, and use the linked specialist studies to see how actual research investigates the coupled system.
Hydrology subject guide · Learning that transfers into life · How X Works hub
Choose your reading route
Route 1: Define the saline surface · Chapters 1–3
Name water activity without confusing it with salt percentage
Route 2: Calculate the driving difference · Chapters 4–5
Route 3: Follow mixing and stores · Chapters 6–8
Allow temperature feedback rather than holding it silently fixed
Route 4: Check mass and inference · Chapters 9–10
Route 5: Transfer and report · Chapters 11–12
Full chapter index · Independent worked case · Questions and practical checks
Chapter index
Define the saline surface · Chapters 1–3
Name water activity without confusing it with salt percentage
Keep the air state and water-surface state separate
Distinguish a driving difference from a complete flux
Calculate the driving difference · Chapters 4–5
Test how humidity changes the apparent salinity effect
Keep the sign when the simplified gradient reverses
Follow mixing and stores · Chapters 6–8
Allow temperature feedback rather than holding it silently fixed
Represent surface dilution and vertical mixing
Weight patches by compatible area and time
Check mass and inference · Chapters 9–10
Keep salt mass and water mass in separate accounts
Distinguish freshwater-equivalent depth from brine-volume change
Transfer and report · Chapters 11–12
Transfer from a pan or one lake without losing assumptions
Report a saline evaporation result with its state definitions
CHAPTER 1 OF 12
1. Name water activity without confusing it with salt percentage
Water activity describes the thermodynamic availability of water in a solution relative to a pure-water reference under the specified conditions. For the simplified surface-equilibrium relationship used here, vapour pressure over the solution is aw times saturation vapour pressure over pure water at the same surface temperature.
Write esaline = aw × es(Tsurface). The activity aw is dimensionless. An activity of 0.80 does not mean the solution is 20% salt by mass, nor does it mean actual evaporation is necessarily 80% of a freshwater flux. Composition, concentration and temperature influence activity through their own relationship.
For a fictional surface temperature with pure-water saturation vapour pressure supplied as 3.0 kPa, aw = 0.80 gives 2.4 kPa over the solution. This is a surface vapour-pressure calculation. It has not yet included the air’s vapour pressure, transport or the energy required for evaporation.
The Mor and colleagues field study examines salinity-dependent exchange over a diluted plume and the surrounding hypersaline lake. The useful principle is a changed surface condition. Its particular measurements should not be treated as a universal activity–concentration table for every salt mixture or lake.
CHAPTER 2 OF 12
2. Keep the air state and water-surface state separate
The air above a lake has an actual vapour pressure ea. A simple outward vapour-transfer model depends on the difference between the surface-supported vapour pressure and ea, not on the surface value alone. Surface temperature matters because it controls the pure-water reference vapour pressure in that difference.
For freshwater, the simplified difference is es(Tsurface) − ea. For saline water it is aw × es(Tsurface) − ea. The two use the same air value only when the comparison explicitly holds the relevant atmospheric state constant.
With supplied es = 3.0 kPa and ea = 1.5 kPa, freshwater has a 1.5 kPa difference and the aw = 0.80 solution has 0.9 kPa. The ratio is 0.9/1.5 = 0.60, not 0.80. Subtraction of the same atmospheric vapour pressure changes the relative effect.
Relative humidity measured at air temperature cannot simply replace ea or be compared numerically with aw when air and water temperatures differ. Convert and define the states properly. A comparison between two dimensionless quantities can still be physically misleading if their reference temperatures differ.
CHAPTER 3 OF 12
3. Distinguish a driving difference from a complete flux
For an intentionally simple classroom model, let evaporation depth rate E = K × [aw es(Tsurface) − ea], where K is a supplied transport coefficient with units chosen to produce depth per time. This model helps inspect the gradient. It is not a complete lake energy-balance method or a universal field formula.
Take K = 2 mm/day per kPa, es = 3.0 kPa and ea = 1.5 kPa. Freshwater gives 3.0 mm/day, while aw = 0.80 gives 1.8 mm/day. These rates follow only because the exercise holds K, surface temperature and air state fixed.
Real evaporation also interacts with available energy, turbulence, stability and heat storage. A change in salinity can alter some of these conditions. The calculation’s strength is that its controlled assumptions are visible; its limitation is that those assumptions cannot be transferred without testing.
If an analyst takes a measured freshwater-pan rate and multiplies by 0.80 merely because activity is 0.80, the operation omits the atmospheric subtraction even in this simple model. The result might accidentally agree with a record under one condition while lacking a defensible explanation.
CHAPTER 4 OF 12
4. Test how humidity changes the apparent salinity effect
Keep the fictional es at 3.0 kPa and aw at 0.80, but change atmospheric vapour pressure. At ea = 0.6 kPa, the freshwater difference is 2.4 kPa and saline difference is 1.8 kPa. Their ratio is 0.75. At ea = 1.5 kPa, the ratio was 0.60.
At ea = 2.1 kPa, the differences are 0.9 and 0.3 kPa, giving a ratio of about 0.333. The activity is unchanged, but its relative effect on this fixed-temperature driving difference is larger in the more humid comparison.
These examples explain why a universal fixed percentage reduction is unreliable. They do not predict how actual humidity, lake temperature and transport will co-evolve. A real observation needs those variables rather than a sequence of substitutions selected to produce a desired rate.
As the freshwater difference approaches zero, ratios can become especially sensitive or cease to be useful. Report the actual signed differences as well as a percentage. A dramatic ratio based on a tiny denominator can sound like a large water-volume change when both flux estimates are small.
CHAPTER 5 OF 12
5. Keep the sign when the simplified gradient reverses
An outward-only evaporation label can hide a change in the direction of net vapour exchange. In the supplied fixed-temperature model, a negative surface-minus-air vapour-pressure difference favours net transfer toward the surface rather than outward evaporation.
With es = 3.0 kPa, aw = 0.80 and ea = 2.6 kPa, freshwater still has a positive 0.4 kPa difference, while the saline surface has −0.2 kPa. The simple model therefore gives different directions under those explicitly fixed states. It is not a claim that every salty lake condenses under a particular regional humidity.
Actual condensation would release latent heat and could change surface temperature, so a sustained prediction needs coupled water and energy accounts. The signed comparison is a diagnostic of the supplied states, not an unlimited flux forecast.
Do not clip negative model results to zero without describing the output definition. A model of gross evaporation constrained non-negative and a model of net vapour exchange answer different questions. The connected dew-collection guide helps distinguish formation, retained water and delivered collection on another kind of surface.
CHAPTER 6 OF 12
6. Allow temperature feedback rather than holding it silently fixed
If evaporation changes, the energy partition and surface temperature can change too. Lower latent heat loss can leave more energy for warming or other heat transfers. A fixed-temperature comparison is useful for isolating a mechanism but not a complete description of the evolving lake.
For a fictional sensitivity test with aw = 0.80 and ea = 1.5 kPa, increasing the supplied pure-water reference es from 3.0 to 3.5 kPa raises the saline difference from 0.9 to 1.3 kPa. The warmer surface partly offsets the activity reduction in this controlled calculation.
At supplied es = 3.75 kPa, the saline difference is 1.5 kPa, matching the original freshwater difference at es = 3.0. This does not establish that a lake will warm by a particular amount or achieve the matching condition. It shows why comparisons must state temperatures rather than assuming similar weather means identical water surfaces.
The lake heat-storage guide explains how radiation, depth and stored energy affect evaporation windows. Use that connection when moving from a local surface relationship to daily or seasonal loss. A thermodynamically sensible surface equation still needs an energy source for sustained water transfer.
CHAPTER 7 OF 12
7. Represent surface dilution and vertical mixing
Freshwater entering a saline lake can create a less saline surface region before mixing fully with deeper water. Evaporation interacts with the surface actually exposed to air, not necessarily a whole-lake average salinity measured at another depth.
Imagine two fictional surface classes at the same supplied temperature and air state. A diluted patch has aw = 0.98; the saltier surface has aw = 0.80. With es = 3.0 and ea = 1.5 kPa, their driving differences are 1.44 and 0.90 kPa respectively. The difference belongs to the defined surfaces, not simply their locations on a map.
Wind can affect both atmospheric transport and water-column mixing. If mixing brings saltier water into a diluted surface, the activity can change while wind also changes turbulence. A simple rule that more wind always causes a proportional evaporation increase can miss this second pathway.
Mor and colleagues studied this coupled problem in a particular diluted plume. Preserve that context when reading their results. A threshold or relative rate observed there is evidence for that study, not a universal wind-speed rule to apply to every stratified saline water body.
CHAPTER 8 OF 12
8. Weight patches by compatible area and time
A whole-lake estimate needs representative surface conditions and their areas through time. One plume observation does not automatically represent the main lake, and one saltier station does not necessarily represent a changing plume.
In a fictional interval, 20% of surface area has a supported evaporation depth of 4 mm and 80% has 2 mm. The area-average depth is 2.4 mm. Over 1 km², that corresponds to 2,400 m³ freshwater-equivalent water loss when the depth definition is appropriate.
Now change plume coverage to 50% while retaining the supplied class depths. The average becomes 3 mm. A change in total loss can therefore arise from area distribution even without changing either class rate. Report changing coverage rather than hiding it inside a fitted lake-wide coefficient.
If area and rates vary together, accumulate interval volumes before summarising. A simple mean of plume fractions multiplied by a simple mean rate can be wrong when their covariance matters. The existing depth–area guide develops this weighting structure for changing reservoir geometry.
CHAPTER 9 OF 12
9. Keep salt mass and water mass in separate accounts
Evaporation of water from a solution generally concentrates retained dissolved material when other pathways are excluded. For a simplified exercise, let Mw be water mass and Ms dissolved salt mass. Salt mass fraction is Ms/(Mw + Ms), not Ms/Mw and not automatically a water-activity value.
Start with a fictional solution containing 90 kg water and 10 kg nonvolatile dissolved salt. Its salt mass fraction is 10%. Remove 10 kg water through evaporation with no salt precipitation, aerosol loss or other transfers. The remaining solution contains 80 kg water and 10 kg salt, so its salt mass fraction is about 11.11%.
The increased concentration does not specify the new water activity without an appropriate composition-dependent relationship. Nor can a concentration increase alone identify evaporation uniquely if inflow, dissolution, salt removal or mixing are possible.
When salts precipitate, distinguish dissolved salt from solid salt retained within the broader system. Total salt can be conserved while dissolved concentration follows a different path. The teaching examples exclude that complication explicitly; a real brine budget cannot remove it simply because the arithmetic is easier without it.
CHAPTER 10 OF 12
10. Distinguish freshwater-equivalent depth from brine-volume change
An evaporated water mass can be expressed as an equivalent liquid-water volume using a stated reference density, then divided by surface area to obtain depth. That is not automatically identical to the reduction in total solution volume, because brine density and composition can change.
For an illustrative account, initial solution mass is 100 kg and its supplied density is 1,100 kg/m³, giving about 0.09091 m³. After 10 kg water evaporates, solution mass is 90 kg and a separately supplied final density is 1,125 kg/m³, giving 0.08000 m³. Solution volume decreases by about 0.01091 m³.
The removed 10 kg water corresponds to 0.01000 m³ at a reference density of 1,000 kg/m³. The two volume changes differ because the exercise explicitly supplied different solution densities. These densities are fictional inputs, not an equation of state or a salinity calibration.
Do not use lake-level decline alone as an evaporation measurement. Geometry, inflow, withdrawals and other pathways also affect level and storage. A saline system adds composition and density considerations to the same water-budget identifiability problem faced in freshwater reservoirs.
CHAPTER 11 OF 12
11. Transfer from a pan or one lake without losing assumptions
First change a freshwater pan to a saline pan while holding a controlled meteorological and geometric comparison. The activity term matters, but surface temperature and heat storage should still be checked. A successful small-container experiment does not establish a lake-wide rate by itself.
Next change from a pan to a deep lake. Fetch, thermal storage, stratification and spatial salinity patterns can differ. The original coefficient may be compensating for pan-specific properties rather than representing a universal salinity effect.
Then change salt composition while retaining the same numerical concentration. Similar mass fractions need not imply identical activities or densities. A coefficient derived for one brine should not be transferred solely because both are labelled salty water.
Finally change the purpose from evaporation depth to future lake level. You need a full water and salt budget, geometry, external forcing and uncertainty over the forecast period. A physical mechanism can inform that model without becoming a complete prediction. This guide intentionally stops short of estimating any real lake’s future level.
CHAPTER 12 OF 12
12. Report a saline evaporation result with its state definitions
A clear statement gives surface temperature, activity or its estimation method, atmospheric vapour state, transport or energy method, interval and spatial support. If a freshwater reference is used, explain exactly what differs and what is assumed common.
For example: under the supplied same-temperature, same-transport model, the saline driving difference was 0.9 kPa against 1.5 kPa for freshwater, giving a rate ratio of 0.60. This is narrower than saying salinity always reduces evaporation by 40%, and it identifies what another reader must change to test a different state.
Keep water-mass loss, freshwater-equivalent depth, brine-volume change and observed level change separately labelled. They are related but not interchangeable measurements. A report becomes much easier to audit when each number retains its boundary and reference density.
For independent recall, explain why activity is not salt mass fraction, why flux ratio is not generally activity, why wind can alter the surface salinity as well as air transport, and why evaporated water volume can differ from solution-volume decline. Those distinctions turn a familiar salt-and-water story into a transferable account of the coupled system.
Independent worked case and evidence checks
Independent case: Aisha compares fictional fresh and saline surfaces using supplied es = 4.0 kPa, ea = 2.0 kPa and K = 1.5 mm/day per kPa. The saline activity is 0.85. Hold temperature and transport identical only for this first calculation.
Freshwater has a driving difference of 2.0 kPa and a modelled rate of 3.0 mm/day. The saline surface has 0.85 × 4.0 − 2.0 = 1.4 kPa and a rate of 2.1 mm/day. The ratio is 0.70, not 0.85. Applying activity directly to the freshwater rate would give 2.55 mm/day and overstate the result of the specified model.
Change only atmospheric vapour pressure to 1.0 kPa. The differences become 3.0 and 2.4 kPa; the rates are 4.5 and 3.6 mm/day. Their ratio is now 0.80, even though water activity remains 0.85. This changed-context test isolates the atmospheric subtraction without pretending to forecast a real weather response.
Now suppose a separately supplied warmer saline surface has es = 4.4 kPa while ea returns to 2.0 kPa. Its driving difference is 0.85 × 4.4 − 2.0 = 1.74 kPa, giving 2.61 mm/day. The state changed, so comparison with the original fixed-temperature saline result should name the temperature feedback rather than call it a coefficient error.
For spatial expansion, use supported class depths of 2.1 mm over 80% of a fictional 2 km² lake and 3.0 mm over the remaining diluted 20%. Average depth is 2.28 mm and freshwater-equivalent volume is 0.00228 × 2,000,000 = 4,560 m³. A single diluted-patch value applied everywhere would give 6,000 m³, answering an unsupported uniform-surface scenario.
For a salt ledger, the exercise starts a separate bounded tank with 180 kg water and 20 kg dissolved salt. Remove 30 kg water, with no other transfers or precipitation of salt. Final salt mass fraction is 20/170, about 11.76%, compared with the initial 10%. This does not supply the final activity; a further composition relationship is required.
Finally compare two possible explanations for a measured salinity increase in a lake. Water removal with salt retained can produce it. Addition of salt or mixing with a saltier store can also produce it. A concentration observation alone cannot choose uniquely among these accounts. Independent water, salt and mixing evidence is the useful next step.
Write the conclusion in four lines: specified surface and air states; conditional transfer calculation; spatial and temporal expansion; and the separate water/salt budget. If one line is missing, resist filling it with a familiar universal factor. The controlled calculation is valuable precisely because its assumptions remain available for challenge.
An additional unit test distinguishes concentration bases. A fictional solution contains 10 kg salt and 90 kg water. It is 10% salt by total solution mass, but the salt-to-water mass ratio is 10/90, about 0.1111. Reporting both as 10% would erase the denominator. A concentration in kilograms per cubic metre additionally requires solution volume and therefore density. None of these numbers is automatically a conductivity reading or an activity. Retain the definition before comparing samples or applying a thermodynamic relationship.
A time-weighting test gives two equal-length intervals with supplied evaporation rates of 2 and 4 mm/day and lake areas of 2 and 1 km² respectively. Their daily losses are both 4,000 m³, totalling 8,000 m³ over two days. Averaging rate to 3 mm/day and area to 1.5 km², then multiplying by two days, gives 9,000 m³. The separate averages hide the relationship between rate and exposed area. Accumulate compatible interval volumes before choosing a compact summary.
A spatial-support test changes only the salinity sampling depth. A deep sample describes a saltier layer while the air exchanges with a diluted surface. Inserting the deep-layer activity into a surface-flux model can underestimate the supplied surface driving difference. The solution is not automatically to discard the deep sample; it answers a different question about the water column. Retain depth, location and time so the temperature and composition used in the surface model actually describe the intended interface.
For a model-identifiability test, suppose one fitted coefficient reproduces an observed lake loss despite incorrect surface temperature and activity inputs. A good fit does not reveal which compensating error occurred. A later season with different humidity or mixing can expose the hidden mismatch. Independent surface temperature, composition and flux evidence help test those links. Calibration should not silently turn a physically incomplete input description into a universally transferable coefficient.
The final useful distinction is between a mechanism and an attribution. Dissolved material can reduce surface-supported vapour pressure at the same temperature. That established mechanism does not by itself prove why a particular lake’s observed level changed over a decade. Changing inflow, withdrawal, geometry and climate still belong in the larger account. Explain the mechanism clearly while retaining the additional observations needed for the broader claim; doing so makes the scientific result more usable rather than less important.
For a final independent explanation, begin with the surface and air states rather than the name of the lake. Reconstruct the signed driving difference, identify the supplied transport and energy assumptions, and only then expand over area and time. That order makes the same reasoning usable in an unfamiliar saline system.
Worked comparison
| Fictional same-temperature state | Freshwater difference | Saline difference at aw 0.80 | Saline/fresh ratio |
|---|---|---|---|
| es 3.0; ea 0.6 kPa | 2.4 kPa | 1.8 kPa | 0.75 |
| es 3.0; ea 1.5 kPa | 1.5 kPa | 0.9 kPa | 0.60 |
| es 3.0; ea 2.1 kPa | 0.9 kPa | 0.3 kPa | About 0.333 |
Questions and practical checks
Is water activity one minus salt percentage?
No. Activity is a thermodynamic quantity, not a simple relabelling of concentration. Use an appropriate composition and temperature relationship.
Can I multiply freshwater evaporation by activity?
Not generally. Even a simple transfer model uses the surface-minus-air vapour-pressure difference. Temperature, transport and energy feedbacks add further considerations.
Does more wind always increase saline-lake evaporation?
Not as a universal rule. Wind affects air transport and can change surface composition through mixing. Examine the actual surface and atmospheric response.
Does evaporation remove salt in the simple examples?
The exercises assume water leaves while nonvolatile salt remains, with no precipitation or other transfers. Real budgets must examine dissolved, solid and transported salt separately when relevant.
Is freshwater-equivalent loss the same as brine-volume decline?
Not automatically. Equivalent water volume uses a reference density; solution volume depends on the brine’s changing mass and density. Lake level adds geometry and other water-budget terms.
Continue with the connected subjects
Follow surface exchange and energy
Evaporation pans and the limits of transferring a coefficient
Lake heat storage and the evaporation energy window
Eddy-covariance fluxes and represented surfaces
Connect depth, composition and lake-level claims
Evaporation depth, changing area and reservoir water balance
Budget closure and why a residual does not identify its cause
The Dead Sea: a wider introduction to salinity and changing water level
Sources and scope
Mor and colleagues (2018): surface salinity, dilution and evaporation over the Dead Sea
Original study: atmospheric boundary-layer processes and evaporation over a shallow saltwater lagoon
Estimation of evaporation from saline water (2020): concentration and depth comparisons
Evaporation from the hypersaline Aral Sea: a water-activity-based modelling study
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.