Potential evapotranspiration (PET) is the rate at which water would pass from a specified land surface to the atmosphere through evapotranspiration if water availability did not restrict the process. It combines evaporation from soil and other wet surfaces with transpiration by plants. Used in hydrology, agriculture, and climatology, PET describes a hypothetical, adequately supplied surface rather than necessarily observed water loss. Its numerical value depends on both weather conditions and assumptions about vegetation and surface properties.
Meaning and related quantities
Evapotranspiration transfers liquid water from the land surface into atmospheric water vapor. Evaporation occurs from soil, intercepted rainfall, and other exposed water, whereas transpiration is water loss from plants, predominantly through leaf stomata.
Three quantities must be distinguished:
- Actual evapotranspiration is the water loss that occurs under existing conditions, including any limitations imposed by soil moisture, vegetation condition, or plant physiology.
- Potential evapotranspiration is the estimated loss under a stated set of surface conditions when water supply is nonlimiting.
- Reference evapotranspiration is the loss from a precisely defined reference surface, allowing consistent comparisons across locations and dates.
PET and reference evapotranspiration are sometimes used interchangeably, especially in climate datasets, but they are not universally equivalent. PET has historically been defined in different ways, including evaporation from extensive, well-watered vegetation or an idealized saturated surface. Reference evapotranspiration deliberately fixes characteristics such as vegetation height, reflectivity, and surface resistance.
The Food and Agriculture Organization’s grass reference surface, for example, is a hypothetical, actively growing, completely shading, adequately watered crop with a height of 0.12 m, an albedo of 0.23, and a surface resistance of . Reference estimates based on this surface are normally denoted . Other standards include a taller, alfalfa-like reference surface.
Consequently, a PET value is fully interpretable only when its calculation method and assumed surface are known. It is not a universal maximum for every landscape: a tall, rough, well-watered canopy can lose more water than a short-grass reference surface under the same measured weather conditions.
Physical controls
Evapotranspiration requires both energy to vaporize water and atmospheric conditions that permit water vapor to leave the surface.
Available energy. Net radiation generally supplies most of the energy for evaporation. Some energy heats the soil or vegetation, and some is exchanged with the air as sensible heat. Warm, dry air moving over an irrigated field can supply additional energy by advection, allowing latent heat loss to exceed the locally available net radiation.
Atmospheric dryness. The vapor-pressure deficit expresses the difference between saturation vapor pressure and actual vapor pressure. A larger deficit generally increases the tendency for water to evaporate, although the response also depends on surface temperature and plant regulation.
Air movement. Wind and turbulence remove moist air from the surface and replace it with drier air. Their influence depends on vegetation height, surface roughness, and atmospheric stability.
Surface characteristics. Vegetation height, leaf area, reflectivity, and canopy resistance affect energy absorption and vapor transfer. Adequate soil water does not eliminate biological controls: stomata may respond to light, atmospheric dryness, and carbon dioxide even when roots have access to water.
Air temperature influences saturation vapor pressure and is often correlated with radiation and season. Nevertheless, temperature alone does not determine PET. Equally warm locations can have substantially different potential water losses because of differences in cloudiness, humidity, wind, and vegetation.
Historical development
The modern concept became established in the mid-twentieth century. In 1948, C. W. Thornthwaite used potential evapotranspiration as a central element of a climatic water-balance classification. His temperature-based approach made it possible to estimate potential water use where detailed meteorological observations were unavailable.
Also in 1948, Howard Penman introduced a combination method that brought together the surface energy balance and aerodynamic transfer of water vapor. This provided a physical basis for estimating evaporation from meteorological observations.
Later work by John Monteith incorporated canopy and aerodynamic resistances into the combination framework. The resulting Penman–Monteith equation became a principal foundation for calculating evapotranspiration from vegetation. Subsequent standardization of reference surfaces helped separate weather-driven evaporative demand from differences among crops and land-cover types.
Estimation methods
Combination and resistance-based methods
The Penman–Monteith equation can be written in energy-flux form as
Here:
- is the latent heat flux associated with evapotranspiration;
- is net radiation;
- is the ground heat flux;
- is the slope of the saturation vapor-pressure curve;
- is air density;
- is the specific heat capacity of air;
- is vapor-pressure deficit;
- is aerodynamic resistance;
- is surface resistance;
- is the psychrometric constant.
With consistent units, the equation combines energy supply with atmospheric vapor transport. Dividing latent heat flux by the latent heat of vaporization and converting units gives an equivalent water-loss rate.
The equation itself is not automatically a PET equation. It can represent different surfaces and conditions depending on its parameters. To calculate potential or reference evapotranspiration, surface resistance and other properties must be assigned consistently with the intended well-watered or reference surface.
The standardized FAO Penman–Monteith method estimates grass-reference evapotranspiration using radiation, temperature, humidity, and wind observations. Its explicit surface definition supports comparison between sites and provides a common basis for crop-water calculations.
Radiation-based methods
Radiation-based approaches estimate evaporation mainly from available energy. The Priestley–Taylor method, for example, modifies an equilibrium evaporation expression with an empirical coefficient. It is often applied to extensive wet surfaces under conditions where local advection is limited.
Such methods require fewer aerodynamic observations than full combination methods. Their performance can deteriorate where dry-air advection or strong winds make atmospheric transport particularly important.
Temperature-based methods
The Thornthwaite method estimates monthly PET from mean air temperature, a heat index, and adjustments for day length and month length. Its modest data requirements have made it influential in historical climate analysis.
The Hargreaves–Samani method uses temperature, the daily temperature range, and extraterrestrial radiation. Temperature range serves partly as an indirect indicator of incoming solar radiation.
These methods are useful where observations are sparse, but their empirical relationships do not transfer equally well among climates. Errors may be substantial in windy, arid, cloudy, mountainous, or strongly maritime environments. Local calibration can improve performance without making the methods physically interchangeable.
Pan observations and direct measurements
Evaporation pans provide an observational indicator of evaporative conditions. Conversion from pan evaporation to reference evapotranspiration requires coefficients that account for pan design, exposure, wind, humidity, and surrounding ground cover.
A lysimeter measures evapotranspiration from a contained soil–vegetation system through changes in mass or a detailed water balance. A well-watered lysimeter with suitable vegetation can help evaluate reference or potential estimates. It still measures a particular surface rather than an abstract, surface-independent PET.
Units and temporal scale
PET is commonly expressed as millimeters per day, month, or year. A depth of over corresponds to one liter of water. It therefore expresses an equivalent depth over an area, not a decline in soil-water content or water-table level.
Values calculated at different time steps are not always directly interchangeable. Because the underlying relationships are nonlinear, an estimate based on monthly mean weather may differ from the sum of estimates calculated from daily or hourly observations. Data products should specify their temporal resolution, aggregation procedure, and treatment of missing observations.
Applications
Climatic water balance and aridity
Comparing PET with precipitation indicates whether climatic water supply is large or small relative to potential atmospheric demand. In a simple water balance,
where is precipitation, is actual evapotranspiration, represents net water export, and is the change in stored water, assuming other inputs are negligible.
PET does not replace actual evapotranspiration in this conservation equation. Instead, hydrological models often use PET as a demand term and constrain actual loss according to soil-water availability and vegetation.
An aridity index is commonly expressed as
Lower values indicate less precipitation relative to potential evaporative demand. Its interpretation depends on the PET method, averaging period, and seasonal distribution of rainfall. Equal annual ratios can conceal very different seasonal water shortages.
Agriculture and irrigation
Reference evapotranspiration provides a weather-based baseline for estimating crop water requirements. Under standard, non-water-stressed conditions, a common relationship is
where is crop evapotranspiration and is a crop coefficient that changes with crop type and growth stage.
This relationship supports irrigation calculations, but potential crop water use is not the same as the required irrigation application. Effective rainfall, stored soil water, irrigation efficiency, and other components must also be considered.
Drought and hydrological modeling
PET is used in drought indices that represent climatic water demand as well as rainfall shortages. It also serves as an input to models of soil moisture, runoff, and groundwater recharge.
A rise in PET can indicate increased potential water stress, but it does not establish that actual evapotranspiration has risen. During severe soil drying, actual water loss may decline while potential demand remains high.
Limitations and interpretation under climate change
PET is an inferred counterfactual: it asks how much water a specified surface could lose if water were sufficiently available. It cannot generally be measured as a unique property of the atmosphere independently of that surface.
Method choice is particularly important in climate change studies. Temperature-only methods may translate warming into increased PET without explicitly representing changes in radiation, humidity, or wind. Physically based methods include more relevant controls, but remain sensitive to weather-data quality and assumptions about surface resistance.
Rising carbon dioxide can alter stomatal behavior and vegetation structure. A fixed-reference calculation intentionally holds its reference vegetation properties constant, so it need not describe the future response of an actual forest, grassland, or crop. Changes in reference evapotranspiration, potential evapotranspiration from a changing canopy, and actual ecosystem water use are therefore distinct quantities.
Comparisons among PET datasets require attention to the definition of the reference surface, the estimation equation, input observations, spatial resolution, and time scale. Differences among these choices can materially affect inferred aridity, drought severity, and projected water availability.