Most of the precipitation that falls on the Colorado River Basin goes back up to the atmosphere through evapotranspiration (ET) before reaching a stream or river. ET encompasses several types of water loss: evaporation of moisture from soils and open water, transpiration from plants and crops, and sublimation from the snowpack. Since ET is such a large component of the water budget, from the field scale to the basin scale, ET can strongly influence the moisture status (dryness/wetness) of the land surface and ultimately, the volume of runoff.
The concepts and terminology around ET can be confusing and are sometimes misused. Strictly speaking, ET refers to the actual loss of water from the land surface. It's analogous to the amount of water that a thirsty person actually drinks. ET is driven by evaporative demand (E0)--the atmosphere's "thirst" for surface moisture--but the amount of ET (water loss) is necessarily limited by the available water at the surface. So cumulative ET over an extended period (e.g., one year) will not exceed precipitation--plus irrigation water, in the case of irrigated cropland. Sometimes, the term Actual ET (AET) is used in place of ET, to better distinguish it from other terms.
Evaporative demand (E0) is the potential loss of water from the land's surface. Evaporative demand is equivalent to the more common term Potential ET (PET). E0 and PET can and usually do exceed ET, just as a person's thirst can exceed the water available to them and that they are able to drink.
Reference ET (ET0) is closely related to E0 and PET. It is an estimate of the upper bound of ET losses (i.e., consumptive water use) from irrigated cropland, and it is the measure typically reported by ag-weather stations (see Data and tools below). In principle, with a fully watered crop, Reference ET would be the same as (actual) ET, but in practice, Reference ET is often greater than ET.
Reference ET and PET can be estimated using an equation that inputs meteorological variables (temperature, solar radiation, humidity, winds), or derived from remote sensing data using a land-surface model. Direct in-situ measurements of PET and Reference ET (e.g., pan evaporation) are very sparse across the Colorado River Basin.
(Actual) ET is more challenging to quantify. ET can be estimated using a land-surface (hydrology) model with meteorological inputs, or by assimilating satellite observations of land-surface temperature into an energy-balance model. More direct measurements of ET can be made using Eddy Covariance (EC) methods, which requires multiple instruments on a single tower to measure the vertical transfer of energy and moisture between the surface and the atmosphere. In September 2022, Reclamation and the Upper Colorado River Commission (UCRC) adopted a satellite-based method, eeMETRIC, as the standard for measuring evapotranspiration (ET) and agricultural consumptive water use in the Upper Basin. Real-time and retrospective eeMETRIC data can be obtained from OpenET, described under Data and tools below.
In the Colorado River Basin, estimates of ET and evaporative demand (E0) and related measures (Reference ET, PET) are used in reservoir operations, irrigation scheduling, and demand and consumptive use modeling. Estimates of watershed-scale ET are also used to validate the water budget simulations in hydrologic models, such as those used by the NOAA Colorado Basin River Forecast Center (CBRFC) for seasonal streamflow forecasts. Estimates of monthly reservoir evaporation and consumptive use by agriculture are also important terms in the Reclamation operations and planning models, and in Reclamation's calculations of natural flows.
Since temperature is the most important driver of evaporative demand, ET and related variables have large seasonal variability, with a maximum in summer and minimum in winter. Consequently, the warming overall climate in the basin has already increased overall E0 and the proportion of precipitation lost to ET, and these trends are projected to continue, increasingly impacting both water supply and water demand.
Data and tools
Several weather-station networks in the basin states, mainly serving the agricultural sector, have instrumentation for all of the variables (temperature, solar radiation, humidity, winds) needed for real-time calculations of Reference ET. Typically, Reference ET is calculated hourly and daily, to the nearest 0.01 in. or 1 mm.
Colorado; 90 active stations in agricultural areas.
Arizona; 27 active stations in agricultural and urban areas.
Utah and adjacent areas in ID, WY, CO, NM; 130 active stations in agricultural and urban areas.
California; 150 active stations in agricultural and urban areas.
EDDI is based on Reference ET (calculated from gridded NLDAS-2 meteorological data, using the same method as for the in-situ data above) and standardized relative to historical Reference ET for each gridpoint over the time window (2 weeks to 1 year) of interest. EDDI can offer early warning of agricultural and hydrologic drought by providing near-real-time information on the emergence or persistence of anomalous evaporative demand.
OpenET is an open-source platform and data viewing tool that brings together several satellite-based ET estimation methods to provide daily, monthly, and annual ET estimates at the field scale. The methods currently in the OpenET ensemble include ALEXI/DisAlexi, eeMETRIC, geeSEBAL, PT-JPL, SIMS, and SSEBop. Free registration required to view data.
State of the Science Report
Chapter 5 of the State of the Science report, Section 5.5 describes ET and evaporative demand, and sources of data, in greater detail.
This section of the OpenET website provides more detail about the several remote-sensing-based ET methods that make up the OpenET ensemble, as well as other datasets (such as Reference ET based on gridded climate data) that inform the Open ET ensemble values.