USGS banner

Water-Resources Investigations Report 01-4234

Estimates of Evapotranspiration from the Ruby Lake National Wildlife Refuge Area, Ruby Valley, Northeastern Nevada, May 1999-October 2000


Seasonal and annual ET rates were determined by estimating site-specific daily ET rates in major habitats and applying these rates to similar areas throughout the refuge. Instrumentation was installed at nine sites to collect micrometeorological data for estimating ET using the Bowen-ratio and eddy-correlation methods. The sites represented wetland, meadow, areas of mixed phreatophytic shrubs and associated bare soil, and desert-shrub upland (table 1).

Year-round accessibility was a factor in the selection of ET sites. In warm weather, access to the wetland sites was by non-motorized canoe. During the winter, when much of the wetland is covered by surface ice, access was by an air boat. The remaining sites were easily reached with field vehicles.

Also considered was fetch, which was deemed adequate at each ET site. Fetch, the horizontal distance from the ET measurement site to a change in surface conditions in the direction of prevailing winds, is assumed to be adequate at 100 times the instrument height (Campbell, 1977). Prevailing winds on the refuge typically came from the southwest during the data-collection period. Adequate fetch implies that the surface is uniform so that the profile of air flow across the area of interest is approximately constant.

Energy Budget and the Bowen-Ratio Method

During the ET process, energy is used to convert water from liquid to vapor and transfer the vapor to the atmosphere. The Bowen-ratio and eddy-correlation methods, which were used in this study to estimate ET, are based on characteristics of the energy budget associated with atmospheric fluxes. The symbols and forms of the equations used in these methods generally follow the nomenclature of Laczniak and others (1999). Detailed information on the equations and methods used in this study to estimate ET is in Nichols (1992) and Laczniak and others (1999).

The balance between incoming and outgoing energy fluxes can be mathematically expressed by the one-dimensional form of the energy-budget equation (eq. 1). In the environment, energy is partitioned by the energy budget into four principal flux components: (1) net radiation, (2) subsurface-heat flux, (3) sensible-heat flux, and (4) latent-heat flux. The term flux refers to flux density, which represents the amount of energy that flows through a horizontal surface of unit area per unit time. Energy terms related to biological processes, such as photosynthesis and the storage of heat in plant biomass, are considered negligible, thus are not included in the energy budget. Energy terms related to the horizontal transfer of heat also are not included because they are assumed to be small compared to the vertical transfer of heat.

Net radiation, which depends on the temperature and reflectivity of the surface exposed, is the major energy source that drives ET processes. Net radiation, the sum of all incoming and outgoing radiation at the surface of the Earth, is considered positive when the sum of incoming radiation exceeds the sum of outgoing radiation (eq. 2).

The subsurface-heat flux is the amount of energy stored in the soil or water column. Because one site was set up over water, the usual soil-heat flux term in the energy-budget equation was replaced with the term subsurface-heat flux (Laczniak and others, 1999, p. 20). The subsurface-heat flux associated with the soil is a function of the change in soil temperature with depth and the thermal and physical properties of the soil (eq. 3). For water, the subsurface-heat flux is a function of the change in temperature of water with depth and the specific heat and density of the water (eq. 4).

Sensible-heat flux is the amount of energy that heats the air directly above the soil, plant canopy, or water surface (eq. 5). Sensible-heat flux is temperature-driven and directly relates to the turbulent transfer of heat.

Energy that is consumed by ET is the latent-heat flux, which is related to the vapor-pressure gradient and the turbulent transfer of vapor (eq. 6). At the Earth's surface, the difference between net radiation and subsurface-heat flux is the energy available (eq. 7) for sensible- and latent-heat fluxes (often called turbulent fluxes).

Net radiation and subsurface-heat flux can be measured in the field using available instrumentation. Sensible-heat and latent-heat fluxes are not easily estimated because turbulent transfer coefficients (kh and kv in eqs. 5 and 6, respectively) are difficult to determine. However, Bowen (1926) determined that if the transfer coefficients are assumed to be equal, the ratio of sensible-heat flux to latent-heat flux is proportional to the ratio of the vertical gradients of temperature and vapor pressure above a surface. This ratio between sensible-heat flux and latent-heat flux is known as the Bowen ratio (eq. 8; Bowen, 1926) and can be approximated from measurements of air temperature and relative humidity at two different heights. Under certain conditions, the Bowen ratio approaches -1 and application of the method is invalidated. When this happens, the calculated value of latent-heat flux (eq. 6) loses numerical meaning (Ohmura, 1982, p. 596). This condition seldom occurred. When it did, however, an average of latent-heat flux from the previous and subsequent time periods was used. In most instances, this condition took place during periods of low ET and probably had little effect on the daily ET computation. The ratio of sensible-heat flux to latent-heat flux was used in a modified form of the energy-budget equation (eq. 9) along with micrometeorological data to compute ET at the four Bowen-ratio sites in the refuge (table 1).

Instrumentation and sensors that could operate for long periods under such adverse weather conditions as high winds, freezing rains, and possible accumulations of snow and ice were required to collect data continuously for application of the Bowen-ratio method. Variations of Bowen-ratio instrumentation also were required to accommodate the differences between sites on land and those over water or bulrush-marsh areas. Solar panels were installed at all sites to recharge batteries used to power the instruments. Energy fluxes and ET were computed every 20 minutes based on a 10- or 30-second sampling interval and were summed to compute daily ET. A schematic of the typical instrument arrangements used to collect micrometeorological data over land and over open water or bulrush marsh is in figure 4.

Instrumentation at the Bowen-ratio land sites (fig. 4A) consisted of:

 two solid-state temperature and relative-humidity probes mounted on an exchange mechanism to measure air temperature and relative humidity at two heights;
 two anemometers to measure wind speed;
 a net radiometer to measure net radiation;
 a set of soil-heat flux plates, thermocouples, and a water-content reflectometer to compute the subsurface-heat flux; and
 two infrared temperature sensors to measure plant-canopy and soil temperatures.

Instrumentation at the open-water and bulrush-marsh sites (figure 4B) was similar to that of the land sites but with a slightly different arrangement. To compute the subsurface-heat flux to or from the water, three thermistor temperature probes extending downward through the water column replaced the heat-flux plates and themocouples. At the open-water site, the temperature and relative humidity of the air were measured at only one height above the water surface. A single anemometer was used to measure wind direction. A floating thermistor measured water-surface temperature, from which (saturated) vapor pressure was calculated. Temperature and vapor pressure differences were computed between the water surface and the elevated sensor. Staff gages were installed at the wetland sites to determine changes in water depth. Data were stored at the Bowen-ratio sites on data loggers and could be retrieved through telecommunication systems.

Eddy-Correlation Method

Eddies are turbulent, highly rotational air flows that move across the surface of the earth transporting water vapor and heat between the surface and the atmosphere. In turbulent air flow, fluxes of water vapor and heat vary irregularly in time and space; for this reason, statistical analyses are used to represent tur-bulent flow. Covariances between two fluctuating variables such as vertical wind speed and water vapor or vertical wind speed and temperature are directly related to turbulent flux (Arya, 1988, p. 118). The eddy-correlation method consists of determining the turbulent fluxes of latent and sensible heat from the covariance of vertical wind speed with vapor density and with air temperature. Latent-heat flux is determined by the covariance of instantaneous departures from the average values of wind speed and vapor density (eq. 10). Latent-heat flux is corrected for oxygen effects (Tanner and Greene, 1989) and for density differences caused by heat and vapor transfer (Webb and others, 1980). Sensible-heat flux is determined by the covariance of instantaneous departures from the average values of wind speed and air temperature (eq. 11).

Net radiation and subsurface-heat flux also are measured at each eddy-correlation site; together with measurements of latent- and sensible-heat flux, these measurements allow an energy budget to be estimated. Evaluation of the energy budget using data collected at eddy-correlation sites provides an indication of instrument efficiency in measuring the available energy. A nonzero energy-budget closure typically suggests instrumentation problems; however, the source of the discrepancy usually is difficult to determine. The un-certainty in computing ET rates by the eddy-correlation method can be inferred from the size of the closure residual. Relative closure of the energy budget is the amount of imbalance relative to the available energy and indicates the amount of available energy that is not accounted for by measurements of turbulent fluxes (eq. 12; Johnson, 1995, p. 7). Although the eddy-correlation method is the most reliable and direct measurement of turbulent fluxes, the method requires sophisticated fast-response instrumentation.

Two similar sets of eddy-correlation instrumentation were used for data collection at five sites. A typical instrumentation configuration for the eddy-correlation method consists of:

 a sonic anemometer with a fine-wire thermocouple to measure instantaneous changes in vertical wind speed and air temperature, respectively, which are used to determine sensible-heat flux;
 a krypton hygrometer to measure instantaneous changes in air-vapor density in combination with changes in vertical wind speed, which are used to determine latent-heat flux;
 a net radiometer to measure net radiation;
 a set of soil-heat flux plates and thermocouples to compute subsurface-heat flux; and
 a solid-state temperature and relative humidity probe to measure air temperature and relative humidity, respectively, at one height (fig. 4B).

Data were stored on data loggers and were retrieved during site visits.

Previous Section: General Description and Setting
Next Section: Evapotranspiration from Habitats
Table of Contents |  Home Page