David I. Stannard

Comparison of Penman-Monteith, Shuttleworth-Wallace, and modified Priestley-Taylor evapotranspiration models for wildland vegetation in semiarid rangeland

Eddy correlation measurements of sensible and latent heat flux are used with measurements of net radiation, soil heat flux, and other micrometeorological variables to develop the Penman-Monteith, Shuttleworth-Wallace, and modified Priestley-Taylor evapotranspiration models for use in a sparsely vegetated, semiarid rangeland. The Penman-Monteith model, a one-component model designed for use with dense crops, is not sufficiently accurate (r2 = 0.56 for hourly data and r2 = 0.60 for daily data). The Shuttleworth-Wallace model, a two-component logical extension of the Penman-Monteith model for use with sparse crops, performs significantly better (r2 = 0.78 for hourly data and r2 = 0.85 for daily data). The modified Priestley-Taylor model, a one-component simplified form of the Penman potential evapotranspiration model, surprisingly performs as well as the Shuttle worth-Wallace model. The rigorous Shuttleworth-Wallace model predicts that about one quarter of the vapor flux to the atmosphere is from bare-soil evaporation. Further, during daylight hours, the small leaves are sinks for sensible heat produced at the hot soil surface.

An interdisciplinary field study of the energy and water fluxes in the atmosphere−biosphere system over semiarid rangelands : description and some preliminary results

Abstract Arid and semiarid rangelands comprise a significant portion of the earths land surface. Yet little is known about the effects of temporal and spatial changes in surface soil moisture on the hydrologic cycle, energy balance, and the feedbacks to the atmosphere via thermal forcing over such environments. Understanding this interrelationship is crucial for evaluating the role of the hydrologic cycle in surface-atmosphere interactions. This study focuses on the utility of remote sensing to provide measurements of surface soil moisture, surface albedo, vegetation biomass, and temperature at different spatial and temporal scales. Remote-sensing measurements may provide the only practical means of estimating some of the more important factors controlling land surface processes over large areas. Consequently, the use of remotely sensed information in biophysical and geophysical models greatly enhances their ability to compute fluxes at catchment and regional scales on a routine basis. However, model cal...

Use of ground‐based remotely sensed data for surface energy balance evaluation of a semiarid rangeland

An interdisciplinary field experiment was conducted to study the water and energy balance of a semiarid rangeland watershed in southeast Arizona during the summer of 1990. Two subwatersheds, one grass dominated and the other shrub dominated, were selected for intensive study with ground-based remote sensing systems and hydrometeorological instrumentation. Surface energy balance was evaluated at both sites using direct and indirect measurements of the turbulent fluxes (eddy correlation, variance, and Bowen ratio methods) and using an aerodynamic approach based on remote measurements of surface reflectance and temperature and conventional meteorological information. Estimates of net radiant flux density (Rn), derived from measurements of air temperature, incoming solar radiation, and surface temperature and radiance compared well with values measured using a net radiometer (mean absolute difference (MAD) ≃ 50 W/m2 over a range from 115 to 670 W/m2). Soil heat flux density (G) was estimated using a relation between G/Rn and a spectral vegetation index computed from the red and near-infrared surface reflectance. These G estimates compared well with conventional measurements of G using buried soil heat flux plates (MAD ≃ 20 W/m2 over a range from −13 to 213 W/m2). In order to account for the effects of sparse vegetation, semiempirical adjustments to the single-layer bulk aerodynamic resistance approach were required for evaluation of sensible heat flux density (H). This yielded differences between measurements and remote estimates of H of approximately 33 W/m2 over a range from 13 to 303 W/m2. The resulting estimates of latent heat flux density, LE, were of the same magnitude and trend as measured values; however, a significant scatter was still observed: MAD ≃ 40 W/m2 over a range from 0 to 340 W/m2. Because LE was solved as a residual, there was a cumulative effect of errors associated with remote estimates of Rn, G, and H.

COMPARISON OF 13 EQUATIONS FOR DETERMINING EVAPOTRANSPIRATION FROM A PRAIRIE WETLAND, COTTONWOOD LAKE AREA, NORTH DAKOTA, USA

Evapotranspiration determined using the energy-budget method at a semi-permanent prairie-pothole wetland in east-central North Dakota, USA was compared with 12 other commonly used methods. The Priestley-Taylor and deBruin-Keijman methods compared best with the energy-budget values; mean differences were less than 0.1 mm d−1, and standard deviations were less than 0.3 mm d−1. Both methods require measurement of air temperature, net radiation, and heat storage in the wetland water. The Penman, Jensen-Haise, and Brutsaert-Stricker methods provided the next-best values for evapotranspiration relative to the energy-budget method. The mass-transfer, deBruin, and Stephens-Stewart methods provided the worst comparisons; the mass-transfer and deBruin comparisons with energy-budget values indicated a large standard deviation, and the deBruin and Stephens-Stewart comparisons indicated a large bias. The Jensen-Haise method proved to be cost effective, providing relatively accurate comparisons with the energy-budget method (mean difference=0.44 mm d−1, standard deviation=0.42 mm d−1) and requiring only measurements of air temperature and solar radiation. The Mather (Thornthwaite) method is the simplest, requiring only measurement of air temperature, and it provided values that compared relatively well with energy-budget values (mean difference=0.47 mm d−1, standard deviation=0.56 mm d−1). Modifications were made to several of the methods to make them more suitable for use in prairie wetlands. The modified Makkink, Jensen-Haise, and Stephens-Stewart methods all provided results that were nearly as close to energy-budget values as were the Priestley-Taylor and deBruin-Keijman methods, and all three of these modified methods only require measurements of air temperature and solar radiation. The modified Hamon method provided values that were within 20 percent of energy-budget values during 95 percent of the comparison periods, and it only requires measurement of air temperature. The mass-transfer coefficient, associated with the commonly used mass-transfer method, varied seasonally, with the largest values occurring during summer.

A theoretically based determination of bowen-ratio fetch requirements

Determination of fetch requirements for accurateBowen-ratio measurements of latent- and sensible-heatfluxes is more involved than for eddy-correlationmeasurements because Bowen-ratio sensors are locatedat two heights, rather than just one. A simplesolution to the diffusion equation is used to derivean expression for Bowen-ratio fetch requirements,downwind of a step change in surface fluxes. Theserequirements are then compared to eddy-correlationfetch requirements based on the same diffusionequation solution. When the eddy-correlation andupper Bowen-ratio sensor heights are equal, and theavailable energy upwind and downwind of the stepchange is constant, the Bowen-ratio method requiresless fetch than does eddy correlation. Differences infetch requirements between the two methods aregreatest over relatively smooth surfaces. Bowen-ratiofetch can be reduced significantly by lowering thelower sensor, as well as the upper sensor. The Bowen-ratio fetchmodel was tested using data from a fieldexperiment where multiple Bowen-ratio systems weredeployed simultaneously at various fetches and heightsabove a field of bermudagrass. Initial comparisonswere poor, but improved greatly when the model wasmodified (and operated numerically) to account for thelarge roughness of the upwind cotton field.

A COMPARISON OF SHORT-TERM MEASUREMENTS OF LAKE EVAPORATION USING EDDY CORRELATION AND ENERGY BUDGET METHODS

Abstract Concurrent short-term measurements of evaporation from a shallow lake, using eddy correlation and energy budget methods, indicate that sensible and latent heat flux between lake and atmosphere, and energy storage in the lake, may vary considerably across the lake. Measuring net radiation with a net radiometer on the lake appeared to be more accurate than measuring incoming radiation nearby and modeling outgoing radiation. Short-term agreement between the two evaporation measurements was obtained by using an energy storage term that was weighted to account for the area-of-influence of the eddy correlation sensors. Relatively short bursts of evaporation were indicated by the eddy correlation sensors shortly after midnight on two of three occasions.

Estimates of fetch-induced errors in Bowen-ratio energy-budget measurements of evapotranspiration from a prairie wetland, Cottonwood Lake Area, North Dakota, USA

Micrometeorological measurements of evapotranspiration (ET) often are affected to some degree by errors arising from limited fetch. A recently developed model was used to estimate fetch-induced errors in Bowen-ratio energy-budget measurements of ET made at a small wetland with fetch-to-height ratios ranging from 34 to 49. Estimated errors were small, averaging −1.90%±0.59%. The small errors are attributed primarily to the near-zero lower sensor height, and the negative bias reflects the greater Bowen ratios of the drier surrounding upland. Some of the variables and parameters affecting the error were not measured, but instead are estimated. A sensitivity analysis indicates that the uncertainty arising from these estimates is small. In general, fetch-induced error in measured wetland ET increases with decreasing fetch-to-height ratio, with increasing aridity and with increasing atmospheric stability over the wetland. Occurrence of standing water at a site is likely to increase the appropriate time step of data integration, for a given level of accuracy. Occurrence of extensive open water can increase accuracy or decrease the required fetch by allowing the lower sensor to be placed at the water surface. If fetch is highly variable and fetch-induced errors are significant, the variables affecting fetch (e.g., wind direction, water level) need to be measured. Fetch-induced error during the non-growing season may be greater or smaller than during the growing season, depending on how seasonal changes affect both the wetland and upland at a site.

Tensiometers: Theory, construction, and use

Standard tensiometers are used to measure matric potential as low as -870 cm of water in the unsaturated zone by creating a saturated hydraulic link between the soil water and a pressure sensor. The direction and, in somecases, quantity of water flux can be determined using multiple installations. A variety of commercial and fabricated tensiometers are commonly used. Saturated porous ceramic materials, which form an interface between the soil water and the bulk water inside the instrument, are available in many shapes, sizes, and pore diameters. A gage, manometer, or electronic pressure transducer is connected to the porous material with small- or large-diameter tubing. Selection of these components allows the user to optimize one or more characteristics, such as accuracy, versatility, response time, durability, maintenance, extent of data collection, and cost. Special designs have extended the normal capabilities of tensiometers, allowing measurement in cold or remote areas, measurement of matric potential as low as -153 m of water (-15 bars), measurement at depths as deep as 6 m (recorded at land surface), and automatic measurement using as many as 22 tensiometers connected to a single pressure transducer. Continuous hydraulic connection between the porous material and soil, and minimal disturbance of the natural infiltration pattern are necessary for successful installation. Avoidance of errors caused by air invasion, nonequilibrium of the instrument, or pressure-sensor inaccuracy will produce reliable values of matric potential, a first step in characterizing unsaturated flow.