Water, energy, and food are linked in intricate ways in irrigated agriculture and understanding the interplay of these components is crucial for sustainable and profitable crop production, particularly in smallholder setting such as in sub-Saharan Africa. This study evaluates water-energy-food linkages, engineering and economic performance, irrigation decision making, and challenges faced around water management in a community-based mechanized irrigation scheme in Rwanda. The research is the first to analyze such as scheme, which uses technology typically used by large farmers in a smallholder setting. The study investigates the variation in water requirements and the relationship and impacts of this variability on crop yield for the crops grown in the scheme: maize, French beans, and dry beans. Observed irrigation decision-making analyses demonstrate a lack of irrigation planning during growth stages and significant field-to-field variation in irrigation; this is linked to yield reduction in major crops. Results suggest that farmers irrigated only 31% of modeled irrigation water in dry beans and 27% of modeled irrigation water in maize. An econometric model assessment is used to understand the relationship between yield and energy inputs. A related policy analysis considers the impacts of changes in crop and water management on field-level profits and system-level financial sustainability. This study has implications for understanding irrigation policies in the context of the water-energy-food nexus and decision-making in SubSaharan Africa.

Irrigation is a policy focus in Sub-Saharan Africa and is viewed as an important mechanism to improve farmers’ income and livelihoods while reducing the impacts of climate change. Water, energy, and food are linked in intricate ways in irrigated agriculture and understanding the interplay of these components is crucial for sustainable and profitable crop production. Although studies have been conducted in different parts of the world to understand water and energy use at a field scale under large irrigation systems, little is known about linkages under farmer-managed mechanized irrigated schemes in Sub-Saharan Africa. This study evaluates water-energy-food linkages, engineering and economic performance, irrigation decision making, and challenges faced around water management in a community-based mechanized irrigation scheme. The research synthesizes intraseasonal water and energy use field data for selected crops in a shared center-pivot irrigation scheme in Rwanda. The major cultivated crops are maize, French beans, and dry beans. A daily soil-water balance is used to estimate irrigation water requirements (IWR) and is simulated in FAO-CROPWAT 8.0. The study investigates the variation in water requirements and the relationship and impacts of this variability on crop yield. Assessment of irrigation performance is done by estimating and comparing crop water productivity (CWP) with global and local averages. Observed irrigation decision-making analyses demonstrate a lack of irrigation planning during growth stages and significant field-to-field variation in irrigation; this is linked to yield reduction in major crops. An econometric model assessment is used to understand the relationship between yield and energy inputs. A related policy analysis considers the impacts of changes in crop and water management on field-level profits and system-level financial sustainability. This study has implications for understanding irrigation policies in the context of the water-energy-food nexus and decision-making in Sub-Saharan Africa.

Effective agricultural water management requires accurate, continuous, and transparent accounting of water use in irrigated agroecosystems, especially in water-limited regions where moratoriums may be imposed. Advances in sensor technologies, networking, and data analytics can aid in fulfilling this task by automatically collecting, analyzing and reporting real-time data to infer irrigators' practices and behaviors, crop water requirements, water applications and use. In this research, an automated irrigation water withdrawal and water use monitoring and data collection system was deployed to monitor actual irrigation dynamics for 31 commercial and large-scale agricultural production fields for three consecutive years. Production scale fields (representing a total of 1050 ha) included center pivot (P), gravity (surface/furrow) (G) and subsurface drip-irrigated (S) fields. On average, irrigation was initiated 40-70 days after planting (DAP) and terminated by 120-140 DAP. The proportion of irrigation systems operating simultaneously and peak water abstraction were highest (70-90 %) during July and August. Mean depth of water applied across all fields was 243, 264 and 284 mm in 2013, 2014 and 2015, respectively. Site-specific monitoring of precipitation, soil moisture, and evaporative demand and a soil-water balance model resulted in mean seasonal irrigation requirement estimates of 394, 242 and 184 mm, in 2013, 2014 and 2015, respectively; for maize; and 307, 163 and 219 mm in 2013, 2014 and 2015, respectively, for soybean. Despite reduction of calculated mean irrigation requirement in 2014 and 2015 by 40-50 %, actual irrigation applications by producers did not change considerably and irrigation applied exceeded irrigation water requirements in 80 % of the fields, suggesting needs for irrigation water management technology implementation and associated educational programs in the region. Some fields showed irrigation applications exceeding the mean annual allocated (moratorium) irrigation depth (305 mm), implying that irrigation decisions are still largely driven by non-scientific and/or technical methods. While substantial farm-to-farm heterogeneity makes it challenging to robustly benchmarking regional water footprint and irrigator behavior, it also creates an opportunity for developing and implementing methodologies and strategies for real-time monitoring of farm-level irrigation dynamics. New advances in technologies with telemetry capabilities as well as internet of things (IoTs) can be leveraged to effectively create databases, track and compare water usage to better plan, allocate, distribute, monitor and manage limited water resources for enhancing agricultural productivity. These processes can also be used for education and demonstration for irrigation professionals to enhance adoption of such technologies. This research has the potential for technology and strategy transfer for advanced water management to other regions in the United States and globally.

Temperature-based grass-reference evapotranspiration (ETo) estimation methods (e.g., Hargreaves-Samani [HS] model) present advantages over combination-based methods that require full-suite weather data. The U.S. High Plains region has scarce and short-term full-suite weather sites. This data scarcity presents challenges for combination-based ETo estimation. The performance of HS model against the American Society of Civil Engineers (ASCE) standardized Penman-Monteith (PM) model was assessed using long-term data at 124 full-suite weather sites across nine states in the U.S. High Plains. The HS model underestimated ETo at arid (mean bias error [MBE] = -1.68 mm d(-1)), semi-arid (MBE = -0.34 mm d(-1)), and dry subhumid sites (MBE = -0.16 mm d(-1)) and overestimated ETo at humid sites (MBE = 0.14 mm d(-1)). There was a significant relationship (p < .01) between HS model performance and aridity index. The HS model performed better (27% lower root mean squared difference [RMSD]) in summer months than the rest of the year at semi-arid and dry subhumid sites. The model performance was non-ideal during the summer months in subhumid climates. Spatio-temporal annual zonal (climate division), monthly zonal, annual site-specific, and monthly site-specific calibration resulted in 12, 16, 20, and 26% reduction in RMSD and 11, 16, 17, and 23% reduction in relative error, respectively. Monthly site-specific calibration performed the best and was used to quantify annual and growing season ETo across the region. The research characterized performance patterns of the HS model over an important agroecosystem-dominated region. Practical data-driven strategies were proposed to better estimate PM ETo using limited weather data at any given site (with similar aridity) and time of the year.

There were inadvertent errors in the authorship line and text. The authorship line should read as: Sumantra Chatterjee, Suat Irmak, Jose O. Payero, Ayse Kilic, Lameck O. Odhiambo, Daran Rudnick, Vivek Sharma, and David Billesbach.

Surface energy balance components, including actual evapotranspiration (ET), were measured in a reducedtill maize-soybean field in south central Nebraska during three consecutive non-growing seasons (2006/2007, 2007/2008, and 2008/2009). The relative fractions of the energy balance components were compared across the non-growing seasons, and surface coefficients (K-c) were determined as a ratio of measured ET to estimated alfalfa (ETr) and grass (ETo) reference ET (ETref). The non-growing season following a maize crop had 25% to 35% more field surface covered with crop residue as compared to the non-growing seasons following soybean crops. Net radiation (R-n) was the dominant surface energy balance component, and its partitioning as latent heat (LE), sensible heat (H), and soil heat (G) fluxes depended on field surface and atmospheric conditions. No significant differences in magnitude, trend, and distribution of the surface energy balance components were observed between the seasons with maize or soybean surface residue cover. The cumulative ET was 196, 221, and 226 mm during the three consecutive non-growing seasons. Compared to ETref, the cumulative total measured ET was 61%, 63%, and 59% of cumulative total ETo and 43%, 46%, and 41% of cumulative total ETr during the three consecutive seasons. The type of residue on the field surface had no significant effect on the magnitude of ET. Thus, ET was primarily driven by atmospheric conditions rather than surface characteristics. The coefficient of determination (R-2) for the daily ET vs. ETr data during the three consecutive non-growing seasons was only 0.23, 0.42, and 0.42, and R-2 for ET vs. ETo was 0.29, 0.46, and 0.45, respectively. Daily and monthly average K-c values varied substantially from day to day and from month to month, and exhibited interannual variability as well. Thus, no single K-c value can be used as a good representation of the surface coefficient for accurate prediction of ET for part or all of the non-growing season. A good relationship was observed between monthly total measured ET vs. monthly total ETref. The R-2 values for monthly total ET vs. monthly total ETref data ranged from 0.71 to 0.89 for both ETr and ETo. Using pooled data for monthly total ET vs. monthly total ETref, R-2 was 0.78 for ETr and 0.80 for ETo. The slopes (S) of the best-fit line with intercept for the monthly total ET vs. monthly total ETref data were consistent for all three non-growing seasons, with S = 0.45 +/- 0.05 for ETr and S = 0.62 +/- 0.08 for ETo. The parity in R-2 and S across the three non-growing seasons suggests that the same regression equation can be used to approximate non-growing season ET for field surfaces with both maize and soybean crop residue covers. Considering the extreme difficulties in measuring ET during winter in cold and windy climates with frozen and/or snow-covered conditions, the approach using a linear relationship between monthly total ET vs. monthly total ETref appears to be a good alternative to using a surface coefficient to approximate non-growing season monthly total ET.The conclusions of this research are based on the typical dormant season conditions observed at the research location and may not be generally transferable to other locations with different climatic and surface conditions.

This research evaluated the relative evaporative losses and water balance components in two soybean [Glycine max (L.) Merr.] fields under subsurface drip irrigation (SDI) and center pivot irrigation (CPI) systems in south-central Nebraska. Meteorological and surface energy balance components, including actual evapotranspiration (ET), above the crop canopy was measured using Bowen ratio energy balance systems installed at the center of both fields. Crop transpiration (T) was estimated based on the variable stomatal resistances using the Penman-Monteith equation in conjunction with fractional green canopy cover. Evaporation (E) losses were estimated as the difference between measured ET and estimated T. Average soil water content (ASWC) in the crop root zone and effective rainfall were estimated using the water balance method. The relative evapotranspiration (ETrel) was 99.8% for the SDI field in both years (2007 and 2008), and it was 103.4% in 2008 and 93.9% in 2010 for the CPI field. The mean ETrel between sites and across years were not significantly different (P > 0.05). There were revealing differences in the relative contribution of T and E to total seasonal ET between the SDI and CPI fields. In the SDI field, T and E were 88 and 12% of ET, respectively, in 2007, and in 2008 they were 84 and 16% of ET, respectively. In the CPI field, T and E were 78 and 22% of ET, respectively, in 2008, and in 2010 they were 75 and 25% of ET, respectively. At full canopy cover, T contributed more than 85% of the total ET. Although E is assumed negligible at the full canopy cover stages, the data indicated that E was 6 and 9% of ET in SDI fields during the midseason growth stage at full canopy cover in 2007 and 2008, respectively, and 14 and 21% of ET in CPI fields during the same growth stage in 2008 and 2010, respectively. On a 2-year average, the CPI field had approximately 10% higher E losses as compared with the SDI field under these experimental conditions, which can have a large impact on crop water productivity when crop production under water-limiting conditions is considered. The reduction in E in the SDI field can be expected to be greater in arid and semiarid regions than in subhumid regions. (C) 2015 American Society of Civil Engineers.

Soybean [Glycine max (L) Merr.] yield, irrigation water use efficiency (IWUE), crop water use efficiency (CWUE), evapotranspiration water use efficiency (ETWUE), and soil water extraction response to eleven treatments of full, limited, or delayed irrigation versus a rainfed control were investigated using a subsurface drip irrigation (SDI) system at a research site in south-central Nebraska. The SDI system laterals were 0.40 m deep in every other row middle of 0.76 m spaced plant rows. Actual evapotranspiration (ETa) was quantified in all treatments and used to schedule irrigation events on a 100% ETa replacement basis in all but three of the eleven treatments (i.e., 75% ETa replacement was used in two, and 60% ETa replacement was used in one). The irrigation amount (I-a) applied at each event was 100% of the ETa amount, except for two 100% ETa treatments in which only 65% or 50% of the water needed to cover the treatment plot area was applied to enable a test of a partial surface area-based irrigation approach. The first irrigation event was delayed until soybean stage R3 (begin pod) in two 100% I-a treatments, but thereafter they were. irrigated with either 100% or 75% ETa replacement. Two 100% ETa and 100% I-a treatments also were used to evaluate soybean response to nitrogen (N) application methods (i.e., a preplant method versus N injection using the SDI system). Soybean ETa varied from 452 mm for the rainfed treatment to 600 mm (30% greater) for the fully irrigated treatment (100% ETa and 100% I-a) in 2007, and from 473 to 579 mm (20% greater) for the same treatments, respectively, in 2008. Among the irrigated treatments, 100% ET and 65% Ia had the lowest 2007 ETa value (557 mm), whereas 100% ETa and 50% I-a had the lowest 2008 ETa (498 mm). The 100%, 75%, and 60% ETa treatments with 100% I-a had respective actual ETa values that declined linearly in 2008 (i.e., 579, 538, and 498 mm), but not in 2007. Seasonal totals for ETa versus I-a exhibited a linear relationship (R-2 = 0.68 in 2007 and R-2 = 0.67 in 2008). Irrigation enhanced soybean yields from rainfed yield baselines of 4.04 ton ha(-1) in 2007 and 4.82 ton ha(-1) in 2008) to a maximum of 4.94 ton ha(-1) attained in 2007 with the delay to R3 irrigation treatment (its yield was significantly greater, p < 0.05, than that of the seven other treatments) and 4.97 ton ha(-1) attained in 2008 with the 100% ETa and 100% I-a preplant N treatment. Seed yield had a quadratic relationship with irrigation water applied and a linear relationship with ETa that was stronger in the drier year of 2007. Each 25.4 mm incremental increase in seasonal irrigation water applied increased soybean yield by 0.323 ton ha(-1) (beyond the intercept) in 2007 and by 0.037 ton ha(-1) in 2008. Each 25.4 mm increase in ETa generated a yield increase of 0.114 ton ha(-1) (beyond the intercept) in 2007, but only 0.02 ton ha(-1) in the wetter year of 2008.This research demonstrated that delaying the onset of irrigation until the R3 stage and practicing full irrigation thereafter for soybean grown on silt loam soils resulted in yields (and crop water productivity) that were similar to full-season irrigation scheduling strategies, and this result may be applicable in other regions with edaphic and climatic characteristics similar to those in south-central Nebraska.

Abstract. Hourly evapotranspiration (ET) crop coefficients (K c ) are needed to optimize the effectiveness and efficiency of high-frequency micro- and sprinkler irrigation practices involving the application of water multiple times a day. However, not much is known about the daily and seasonal patterns and magnitudes in hourly K c values for soybean. In addition, locally developed K c values are necessary for more robust within-season irrigation management, crop ET estimation, and water balance analyses. Hourly and daily K c functions were developed for soybean in south-central Nebraska through extensive field research. Actual crop evapotranspiration (ET a ) was measured using a Bowen ratio energy balance system. Daily crop coefficients were calculated as K c = ET a /ET ref , wherein reference (potential) evapotranspiration (ET ref ) was calculated using the Penman-Monteith equation with a fixed canopy resistance for both alfalfa-reference (ET r ) and grass-reference (ET o ) surfaces. The K c values were derived in two forms: (1) a single (normal or average K c ) K cr based on ET r , and K co based on ET o ; and (2) a basal coefficient (K cbr ) based on ET r , and K cbo based on ET o . The seasonal patterns of variation of K cr , K co , K cbr , and K cbo were examined on five different temporal base scales: days after emergence (DAE), cumulative growing degree days (GDD), leaf area index (LAI), fractional green canopy groundcover (CC), and plant phenology (V and R stages). The 2007 and 2008 growing season ET a totals were 535 and 514 mm, respectively. Extreme hourly K c values were frequently observed in the early morning and late afternoon hours when ET a was very low relative to ET r and ET o . Daily means of the 10 to 13 hourly values computed for K cr ranged from 0.25 to 1.06 in 2007 and from 0.15 to 1.02 in 2008, whereas those computed for K co ranged from 0.39 to 1.37 in 2007 and from 0.22 to 1.29 in 2008. Daily K cr and K co values calculated based on daily data ranged from 0.20 to 1.12 and from 0.27 to 1.47, respectively. Comparison of all daily means of hourly coefficients with the corresponding daily coefficients in one-to-one graphs and zero-origin based regression of the former on the latter revealed linear regression coefficients of 0.92 (2007 K cr ), 0.95 (2008 K cr ), 0.96 (2007 K co ), and 0.97 (2008 K co ), with R 2 values of 0.78 or better. On average, hourly K c values were about 4% to 8% lower that the corresponding daily values. Substantial diurnal variability was observed in K co and K cr measured during daylight hours (ranging from 0.1-0.2 to 1.5-1.6) from early morning to late afternoon (8:00 to 18:00), and the range of variability was substantially dependent on the coincident V and R stages. The relationship between K c and LAI was best represented by two regression trend lines: one representing crop development from its beginning up to the start of senescence, and the other representing crop development thereafter. A similar break in the regression trend line was observed in the relationship between basal K c and GDD. In contrast, the relationship between K c and fractional CC was not biphasic and could be modeled with one regression trend line. The FAO-56 tabulated K co values and those measured in this research were significantly different (p a and crop water requirement estimates. Because this research proved that K co and K cr values are not constant during the day from dawn to dusk, using daily average K co or K cr values would not be able to provide robust and precise determination of crop irrigation requirements for irrigation practices delivered more than once per day. The crop coefficients developed in this research as a function of several base scales should provide crop consultants, extension service personnel, agronomists, irrigation practitioners, and other irrigation and water management professionals with robust and accurate methods for choosing and applying crop coefficients to be used for more precise determination of ET a and water requirements, thus leading to more efficient and effective seasonal soybean irrigation management.

Single and dual crop coefficient methods are used in conjunction with grass reference evapotranspiration (ETo) to estimate actual crop evapotranspiration (ETc). However, the impact of soil surface residue cover on the accuracy of ETc estimated with these methods is not well understood. The objective of this study is to evaluate and compare the accuracy of the FAO-56 single crop coefficient (single-Kc) and dual crop coefficient (dual-Kc) methods for estimating soybean [Glycine max (L.) Merr.] ETc in a partially residue covered field. The study was conducted at the University of Nebraska-Lincoln, South Central Agricultural Laboratory (SCAL), Nebraska, during the 2007 and 2008 growing seasons. The field was under reduced-tillage (ridge till) on a silt loam soil and irrigated using a subsurface drip irrigation system. Evapotranspiration flux (ETm) above the crop canopy was measured using a deluxe version of a Bowen ratio energy balance system (BREBS) and ETo was calculated with the Penman–Monteith method. The single-Kc and dual-Kc-estimated ETc values, both unadjusted for residue cover, were compared to ETm. The unadjusted FAO-56 Kc values performed poorly as the single-Kc underestimated ETm during the initial crop growth stage by 21% in 2007 and 33.6% in 2008 while the dual-Kc overestimated ETm during the same growth stage by 16.8% in 2007 and 16.5% in 2008. Extended simulations were conducted to determine the magnitude by which ETc is reduced for each 10% of soil surface covered with crop residue. Downward adjustments in soil water evaporation (Es) for every 10% of the soil surface covered with crop residue improved the accuracy of ETc estimated by the dual-Kc method. The largest changes in ETc due to adjustments in Es occur during the initial stage of the growing season. The best estimates for seasonal ETc were obtained by reducing Es by 5% for every 10% of surface covered with residue in 2007 (R2 = 0.77, RMSD = 0.87 mm d−1, E = 0.94) and 2008 (R2 = 0.83, RMSD = 0.84 mm d−1, E = 0.95). Greater improvements in the accuracy of estimated seasonal ETc were obtained by reducing Es by 2.5% for each 10% of surface covered with residue during the initial stage and by 5% during the rest of the crop growth stage. These results suggest that the more computationally-involved dual-Kc method with adjustments in Es for each 10% of surface covered with residue improves the prediction of ETc in fields with soil surface residue cover, especially during the initial growth stage. Inaccurate selection of percentage reduction in Es can result in substantial overestimation or underestimation of seasonal ETc by the dual-Kc method.

Estimation of actual evapotranspiration (ET), especially its partitioning into plant transpiration (T) and soil evaporation (E), in agricultural fields is important for effective soil water management and conservation and for understanding the interactions between ET, T, and E with the management practices. Direct field measurements of ET, T, and E rates are difficult and costly; hence, mathematical models are used for estimating them. The objective of this study was to evaluate the practical applicability of the Shuttleworth-Wallace (S-W) model to estimate and partition ET in a subsurface drip-irrigated soybean (Glycine max L. Merr) field with partial residue cover. While its performance has been studied for various surfaces, the performance evaluation of the S-W model for such surface has not been carried out. An integrated approach of calculating bulk stomatal resistance (r(s)(c)) as a function of soil water content (theta(i)) was incorporated into the model to allow simulation of T over a range of 01, and a residue decomposition function was introduced to account for surface residue decay over time to more accurately account for the actual residue cover in field conditions. The model performance was evaluated for different plant growth stages during the 2007 and 2008 growing seasons at the University of Nebraska-Lincoln, South Central Agricultural Laboratory near Clay Center, Nebraska. The sum of estimated T and E was compared to the Bowen Ratio Energy Balance System (BREBS)-measured actual ET on a daily time-step. The model was able to capture the trends and magnitudes of measured ET, but its performance differed for various plant physiological growth stages. The root mean square difference (RMSD) values between the model-estimated and measured ET values for the growing season (day after emergence until physiological maturity) were 1.26 and 1.03 mm d(-1) for 2007 and 2008, respectively. Best performance was observed during the mid-season during full canopy cover with a two-year average r(2) of 0.87, average RMSD of 0.94 mm d(-1), and average mean biased error (MBE) of 0.30 mm d(-1). Estimates for both initial and late season growth stages where E was dominant had the least agreement with BREBS measurements. The proportion of T and E in the estimated ET varied with growth stage. The S-W-estimated seasonal total ET and BREBS measurements were equal in 2007 (S-W model ET = 496 mm and BREBS ET = 498 mm), and in 2008 the model underestimated by only 8.2% (S-W model ET = 452 mm and BREBS ET = 489 mm). While, in general, the model was successful in tracking the trends and magnitude of the BREBS-measured ET, further re-parameterization of the T module of the model can improve its accuracy to estimate ET, especially T, during the initial and late season (before full canopy cover and after physiological maturity) for a subsurface drip-irrigated soybean canopy. Other enhancements needed in the model for improved estimation of the E component include accurate determination of soil surface resistance coefficients and accounting for direct evaporation of intercepted rainfall on the canopy.

Estimation of actual evapotranspiration (ET), especially its partitioning into plant transpiration (T) and soil evaporation (E), in agricultural fields is important for effective soil water management and conservation and for understanding the interactions between ET, T, and E with the management practices. Direct field measurements of ET, T, and E rates are difficult and costly; hence, mathematical models are used for estimating them. The objective of this study was to evaluate the practical applicability of the Shuttleworth-Wallace (S-W) model to estimate and partition ET in a subsurface drip-irrigated soybean (Glycine max L. Merr.) field with partial residue cover. While its performance has been studied for various surfaces, the performance evaluation of the S-W model for such surface has not been carried out. An integrated approach of calculating bulk stomatal resistance (r s c ) as a function of soil water content (θ i ) was incorporated into the model to allow simulation of T over a range of θ i , and a residue decomposition function was introduced to account for surface residue decay over time to more accurately account for the actual residue cover in field conditions. The model performance was evaluated for different plant growth stages during the 2007 and 2008 growing seasons at the University of Nebraska-Lincoln, South Central Agricultural Laboratory near Clay Center, Nebraska. The sum of estimated T and E was compared to the Bowen Ratio Energy Balance System (BREBS)-measured actual ET on a daily time-step. The model was able to capture the trends and magnitudes of measured ET, but its performance differed for various plant physiological growth stages. The root mean square difference (RMSD) values between the model-estimated and measured ET values for the growing season (day after emergence until physiological maturity) were 1.26 and 1.03 mm d -1 for 2007 and 2008, respectively. Best performance was observed during the mid-season during full canopy cover with a two-year average r 2 of 0.87, average RMSD of 0.94 mm d -1 , and average mean biased error (MBE) of 0.30 mm d -1 . Estimates for both initial and late season growth stages where E was dominant had the least agreement with BREBS measurements. The proportion of T and E in the estimated ET varied with growth stage. The S-W-estimated seasonal total ET and BREBS measurements were equal in 2007 (S-W model ET = 496 mm and BREBS ET = 498 mm), and in 2008 the model underestimated by only 8.2% (S-W model ET = 452 mm and BREBS ET = 489 mm). While, in general, the model was successful in tracking the trends and magnitude of the BREBS-measured ET, further re-parameterization of the T module of the model can improve its accuracy to estimate ET, especially T, during the initial and late season (before full canopy cover and after physiological maturity) for a subsurface drip-irrigated soybean canopy. Other enhancements needed in the model for improved estimation of the E component include accurate determination of soil surface resistance coefficients and accounting for direct evaporation of intercepted rainfall on the canopy.

Net radiation (R-n) is the main driving force of evapotranspiration (ET) and is a key input variable to the Penman-type combination and energy balance equations. However, R-n is not commonly measured. This paper analyzes the impact of 19 net radiation models that differ in model structure and intricacy on estimated grass and alfalfa-reference ET (ETo and ETr, respectively) and investigates how climate, season and cloud cover influence the impact of the R-n models on ETo and ETr. Datasets from two locations (Clay Center, Nebraska, subhumid; and Davis, California, a Mediterranean-type semiarid climate) were used. R-n values computed from the 19 models were used in the standardized ASCE-EWRI Penman-Monteith equation to estimate ETo and ETr on a daily time step. The influence of seasons on the estimation of R-n and on estimated ETo and ETr was investigated in winter (November-March) and summer (May-September) months. To analyze the influence of clouds on the impact of R-n models, relative shortwave radiation (R-rs) was used as a means to express the cloudiness of the days as: 0 <= R-rs <= 0.35 for completely cloudy days; 0.35 <= R-rs <= 0.70 for partially cloudy days; and 0.70 <= R-rs <= 1.0 for clear sky days. The performances of R-n models showed variations at the same location and between the locations for the same model based on methods used to calculate various model parameters. The most significant impact of R-n on estimated ETo and ETr was related to the methods used to calculate atmospheric emissivity (epsilon) rather than methods used to calculate clear sky solar radiation (R-so) or cloud adjustment factor (f). R-n models that used average air temperature to compute epsilon and an estimated f resulted in good performances at both locations. Empirical models that assumed f = 1.0 showed poor to average performances at both locations. While model performances varied based on methods used to calculate R-so, f, and epsilon, there were significant seasonal variations in performances of models that calculated epsilon as a function of actual vapor pressure of the air (e(a)). The seasonal variations in performances of these models were greater under subhumid climate at Clay Center than in semiarid climate at Davis, Calif. The models that calculated epsilon as a function of e(a) performed better under completely cloudy days than on other days, more so at Clay Center. Methods used to calculate epsilon have a significant impact on the R-n model performance, especially in unstable climatic conditions such as at Clay Center where there are frequent and rapid changes in climatic variables in a given day and throughout the year. The results of this study can be used as a reference tool to provide practical information on which method to select based on the data availability for reliable estimates of daily R-n relative to the ASCE-EWRI R-n method in subhumid and semiarid climates similar to Clay Center, Neb. and Davis, Calif.

Estimation of reference evapotranspiration (ETref) using measured microclimatic data and the Penman-Monteith (PM) method provides a powerful means of quantifying actual plant evapotranspiration (ETa) needed for use in various disciplines. When applying the PM method to estimate ETref, it is desirable to measure the required microclimatic data over a reference grass or alfalfa surface rather than above non-reference surfaces. However, in reality, establishing and maintaining a reference surface for long periods of time is a difficult task. Other surface energy balance systems, such as the Bowen ratio energy balance system (BREBS), eddy covariance system, and surface renewal, are increasingly used to measure surface energy fluxes along with the microclimatic data above various plant canopies. These systems could be another source of data for ETref estimations when reference weather station data are not available due to logistical difficulties associated with establishing and maintaining a separate reference weather station. In many cases, data measured above other vegetation surfaces using the surface energy balance systems are the only source of data for ETref and ETa estimations due to the absence of reference weather stations. There is little information on how microclimatic data measured above different plant canopies impact the calculated ETref if used in the PM method in place of data collected from a reference surface. This study compares data measured above grass and maize (Zea mays L.) canopies and assesses how the variables measured above two canopies impact ETref calculated using the ASCE standardized Penman-Monteith (ASCE-EWRI PM) equation. Two years (2005 and 2006) of hourly microclimatic data measured above a grass surface using an automated weather station and above a maize canopy using BREBS installed on a well-watered maize field were used. The results obtained indicate very good agreements between the microclimatic variables measured above grass and maize, and between ETref calculated with data measured above the two surfaces. The measured rainfall was the same for both sites (316 and 323 mm in 2005 for the weather station and BREBS, respectively, and 368 and 366 mm in 2006). The main difference between the two surfaces was in wind speed (u(2)) and aerodynamic resistance (r(a)). On a seasonal average basis, u(2) was 15% and 20% higher over the grass canopy than the maize canopy for 2005 and 2006, respectively. The maximum difference in r(a) between the two surfaces occurred when the maize was at its maximum height (2.45 m). On a seasonal average, the r(a) above the maize canopy was 37 s m(-1) higher than the ra above the grass surface. However, the impact of u(2) and r(a) on EZ(ref) was insignificant. The grass and alfalfa-reference ET (ETo and ETr) estimated using the data measured above maize (ETo-maize and ETr-maize) and above grass (ETo-grass and ETr-grass) were very similar in both years. In 2005, ETo-maize (816 mm) and ETo-grass (824 mm) were within 1 %, and ETr-maize (1,033 nun) and ETr-grass (1,070 mm) were within 3%. The same percentages were obtained in 2006 (ETo-maize = 671 mm, ETo-grass = 675 mm, ETr-graze = 838 mm, ETr-grass = 868 mm).Thus, in practice, data measured above a well-watered maize canopy can be a ubstitute for the microclimatic data measured above a reference surface in ETref estimations when "reference" weather station data are riot available to solve the PM equation in areas with similar rainfall (1300 mm) during the growing season, as observed in this study.

Soils can be non-intrusively mapped by observing similar patterns within ground-penetratingradar (GPR) profiles. We observed that the intricate and often indiscernible textural variability found withina complex GPR image possesses important parameters that help delineate regions of similar soilcharacteristics. Therefore, in this study, we examined the feasibility of using textural features extractedfrom GPR data to automate soil characterizations. The textural features were matched to a fingerprintdatabase of previous soil classifications of GPR textural features and the corresponding ground truths ofsoil conditions. Four textural features (energy, contrast, entropy, and homogeneity) were selected forinputs into a neural-network classifier. This classifier was tested and verified using GPR data obtainedfrom two distinctly different field sites. The first data set contained features that indicate the presence orlack of sandstone bedrock in the upper 2 m of a shallow soil profile of fine sandy loan and loam. Thesecond data set contained columnar patterns that correspond to the presence or the lack of verticalpreferential-flow paths within a deep loess soil. The classifier automatically grouped each of these datasets into one of the two categories. Comparing the results of classification using extracted texturalfeatures to the results obtained by visual interpretation found 93.6% of the sections that lack sandstonebedrock correctly classified in the first set of data, and 90% of the sections that contain pronouncedcolumnar patterns correctly classified in the second set of data. The classified profile sections weremapped using integrated GPR and GPS data to show surface boundaries of different soil categories.These results indicate that extracted textural features can be utilized for automatic characterization ofsoils using GPR data.

A unique survey protocol has been developed that maps near-subsurface preferential flow using integrated ground-penetrating radar (GPR) and a differential geographical positioning system (DGPS). The survey protocol consists of a mobile GPR system that spirals outward along a prescribed course, continuously gathering subsurface data for an extended period. Metered water is applied to a centrally located water-ponding ring, after first capturing the initial dry-state pattern signatures. The water radiates outward beneath the surface as it follows preferential flow pathways, which the GPR instrumentation spiraling above highlights. After data are collected, pre- and post-water time-elapsed images profiles are segmented by pattern dissimilarities. The specific locales that exhibit pattern shifts from the initial dry state are identified as dynamic water movement. Locales that exhibit pattern shifts are mapped to indicate the rate and direction of preferential flow about the near surface.

Estimated daily reference crop evapotranspiration (ETo) is normally used to determine the water requirement ofcrops using the crop factor method. Many ETo estimation methods have been developed for different types of climatic data,and the accuracy of these methods varies with climatic conditions. In this study, pair-wise comparisons were made betweendaily ETo estimated from eight different ETo equations and ETo measured by lysimeter to provide information helpful inselecting an appropriate ETo equation for the Cumberland Plateau located in the humid Southeast United States. Based onthe standard error of the estimate (Syx), the relationship between the estimated and measured ETo was the best using theFAO-56 Penman-Monteith equation (coefficient of determination (r2) = 0.91, Syx = 0.31 mm d-1, and a coefficient ofefficiency (E) = 0.87), followed by the Penman (1948) equation (r2 = 0.91, Syx = 0.34 mm d-1, and E = 0.88), and Turcsequation (r2 = 0.90, Syx = 0.36 mm d-1, and E = 0.88). The FAO-24 Penman and Priestly-Taylor methods overestimated ETo,while the Makkink equation underestimated ETo. The results for the Hargreaves-Samani equation showed low correlationwith lysimeter ETo data (r2 = 0.51, Syx = 0.68 mm d-1, and E = 0.20), while those for the Kimberly Penman were reasonable(r2 = 0.87, Syx = 0.40 mm d-1, and E = 0.87). These results support the adoption of the FAO-56 Penman-Monteith equationfor the climatological conditions occurring in the humid Southeast. However, Turcs equation may be an attractive alternativeto the more complex Penman-Monteith method. The Turc method requires fewer input parameters, i.e., mean air temperatureand solar irradiance data only.

Distributed irrigation control (DIC) for site-specific management and/or operation of fixed irrigation systems is easier to install and maintain as compared to centralized irrigation control (CIC), but requires multiple controllers in the field. The advantages of DIC over CIC systems include: (1) reduced wiring and piping requirements, (2) a lower risk of complete system failure due to mechanical damage or lightening strikes, and (3) more flexibility when modifying or extending the system. In this study, a low cost solar-powered feedback controller for DIC of fixed irrigation systems was developed and tested. The specific tasks included the controller design (hardware and software), performance evaluation, and power optimization. The controller uses soil water potential (SWP) measurements to control the amount of water applied to each specific management area of a field, and measured system hydraulic pressure to communicate with other controllers. Each controller is autonomously powered by a solar panel and battery, eliminating hard-wire connections among control units. The results indicate that the controller was effective in maintaining the SWP in the root zone close to a predetermined management allowed deficit (MAD). The power supply was optimized using simulated and measured solar radiation data from two locations.

The effects of some common vapor pressure deficit (VPD) and net irradiance (R-n) calculation methods on the accuracy of ETo values estimated by using the standardized ASCE Penman-Monteith (ASCE-PM) equation for short grass were examined by comparing the estimated ETo values with measured ETo values in a humid climate. Sensitivity analysis showed 17% and 84% change in the estimated daily ET, values per unit change in the calculated VPD and R-n values, respectively. A total of 12 VPD and 27 R-n calculation methods were examined. Analyses of variance indicated lack of equality in the means of estimated ET, values obtained by different VPD and R-n methods. The percent mean error in the estimated ETo values ranged from -0.9 to -8.4% for VPD methods and from -0.3 to -19.7% for R-n methods. On the basis of the coefficient of determination (r(2)) and the standard error of the estimated (S-y/x) values, the VPD calculated from saturation vapor pressure (e(s)), estimated by averaging the e(s) at the maximum and minimum daily air temperatures, and actual vapor pressure (e(a)), estimated by using either the average of minimum and maximum relative humidity or the dew-point temperature, gave more accurate results. Net irradiance (R-n) estimated by using a regression of relative short-wave solar irradiance, as well as a linear regression on the square root of e(a), resulted in relatively more accurate estimates of ETo than that obtained by methods based on e(a) or clear-sky data alone. These results indicate that in a humid climate, some of the VPD and R-n methods have a significant effect on the accuracy of the ETo estimated by using the standardized ASCE-PM equation.

Accurate knowledge of nitrate distribution in the soil under fertigation through drip-irrigation systems is fundamentally important for system design and management. The determination of nitrate distribution through modeling represents a highly complex nonlinear problem that includes adsorption, transformation, convection, and dispersion. For this reason, an alternative methodology is proposed, which combines artificial neural networks (ANN) and laboratory experiments. Seventeen experiments with apparent discharge rates varying from 0.6 to 7.8 l/h, the apparent cylindrical applied volume from 6 to 15 l, and the input concentration from 100 to 700 mg/l were conducted to provide a database for establishing the ANN architecture. The model input parameters were initial soil water content, initial nitrate concentration in the soil, discharge rate, input concentration of fertilizer (NH 4 NO 3 ), applied volume, and final soil water content. The model output was nitrate concentration in the soil after fertigation. A total of 298 vectors were used to train the ANN model, and 212 independent vectors were used to test the model. Results of the test show a good correspondence with a determination coefficient ( r 2 ) of 0.83 between the model-estimated nitrate concentration in the soil and laboratory-measured nitrate concentration in the soil. These results show that the optimized ANN models are reasonably accurate and can provide an easy and efficient means of estimating nitrate distribution in the soil under fertigation through drip-irrigation systems.