Evapotranspiration Algorithm Research Articles

Uncertainty and Sensitivity Analysis of a Remote-Sensing-Based Penman–Monteith Model to Meteorological and Land Surface Input Variables

This study analysed the uncertainty and sensitivity of core and intermediate input variables of a remote-sensing-data-based Penman–Monteith (PM-Mu) evapotranspiration (ET) model. We derived absolute and relative uncertainties of core measured meteorological and remote-sensing-based atmospheric and land surface input variables and parameters of the PM-Mu model. Uncertainties of important intermediate data components (i.e., net radiation and aerodynamic and surface resistances) were also assessed. To estimate the instrument measurement uncertainties of the in situ meteorological input variables, we used the reported accuracies of the manufacturers. Observational accuracies of the remote sensing input variables (land surface temperature (LST), land surface emissivity (εs), leaf area index (LAI), land surface albedo (α)) were derived from peer-reviewed satellite sensor validation reports to compute their uncertainties. The input uncertainties were propagated to the final model’s evapotranspiration estimation uncertainty. Our analysis indicated relatively high uncertainties associated with relative humidity (RH), and hence all the intermediate variables associated with RH, like vapour pressure deficit (VPD) and the surface and aerodynamic resistances. This is in contrast to other studies, which reported LAI uncertainty as the most influential. The semi-arid conditions and seasonality of the regional South African climate and high temporal frequency of the variations in VPD, air and land surface temperatures could explain the uncertainties observed in this study. The results also showed the ET algorithm to be most sensitive to the air-land surface temperature difference. An accurate assessment of those in situ and remotely sensed variables is required to achieve reliable evapotranspiration model estimates in these generally dry regions and climates. A significant advantage of the remote-sensing-based ET method remains its full area coverage in contrast to classic-point (station)-based ET estimates.

Automatically selecting hot and cold pixels for satellite actual evapotranspiration estimation under different topographic and climatic conditions

The persistent monitoring of evapotranspiration (ET) over the regions suffering from water scarcity is critical for sustainable agricultural water management. Remote sensing provides time- and cost-effective capability to investigate daily ET rates at large scales. Satellite-based actual evapotranspiration (ETa) algorithms typically rely on specifying the upper and lower boundaries of ETa rate over agricultural and pasture fields, commonly known as hot (dry) and cold (wet) pixels selection. These boundaries are to be recognized by an expert through a subjective and labor-intensive task. In this study, a method has been introduced to automatically select appropriate anchor pixels (i.e., hot and cold pixels) independent from land use/cover maps with the simplest possible way, quickly applied even by an inexperienced operator. Subsequently, ETa was calculated using Mapping EvapoTranspiration at high Resolution with Internalized Calibration (METRIC), Surface Energy Balance Algorithm for Land (SEBAL), and Surface Energy Balance System (SEBS) algorithms and evaluated against measured data. In this method, the mountains and foothills were removed using the Shuttle Radar Topography Mission Digital Elevation Model (SRTM DEM) and the subsequent product was a slope mask. Then, filters were applied based on the Normalized Difference Vegetation Index (NDVI), Albedo, and Land Surface Temperature (LST) images to identify potential candidate pixels for hot and cold pixels. In the end, the best-conditioned pixel being closest to the meteorological station was selected. The method was assessed in five different regions with different topographic and climatic conditions. The selected pixels were first visually validated in Landsat images, and then the fluctuations and values were discussed in time series of anchor pixels and LST histograms. The visual interpretation was indicative of selecting the anchor pixels in fallow (hot pixel) and densely vegetated (cold pixel) surfaces. Also, the hot and cold pixels were suitably situated in the upper and lower quartiles of the LST histogram, respectively. The range of cold pixels variations throughout the study periods was lower compared with the hot pixels (44.2, 55.3, 35.5, 66.5, and 25.2 K for hot pixels against 34.2, 45.7, 25.6, 52.2, and 17 K for cold pixels) as expected, which emanated from the lower fluctuations of temperature over vegetation against the soil. The results were indicative of the better performance of METRIC compared with SEBAL and SEBS with greater values of R2 in all the regions. Therefore, using the introduced method, the expert subjective interference was eliminated and processing time reduced significantly from about 1 h per image to a few minutes.

Using SMAP Level-4 soil moisture to constrain MOD16 evapotranspiration over the contiguous USA

Evapotranspiration (ET) is a key hydrologic variable linking the Earth's water, carbon and energy cycles. At large spatial scales, remote sensing-based (RS) models are often used to quantify ET. Despite the large number of RS ET models available, few include soil moisture as a key environmental input, which can degrade model accuracy and utility. Here, we use model assimilation enhanced soil moisture estimates from the NASA SMAP (Soil Moisture Active Passive) mission as a water supply control in the MOD16 ET algorithm framework. SMAP-derived daily surface (0-5 cm depth) and root zone (0-1 m depth) soil moisture are used with MODIS (Moderate Resolution Imaging Spectroradiometer) vegetation observations, and 4 km gridded regional surface meteorology (Gridmet) as primary inputs for estimating daily ET and underlying model soil and stomatal conductance terms. We calibrated the model environmental response parameters using tower eddy covariance ET observations representing major North American biomes. The model ET results were validated using a holdout set of tower observations spanning a large regional climate gradient. The updated ET estimates outperform the baseline MOD16 product across all tower validation sites (RMSE = 0.758 vs 1.108 mm day−1; R2 = 0.68 vs 0.45, respectively). Smaller relative improvements were obtained using a recalibrated model with 4 km Gridmet meteorology, but no soil moisture control (RMSE = 0.813 mm day−1; R2 = 0.66), indicating that these changes are essential for the improved model performance. The soil moisture-constrained model improvements and relative benefits from the SMAP observations are greater in arid climates, consistent with stronger soil moisture control on ET in water-limited regions. The use of SMAP soil moisture as an additional model constraint improves MOD16 regional performance and provides a new framework for investigating both soil and atmosphere controls on ET.

Assessing MODIS carbon and water fluxes in grasslands and shrublands in semiarid regions using eddy covariance tower data

ABSTRACT Gross primary productivity (GPP) and evapotranspiration (ET) are two important fluxes between the terrestrial ecosystems and the atmosphere. Remote sensing data-driven models have been successfully used to estimate carbon and water fluxes in various ecosystems, but the models are still underperforming in dryland. In this study, the agreement between the Moderate Resolution Imaging Spectroradiometer (MODIS) products, MODIS data-driven models, and the eddy covariance (EC) tower observation data were tested for two different ecosystem sites in arid regions in Xinjiang, China. The results convincingly indicated that the MODIS products can successfully capture the temporal GPP and ET variables for grasslands and shrublands, but these results have large biases. The MODIS GPP products certified contributed 88% of the EC observed GPP for the grassland but 16% for the shrubland. The temperature and greenness (TG) model clearly showed favourable correspondence with tower GPP observed in arid regions, with the coefficient of determination (R 2) of 0.91 and root mean square error (RMSE) of 17.65 g C m−2 16 days−1. The ET performance was evaluated at two sites, yielding R 2 values of 0.77 and 0.34 for the grassland and shrubland, respectively. Our study showed that the source of uncertainties comes from the remotely sensed data input in GPP and ET algorithms.

Ecohydrological responses to surface flow across borders: Two decades of changes in vegetation greenness and water use in the riparian corridor of the Colorado River delta

AbstractHydrological and bioclimatic processes that lead to drought may stress plants and wildlife, restructure plant community type and architecture, increase monotypic stands and bare soils, facilitate the invasion of non‐native plant species and accelerate soil erosion. Our study focuses on the impact of a paucity of Colorado River surface flows from the United States (U.S.) to Mexico. We measured change in riparian plant greenness and water use over the past two decades using remotely sensed measurements of vegetation index (VI), evapotranspiration (ET) and a new annualized phenology assessment metric (PAM) for ET. We measure these long‐term (2000–2019) metrics and their short‐term (2014–2019) response to an environmental pulse flow in 2014, as prescribed under Minute 319 of the 1944 Water Treaty between the two nations. In subsequent years, small‐directed flows were provided to restoration areas under Minute 323. We use 250 m MODIS and 30 m Landsat imagery to evaluate three vegetation indices (NDVI, EVI, EVI2). We select EVI2 to parameterize an optical‐based ET algorithm and test the relationship between ET from Landsat and MODIS by regression approaches. Our analyses show significant decreases in VIs and ET for both the 20‐year and post‐pulse 5‐year periods. Over the last 20 years, EVI Landsat declined 34% (30% by EVIMODIS) and ETLandsat‐EVI declined 38% (27% by ETMODIS‐EVI), overall ca. 1.61 mm/day or 476 mm/year drop in ET; using PAM ETLandsat‐EVI the drop was from 1130 to 654 mm/year. Over the 5 years since the 2014 pulse flow, EVILandsat declined 20% (13% by EVIMODIS) and ETLandsat‐EVI declined 23% (4% by ETMODIS‐EVI) with a 0.77 mm/day or a 209 mm/year 5‐year drop in ET; using PAM ETLandsat‐EVI the drop was from 863 to 654 mm/year. Data and change maps show the pulse flow contributed enough water to slow the rate of loss, but only for the very short‐term (1–2 years). These findings are critically important as they suggest further deterioration of biodiversity, wildlife habitat and key ecosystem services due to anthropogenic diversions of water in the U.S. and Mexico and from land clearing, fires and plant‐related drought which affect hydrological processes.

Comparison of evapotranspiration methods in the DSSAT Cropping System Model: II. Algorithm performance

Accurate calculations of evapotranspiration (ET) are highly important for agroecosystem model simulations, and improvement of ET algorithms is an on-going model development goal. The objective of this study was to evaluate and compare six ET methods in the Decision Support System for Agrotechnology Transfer (DSSAT) Cropping System Model (CSM) using agronomic and weighing lysimetry data from cotton field studies at Bushland, Texas. Three options were tested for estimating potential ET as required by the DSSAT-CSM: 1) a Priestley-Taylor method, 2) a Penman-Monteith combination equation estimate of grass reference ET with a DSSAT-specific single crop coefficient equation, and 3) the ASCE Standardized Reference ET Equation combined with a dual crop coefficient method for non-stressed conditions. The latter two reference ET methods were adapted to provide reasonable estimates for DSSAT-required potential ET. Additionally, two methods for calculation of soil water evaporation were tested, including both the original and updated formulations of Ritchie approaches for DSSAT-CSM. The combinations of the three potential ET and two soil water evaporation approaches led to six possible ET simulation options in the model. A computationally-intensive multiobjective optimization method was used to select among model parameterization options and ensure that modeler bias did not influence ET method comparisons. Among 23 agroecosystem metrics that included lysimeter-based ET, various cotton growth variables, and soil water content in multiple soil layers, the original Ritchie soil water evaporation approach performed statistically equivalent to or better than the more recent Ritchie method (p⩽0.05). The default ET method in the model, which involved Priestley-Taylor potential ET with the more recent Ritchie soil water evaporation method, was outperformed by other ET methods for 14 of 23 agroecosystem metrics (p⩽0.05). When the original Ritchie soil water evaporation method was combined with potential ET from the ASCE reference ET and dual crop coefficient method, the model performed statistically equivalent to or better than the other five ET options for all but 1 of 23 agroecosystem metrics (p⩽0.05). Based on three years of cotton data from the Bushland lysimetry fields, a DSSAT-CSM ET approach based on the standardized ET methodologies described by ASCE and FAO-56 combined with the original Ritchie soil water evaporation method provided holistic improvements to model simulations among multiple agroecosystem metrics.

Comparison of evapotranspiration methods in the DSSAT Cropping System Model: I. Global sensitivity analysis

Global sensitivity analysis (GSA) is useful for evaluating the responsiveness of agroecosystem models to input parameter adjustments. Evaluations of model sensitivity for diverse water status conditions and evapotranspiration (ET) algorithms will facilitate better use of models to provide water management recommendations. The objective of this study was to conduct a GSA to identify influential parameters in the Decision Support System for Agrotechnology Transfer (DSSAT) Cropping System Model (CSM), specifically using the CROPGRO-Cotton module with data from cotton field studies conducted in 2000, 2001, and 2008 at Bushland, Texas. The field studies tested fully-irrigated, deficit-irrigated, and dryland cotton production in a semi-arid environment. Using high performance computing resources, a GSA was conducted to evaluate the sensitivity of 24 model outputs with respect to 37 model input parameters. The GSA was conducted for six different ET methodologies available in the model. With first-order sensitivity indices <0.05, nearly half of the tested input parameters did not influence any model output for any of the tested simulation scenarios. Among the parameters with first-order sensitivity indices >0.05, eleven were cultivar parameters that controlled crop development and growth, and five were soil parameters that specified initial soil water conditions, soil water limits, drainage rate, and root growth characteristics. The influences of another soil parameter and one ET parameter were relevant only for the ET methods that required them. Large differences in sensitivity indices were found based on the choice between two soil water evaporation methods. In addition to providing insights for other applications of this model, the results specifically informed further efforts to evaluate the model using measured data from the Bushland cotton studies to compare performance among the six ET methods, as reported in a companion paper.

Evaluation of Evapotranspiration Models Using Different LAI and Meteorological Forcing Data from 1982 to 2017

We evaluated the performance of three global evapotranspiration (ET) models at local, regional, and global scales using the multiple sets of leaf area index (LAI) and meteorological data from 1982 to 2017 and investigated the uncertainty in ET simulations from the model structure and forcing data. The three ET models were the Simple Terrestrial Hydrosphere model (SiTH) developed by our team, the Priestley–Taylor Jet Propulsion Laboratory model (PT-JPL), and the MODerate Resolution Imaging Spectroradiometer (MODIS) ET algorithm (MOD16). Comparing the observed with simulated monthly ET by the three models over 43 Fluxnet sites, we found that SiTH overestimated ET for forests with mean slope from 1.25 to 1.67, but it performed better than the other two models over short vegetation. MOD16 and PT-JPL models simulated well for forests but poorly in dryland biomes (slope = 0.25~0.55; R2 = 0.02~0.46). At the catchment scale, all models performed well, except for some tropical and high latitudinal catchments, with NSE values lower than 0 and RMSE and MAE values far beyond their mean values. At the global scale, SiTH highly overestimated ET in tropics, while PT-JPL slightly underestimated ET between 30°N and 60°N and MOD16 underestimated ET between 15°S and 30°S. Generally, the PT-JPL provided the better performance than SiTH and MOD16 models. This study also revealed that the estimated ET by SiTH and especially PT-JPL model were influenced by the uncertainty in meteorological data, and the estimated ET was performed better using MERRA-2 datasets for PT-JPL and using ERA5 datasets for SiTH. While the estimated ET by MOD16 were relatively sensitive to LAI data. In addition, our results suggested that the GLOBMAP and GIMMS datasets were more suitable for long-term ET simulations than the GLASS dataset.

Comparison of Remote Sensing based Multi-Source ET Models over Cropland in a Semi-Humid Region of China

The estimation of cropland evapotranspiration (ET) is essential for agriculture water management, drought monitoring, and yield forecast. Remote sensing-based multi-source ET models have been widely applied and validated in the semi-arid region of China. However, careful investigation of the models’ performances for different crop types (winter wheat and summer maize) over the semi-humid region is still necessary. This study used remote sensing data (Landsat 8 and ASTER) and compared three mainstream multi-source ET models: (i) the two-source energy balance model, i.e., TSEB; (ii) the Penman-Monteith based four-source model, i.e., 4s-PM; (iii) the Priestley Taylor-Jet Propulsion Laboratory ET algorithm, i.e., PT-JPL. The measurements of the eddy-covariance (EC) flux tower located in Guantao county of North China were used to validate the models. The results showed that the TSEB model performed the best in estimating latent heat flux (LE) of maize, with an RMSE of 75.0 W/m2 and an R2 of 0.9, and the 4s-PM model had the highest accuracy of LE estimation for wheat, with an RMSE of 61.0 W/m2 and an R2 of 0.91. The LE spatial distribution comparison indicated that the PT-JPL model had more capacity to exhibit crop ET heterogeneity. The major environmental factors affecting ET varied with crop types and crop growth stages. Without taking soil moisture into account, the 4s-PM and TSEB models overestimated LE under water deficit in the maturation stage of wheat. The plant moisture stress based on vegetation index in the PT-JPL model underestimated the evaporation in the maturation stage while the cropland was still wet.

Sensitivity and uncertainty quantification for the ECOSTRESS evapotranspiration algorithm – DisALEXI

Evapotranspiration (ET) is a measure of plant water use that is utilized regionally for drought detection and monitoring, and locally for agricultural water resource management. Understanding the uncertainty associated with this measurement is vital for science predictions and analysis and for water resource management decision making. In this manuscript, the uncertainty in disaggregated Atmosphere-Land Exchange (disALEXI) is quantified; disALEXI is an ET algorithm that utilizes land surface temperature (LST) derived from the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS), as well ancillary inputs for landcover, elevation, vegetation parameters, and meteorological inputs. Since each of these inputs has an associated, and potentially unknown, uncertainty, in this study a Monte Carlo simulation based on a spatial statistical model is used to determine the algorithm's sensitivity to each of its inputs, and to quantify the probability distribution of algorithm outputs. Analysis shows that algorithm is most sensitive to LST (the input derived from ECOSTRESS). Significantly, the output uncertainty distribution is non-Gaussian, due to the non-linear nature of the algorithm. This means that ET uncertainty cannot be prescribed by accuracy and precision alone. Here, uncertainty was represented using five quantiles of the output distribution. The distribution was consistent across five different datasets (mean offset is 0.01 mm/day, and 95% of the data is contained within 0.3 mm/day). An additional two datasets with low ET, showed higher uncertainty (95% of the data is within 1 mm/day), and a positive bias (i.e., ET was overestimated by an average of 0.12 mm/day when ET was low).

Estimating Actual Evapotranspiration over a Large and Complex Irrigation System of the Nile Delta in Egypt

In the Nile Delta region of Egypt, measurements of Actual Evapotranspiration (ETa) and Potential Evapotranspiration (ETp) are difficult, expensive and labor-intensive. The current paper aimed at finding the superior method for estimating the ETa and ETa/ETp in the Nile Delta governorates by comparing different methods. Three different methods were used for ETa estimates being; Remote Sensing approach by Moderate Resolution Imaging Spectroradiometer (MODIS) ETa-product (MOD16A2), FAO (33) method, and Irrigation Water Balance Calculation (IWBC) method. The three ETa methods were applied for five governorates in 2017, where the data for IWBC were available. However, only MOD16A2 was compared with FAO (33) for ETa/ETp ratio and the two selected methods were applied for all eight Nile Delta governorates for the period of 2008 -2017. The MOD16A2 product was derived from the MODIS satellite images using an improved evapotranspiration algorithm based on the Penman-Monteith equation. FAO (33) was based on the relationship between the relative yield loss of any crop to relative reduction of water consumption. The IWBC required estimation of both; the field application and conveyance water losses as the only unknown elements of irrigation water balance in Delta governorates. For comparison between applied methods, descriptive statistics, analysis of variance (ANOVA) for checking difference, and cluster analysis were applied. The results showed a significant difference values between MOD16A2 and FAO (33) for estimating ETa/ETp ratios. However, the difference for ETa estimation was insignificant between the three methods indicating a significant relationship, with a strong correlation between MOD16A2 and IWBC. It was observed that ETa values were impacted by the cropping pattern, since they were very close in governorates having the same dominant crops. In conclusion, both MOD16A2 and IWBC can be utilized for ETa estimation. Both MOD16A2 and FAO (33) are not confirmed for estimation of ETa/ETp ratio due to the significant difference between both results. FAO (33) cannot be utilized for both ETa and ETa/ETp ratio estimates. Further data collection and investigation on ETa and ETp estimates methods are recommended.

A three-source satellite algorithm for retrieving all-sky evapotranspiration rate using combined optical and microwave vegetation index at twenty AsiaFlux sites

Evapotranspiration (ET) composed of transpiration from vegetation (ETveg) and evaporation from soil (ETsoil) and interception water (ETint) is a crucial process for driving global water and energy cycle. In this study, a three-source satellite ET algorithm is developed for estimating instantaneous ET rate (ETcal) under clear and non-rainy cloudy sky conditions and is validated at 20 AsiaFlux sites of multiple vegetation types (evergreen and deciduous forests, croplands and grasslands) in East Asia. The ET algorithm is completely independent from ground-based measurements and thus has the potential to be used on regional scale. All-sky surface net radiation flux is derived from satellite products. Satellite optical Normalized Difference Vegetation Index (NDVI) is used to partition the surface available energy into vegetated area and bare soil area. Satellite microwave Emissivity Difference Vegetation Index (EDVI), an indicator of vegetation water content, is utilized to estimate canopy resistance and transpiration from vegetation. Evaporation components from canopy interception water and soil surface are estimated based on Priestley-Taylor method. The estimated evaporation fraction (EF) for all sites can produce a small absolute bias of less than 0.03 (or 8.1%) with a good determination coefficient (R2) larger than 0.42 when validated against the corrected in-situ measured EF on instantaneous and long-term mean scales. Instantaneous ETcal under all sky conditions are found to have a R2 of 0.61, a bias of 16.5 Wm-2 (13.7%) and a root mean square error (RMSE) of 63.6 Wm-2 (52.5%) over all sites, respectively. The performances of this algorithm are better achieved on 16-day (R2 = 0.66, RMSE = 44.1 Wm-2) and monthly (R2 = 0.70, RMSE = 39.1 Wm-2) scales. More importantly, EDVI-based ET method is able to produce ETcal with stable and good accuracy from clear sky to overcast sky (bias < 22 Wm-2 or 18%, RMSE < 77 Wm-2 or 59%). The daily ET and its components (ETveg, ETsoil and ETint) derived from our algorithm are further compared with those from Global Land Evaporation Amsterdam Model (GLEAM) ET products. EDVI-based ET (EDVI-ET) generally shows equal performances with GLEAM-ET when using in-situ ET as reference. EDVI-ETveg shows a fairly good correlation (R2 = 0.21) and an underestimation of -0.32 mm/day compared with GLEAM-ETveg. EDVI-ETint shows a significant correlation (R2 = 0.30) with GLEAM-ETveg after excluding three outlier tropical forest sites. Meanwhile, EDVI-ETsoil shows a fairly good positive correlation (R2 = 0.15) and an overestimation of 0.23 mm/day compared with GLEAM-ETsoil. The results indicate that our method has great potential for regional ET estimation under various cloud conditions when combined with satellite optical and microwave remote sensing.

Analysis of Dry/Wet Variations in the Poyang Lake Basin Using Standardized Precipitation Evapotranspiration Index Based on Two Potential Evapotranspiration Algorithms

Global warming has resulted in unevenly distributed changes in precipitation and evapotranspiration, which has some influence on dry/wet conditions, thus exerting a tremendous impact on national life and the social economy, especially agricultural production. In order to characterize the dry/wet variations in the Poyang Lake basin during 1958–2013, based on the potential evapotranspiration (PET) estimated by the Thornthwaite (TH) and Penman–Monteith (PM) formulas, two types of Standardized Precipitation Evapotranspiration Index (SPEI), namely SPEI_th and SPEI_pm, were calculated in this study. The characteristic of dry/wet variations in the Poyang Lake basin was analyzed and a comparative analysis of two SPEIs was conducted. The results indicate that both SPEI series showed a wet trend in the Poyang Lake basin on an annual scale as well as seasonal scales during 1958–2013, except for spring and autumn. A drying trend was observed in spring, while in autumn, the dry and wet conditions in two SPEIs had opposite trends. However, all trends from two SPEIs were not significant, except for summer SPEI_pm. Meanwhile, significant positive correlations were detected between precipitation and two SPEIs, with the correlation coefficients above 0.95, whereas negative correlations were detected between PET and two SPEIs, with the correlation coefficients ranging from −0.17 to −0.85. This indicates that precipitation was the main climatic factor to determine change in dry/wet conditions in the Poyang Lake basin. Although there were obvious differences between the accumulated values of the Penman–Monteith-based PET (ET_pm) and Thornthwaite-based PET (ET_th), trends in the SPEI_pm series were generally consistent with those in the SPEI_th series, revealing that the method for PET calculation was not critical to the change in dry/wet conditions. Moreover, the results of the conditional probability of SPEI_pm and SPEI_th show that both SPEI_pm and SPEI_th could detect wet or dry events that were identified by SPEI_pm or SPEI_th.

Novel methodology to evaluate and compare evapotranspiration algorithms in an agroecosystem model

Efforts to improve agroecosystem models require methods for unbiased comparisons among simulation algorithms. With focus on evapotranspiration (ET) in Cotton2K, the objectives were to develop a novel methodology for evaluating model parameterization options and to compare model performance using three ET algorithms. The Cotton2K model was updated to include a standardized ET method, and two Penman approaches were also tested. Sobol global sensitivity analysis and multiobjective optimization were used to identify influential parameters and select feasible parameterization options. The three ET methods led to differences in simulation accuracy for ET, soil water contents, and several plant growth metrics (p<0.05). However, no ET method could consistently outperform the other two methods, and ET simulation errors were up to 60%. The simulation methodology permitted unbiased comparison of three ET methods in Cotton2K and highlighted areas for model improvement, including the surface evaporation simulation and the linkage between simulated ET and crop growth.

Evapotranspiration in North America: implications for water resources in a changing climate

Accurate quantification of evapotranspiration (ET) is important to understanding its role in the global hydrological cycle of terrestrial ecosystems and feedbacks to the climate system. This study improves ET quantification in North America using a data assimilation technique and a process-based Terrestrial Ecosystem Model as well as in situ and satellite data. ET is modeled using the Penman-Monteith equation with an improved leaf area index (LAI) algorithm in a biogeochemistry model, the Terrestrial Ecosystem Model (TEM). The evaluated TEM was used to estimate ET at site and regional scales in North America from 2000 to 2010. The estimated annual ET varies from 420 to 450 mm year−1 with the improved model, close to Moderate Resolution Imaging Spectroradiometer monthly data with a root-mean-square error less than 10 mm month−1 for the study period. Alaska, Canada, and the conterminous US account for 33%, 6%, and 61% of the regional ET, respectively. Water availability, the difference between precipitation and ET, is 181 mm month−1, averaged from 2000 to 2010. Under IPCC Representative Concentration Pathway (RCP) 2.6 and RCP 8.5 scenarios, the regional ET increases by 11% and 24%, respectively. Consequently, the water availability decreases in the region by 2.4% and 23.7%, respectively. For the period of 2020–2100, due to uncertain parameters, TEM versions integrated with three different ET algorithms estimated that the regional ET in the USA are 430.5 ± 10.5 mm year−1, 482.1 ± 11.2 mm year−1, and 489.7 ± 13.4 mm year−1, and the available water is − 105.3 ± 8.7 mm year−1, − 20.3 ± 11.9 mm year−1, and − 126.2 ± 15.4 mm year−1, respectively, by the end of the twenty-first century. Our analysis suggests natural terrestrial ecosystem soils in North America will get drier under future climate conditions, which will impact the regional water resource and the climate system. Based on our ET simulation under three climate change scenarios, our study suggests that the RCP 2.6 is the optimum trajectory for preserving freshwater resources in North America and other regions in the globe.

Intercomparison of remote-sensing based evapotranspiration algorithms over amazonian forests

Evapotranspiration (ET) is considered a key variable in the understanding of the Amazonian tropical forests and their response to climate change. Remote-Sensing (RS) based evapotranspiration models are presented as a feasible means in order to provide accurate spatially-distributed ET estimates over this region. In this work, the performance of four commonly used ET RS models was evaluated over Amazonia using Moderate Resolution Imaging Spectroradiometer (MODIS) data. RS models included i) Priestley-Taylor Jet Propulsion Laboratory (PT-JPL), ii) Penman-Monteith MODIS operative parametrization (PM-Mu), iii) Surface Energy Balance System (SEBS), and iv) Satellite Application Facility on Land Surface Analysis (LSASAF). These models were forced using two ancillary meteorological data sources: i) in-situ data extracted from Large-Scale Biosphere-Atmosphere Experiment (LBA) stations (scenario I), and ii) three reanalysis datasets (scenario II), including Modern-Era Retrospective analysis for Research and Application (MERRA-2), European Centre for Medium-range Weather Forecasts (ECMWF) Re-Analysis-Interim (ERA-Interim), and Global Land Assimilation System (GLDAS-2). Performance of algorithms under the two scenarios was validated using in-situ eddy-covariance measurements. For scenario I, PT-JPL provided the best agreement with in-situ ET observations (RMSE = 0.55 mm/day, R = 0.88). Neglecting water canopy evaporation resulted in an underestimation of ET measurements for LSASAF. SEBS performance was similar to that of PT-JPL, nevertheless SEBS estimates were limited by the continuous cloud cover of the region. A physically-based ET gap-filling method was used in order to alleviate this issue. PM-Mu tended to overestimate in-situ ET observations. For scenario II, quality assessment of reanalysis input data demonstrated that MERRA-2, ERA-Interim and GLDAS-2 contain biases that impact model performance. In particular, biases in radiation inputs were found the main responsible of the observed biases in ET estimates. For the region, MERRA-2 tends to overestimate daily net radiation and incoming solar radiation. ERA-Interim tends to underestimate both variables, and GLDAS tends to overestimate daily radiation while underestimating incoming solar radiation. Discrepancies amongst these reanalysis inputs generally explain the observed discrepancies in model spatial and temporal patterns.

LandMOD ET mapper: A new matlab-based graphical user interface (GUI) for automated implementation of SEBAL and METRIC models in thermal imagery

Remote sensing based surface energy balance (SEB) models, such as SEB Algorithm for Land (SEBAL) and Mapping evapotranspiration (ET) at high Resolution with Internalized Calibration (METRIC) are widely used to monitor ET. However, their operational use is challenged by model complexity and the tedious process of selecting hot and cold pixels. We developed a new Matlab-based ET mapping toolbox named “LandMOD ET Mapper” for automated implementation of these models to facilitate their widespread applications among new users. The toolbox is demonstrated using one Landsat and one Moderate Resolution Imaging Spectroradiometer (MODIS) Terra image covering three AmeriFlux sites in Nebraska, U.S. Daily ET estimates were comparable with the observed ET for five days when near cloud-free Landsat and MODIS Terra images were available in the 2006 maize growing season. The toolbox provides a viable ET mapping option to new and inexperienced users with reduced time and labor demands.

Evaluating a new algorithm for satellite-based evapotranspiration for North American ecosystems: Model development and validation

We introduce a different operational approach to estimate 8-day average daily evapotranspiration (ET) using both routinely available data and the Penman-Monteith (P-M) equation for canopy transpiration and evaporation of intercepted water and Priestley and Taylor for soil evaporation. Our algorithm considered the environmental constraints on canopy resistance and ET by (1) including vapor pressure deficit (VPD), incoming solar radiation, soil moisture, and temperature constraints on stomatal conductance; (2) using leaf area index (LAI) to scale from the leaf to canopy conductance; and (3) calculating canopy resistance as a function of environmental variables such as net radiation and VPD. Remote sensing data from the Moderate Resolution Spectroradiometer (MODIS) and satellite soil moisture data were used to derive the ET model. The algorithm was calibrated and evaluated using measured ET data from 20 AmeriFlux Eddy covariance flux sites for the period of 2003–2012. We found good agreements between our 8-day ET estimates and observations with mean absolute error (MAE) ranges from 0.17 mm/day to 0.94 mm/day compared with MAE ranging from 0.28 mm/day to 1.50 mm/day for MODIS ET. Compared to MODIS ET, our proposed algorithm has higher correlations and higher Willmott’s index of agreement with observations for the majority of the Ameriflux sites. The strong relationship between the model estimated ET and the flux tower observations implies that our model has the potential to be applied to different ecosystems and at different temporal scales.

SMAP soil moisture improves global evapotranspiration

Accurate estimation of global evapotranspiration (ET) is essential to understand water cycle and land-atmosphere feedbacks in the Earth system. Satellite-driven ET models provide global estimates, but many of the ET algorithms have been designed independently of soil moisture observations. As water for ET is sourced from the soil, incorporating soil moisture into global remote sensing algorithms of ET should, in theory, improve performance, especially in water-limited regions. This paper presents an update to the widely-used Priestley Taylor-Jet Propulsion Laboratory (PT-JPL) ET algorithm to incorporate spatially explicit daily surface soil moisture control on soil evaporation and canopy transpiration. The updated algorithm is evaluated using 14 AmeriFlux eddy covariance towers co-located with COsmic-ray Soil Moisture Observing System (COSMOS) soil moisture observations. The new PT-JPLSM model shows reduced errors and increased explanation of variance, with the greatest improvements in water-limited regions. Soil moisture incorporation into soil evaporation improves ET estimates by reducing bias and RMSE by 29.9% and 22.7% respectively, while soil moisture incorporation into transpiration improves ET estimates by reducing bias by 30.2%, RMSE by 16.9%. We apply the algorithm globally using soil moisture observations from the Soil Moisture Active Passive Mission (SMAP). These new global estimates of ET show reduced error at finer spatial resolutions and provide a rich dataset to evaluate land surface and climate models, vegetation response to changes in water availability and environmental conditions, and anthropogenic perturbations to the water cycle.

Comparison of MODIS and SWAT evapotranspiration over a complex terrain at different spatial scales

Abstract. In most hydrological systems, evapotranspiration (ET) and precipitation are the largest components of the water balance, which are difficult to estimate, particularly over complex terrain. In recent decades, the advent of remotely sensed data based ET algorithms and distributed hydrological models has provided improved spatially upscaled ET estimates. However, information on the performance of these methods at various spatial scales is limited. This study compares the ET from the MODIS remotely sensed ET dataset (MOD16) with the ET estimates from a SWAT hydrological model on graduated spatial scales for the complex terrain of the Sixth Creek Catchment of the Western Mount Lofty Ranges, South Australia. ET from both models was further compared with the coarser-resolution AWRA-L model at catchment scale. The SWAT model analyses are performed on daily timescales with a 6-year calibration period (2000–2005) and 7-year validation period (2007–2013). Differences in ET estimation between the SWAT and MOD16 methods of up to 31, 19, 15, 11 and 9 % were observed at respectively 1, 4, 9, 16 and 25 km2 spatial resolutions. Based on the results of the study, a spatial scale of confidence of 4 km2 for catchment-scale evapotranspiration is suggested in complex terrain. Land cover differences, HRU parameterisation in AWRA-L and catchment-scale averaging of input climate data in the SWAT semi-distributed model were identified as the principal sources of weaker correlations at higher spatial resolution.