Improved Evapotranspiration Estimation in Coarse-Textured Soils Using a Comprehensive Water Balance Model

Thermal infrared remote sensing observations have been widely used to provide useful information on surface energy and water stress for estimating evapotranspiration (ET). However, the revisit time of current high spatial resolution (<100 m) thermal infrared remote sensing systems, sixteen days for Landsat for example, can be insufficient to reliably derive ET information for water resources management. We used in situ ET measurements from multiple Ameriflux sites to (1) evaluate different scaling methods that are commonly used to derive daytime ET estimates from time-of-day observations; and (2) quantify the impact of different revisit times on ET estimates at monthly and seasonal time scales. The scaling method based on a constant evaporative ratio between ET and the top-of-atmosphere solar radiation provided slightly better results than methods using the available energy, the surface solar radiation or the potential ET as scaling reference fluxes. On average, revisit time periods of 2, 4, 8 and 16 days resulted in ET uncertainties of 0.37, 0.55, 0.73 and 0.90 mm per day in summer, which represented 13%, 19%, 23% and 31% of the monthly average ET calculated using the one-day revisit dataset. The capability of a system to capture rapid changes in ET was significantly reduced for return periods higher than eight days. The impact of the revisit on ET depended mainly on the land cover type and seasonal climate, and was higher over areas with high ET. We did not observe significant and systematic differences between the impacts of the revisit on monthly ET estimates that are based on morning or afternoon observations. We found that four-day revisit scenarios provided a significant improvement in temporal sampling to monitor surface ET reducing by around 40% the uncertainty of ET products derived from a 16-day revisit system, such as Landsat for instance.

Estimating evapotranspiration (ET), the main water output flux within basins, is an important step in assessing hydrological changes and water availability. However, direct measurements of ET are challenging, especially for large regions. Global products now provide gridded estimates of ET at different temporal resolution, each with its own method of estimating ET based on various data sources. This study investigates the differences between ERA5, GLEAM, and GLDAS datasets of estimated ET at gridded points across Iran, and their accuracy in comparison with reference ET. The spatial and temporal discrepancies between datasets are identified, as well as their co-variation with forcing variables. The ET reference values used to check the accuracy of the datasets were based on the water balance (ETwb) from Iran’s main basins, and co-variation of estimated errors for each product with forcing drivers of ET. The results indicate that ETERA5 provides higher base average values and lower maximum annual average values than ETGLEAM. Temporal changes at the annual scale are similar for GLEAM, ERA5, and GLDAS datasets, but differences at seasonal and monthly time scales are identified. Some discrepancies are also recorded in ET spatial distribution, but generally, all datasets provide similarities, e.g., for humid regions basins. ETERA5 has a higher correlation with available energy than available water, while ETGLEAM has higher correlation with available water, and ETGLDAS does not correlate with none of these drivers. Based on the comparison of ETERA5 and ETGLEAM with ETwb, both have similar errors in spatial distribution, while ETGLDAS provided over and under estimations in northern and southern basins, respectively, compared to them (ETERA5 and ETGLEAM). All three datasets provide better ET estimates (values closer to ETWB) in hyper-arid and arid regions from central to eastern Iran than in the humid areas. Thus, the GLEAM, ERA5, and GLDAS datasets are more suitable for estimating ET for arid rather than humid basins in Iran.

A robust method for characterizing the biophysical environment of terrestrial vegetation uses the relationship between Actual Evapotranspiration (AET) and Climatic Water Deficit (CWD). These variables are usually estimated from a water balance model rather than measured directly and are often more representative of ecologically-significant changes than temperature or precipitation. We evaluate trends and spatial patterns in AET and CWD in the Continental United States (CONUS) during 1980–2019 using a gridded water balance model. The western US had linear regression slopes indicating increasing CWD and decreasing AET (drying), while the eastern US had generally opposite trends. When limits to plant performance characterized by AET and CWD are exceeded, vegetation assemblages change. Widespread increases in aridity throughout the west portends shifts in the distribution of plants limited by available moisture. A detailed look at Sequoia National Park illustrates the high degree of fine-scale spatial variability that exists across elevation and topographical gradients. Where such topographical and climatic diversity exists, appropriate use of our gridded data will require sub-setting to an appropriate area and analyzing according to categories of interest such as vegetation communities or across obvious physical gradients. Recent studies have successfully applied similar water balance models to fire risk and forest structure in both western and eastern U.S. forests, arid-land spring discharge, amphibian colonization and persistence in wetlands, whitebark pine mortality and establishment, and the distribution of arid-land grass species and landscape scale vegetation condition. Our gridded dataset is available free for public use. Our findings illustrate how a simple water balance model can identify important trends and patterns at site to regional scales. However, at finer scales, environmental heterogeneity is driving a range of responses that may not be simply characterized by a single trend.

Due to the proliferation of precision agriculture, the obstacle of estimating evapotranspiration (ET) and its components from shadow pixels acquired from remote sensing technology should not be neglected. To accurately detect shaded soil and leaf pixels and quantify the implications of shadow pixels on ET inversion, a two-year field-scale observation was carried out in the growing season for a pinot noir vineyard. Based on high-resolution remote sensing sensors covering visible light, thermal infrared, and multispectral light, the supervised classification was applied to detect shadow pixels. Then, we innovatively combined the normalized difference vegetation index with the three-temperature model to quantify the proportion of plant transpiration (T) and soil evaporation (E) in the vineyard ecosystem. Finally, evaluated with the eddy covariance system, we clarified the implications of the shadow pixels on the ET estimation and the spatiotemporal patterns of ET in a vineyard system by considering where shadow pixels were presented. Results indicated that the shadow detection process significantly improved reliable assessment of ET and its components. (1) The shaded soil pixels misled the land cover classification, with the mean canopy cover ignoring shadows 1.68–1.70 times more often than that of shaded area removal; the estimation accuracy of ET can be improved by 4.59–6.82% after considering the effect of shaded soil pixels; and the accuracy can be improved by 0.28–0.89% after multispectral correction. (2) There was a 2 °C canopy temperature discrepancy between sunlit leaves and shaded leaves, meaning that the estimation accuracy of T can be improved by 1.38–7.16% after considering the effect of shaded canopy pixels. (3) Simultaneously, the characteristics showed that there was heterogeneity of ET in the vineyard spatially and that E and T fluxes accounted for 238.05 and 208.79 W·m−2, respectively; the diurnal variation represented a single-peak curve, with a mean of 0.26 mm/h. Our findings provide a better understanding of the influences of shadow pixels on ET estimation using remote sensing techniques.

The paper presents an analysis of the dependence of summertime daily maximum temperature on antecedent soil moisture using daily surface observations from a selection of stations in the contiguous United States and daily time series of soil moisture computed with a simple local water balance model. The computed soil moisture time series are offered as an alternative to Palmer’s soil moisture anomaly (Z) index, the Palmer Drought Severity Index (PDSI), and other such time series. In contrast to other water balance models that have been designed for the computation of soil moisture time series, the model herein is driven by daily rather than monthly data, uses the Priestley–Taylor method in lieu of Thornthwaite’s method to calculate potential evapotranspiration, allows for runoff during dry periods as well as when soil moisture is not at field capacity, includes a crude scheme for taking into account the effects of snowmelt on the water balance, and permits geographical variations in soil water cap...

This paper presents an application of a long-term, large catchment-scale, water balance model developed to predict the effects of forest clearing in the south-west of Western Australia. The conceptual model simulates the basic daily water balance fluxes in forested catchments before and after clearing. The large catchment is divided into a number of sub-catchments (1–5 km2 in area), which are taken as the fundamental building blocks of the large catchment model. The responses of the individual subcatchments to rainfall and pan evaporation are conceptualized in terms of three inter-dependent subsurface stores A, B and F, which are considered to represent the moisture states of the subcatchments. Details of the subcatchment-scale water balance model have been presented earlier in Part 1 of this series of papers. The response of any subcatchment is a function of its local moisture state, as measured by the local values of the stores. The variations of the initial values of the stores among the subcatchments are described in the large catchment model through simple, linear equations involving a number of similarity indices representing topography, mean annual rainfall and level of forest clearing. The model is applied to the Conjurunup catchment, a medium-sized (39·6 km2) catchment in the south-west of Western Australia. The catchment has been heterogeneously (in space and time) cleared for bauxite mining and subsequently rehabilitated. For this application, the catchment is divided into 11 subcatchments. The model parameters are estimated by calibration, by comparing observed and predicted runoff values, over a 18 year period, for the large catchment and two of the subcatchments. Excellent fits are obtained.

Reduction in agricultural water use and increasing the sustainability of water resources can be achieved by studying the water balance of the area and crop water demand. In this paper, by using a simple water balance model, Evapotranspiration, Rainfall, Runoff, Water Demand and Water Requirement different crops are estimated. The crop water requirement and crop water demand for different crops grown in the Periya Pallam Catchment of Upper Bhavani Basin, Tamilnadu, was estimated. Water balance estimation of the area reveals that out of the annual rainfall, runoff is estimated to be 129 mm, effective rainfall is 252 mm, and deep percolation is about 67 mm. The demand for water for agriculture in the study area is about 61 million cubic meters (MCM), but only 19 MCM of water is available through precipitation in the form of effective rainfall. Hence, the remaining 43 MCM of water is supplied through groundwater and other sources. The results will pave the way for sustainable crop water use planning and would achieve water security in the basin.

Precipitation and evapotranspiration partitioning on the Three-River Source Region: A comparison between water balance and energy balance models

Drought constraints and transpiration of rubber (Hevea brasiliensis) plantations under different evaporative demand regimes were assessed by the simple water balance model. A lump water model, BILJOU (BILan hydrique JOUrnalier), is the daily water balance model. This model requires daily potential evapotranspiration (ETo) and rainfall as input climatic data, also requires site and stand parameters are maximum extractable soil water and leaf area index (LAI). The study was carried out two sites; namely Songkhla and Chachoengsao province, Thailand, traditional and new plantation area, respectively. The calibration of this model was done with sap flow measurements. Soil water derived by tensiometer for Songkhla and soil sampling for Chachoengsao were used to validate the model. Under non limiting soil water and full canopy, transpiration of rubber was influenced by evaporative demand. Consistently, under limited soil water represented as threshold of relative extractable water (REWc < 0.4), transpiration was influenced by REW. In the new plantation area; Chachoengsao, drought constraints were evident annually from the beginning of senescence until the new accomplished flushing; December to June. However, at Songkhla site, transpiration and soil water was mainly driven by evaporative demand. And the limitation of soil water represented shortly during the plateau stage of LAI.

One of the most complex hydrogeological problems in the design and maintenance of drainage systems in abandoned mining sites is quantifying the maximum water infiltration and, therefore, the amount that is potentially drainable by the tunnels. This problem is compounded when water-inflow data are limited or lacking. The aim of the study was to present a single but reliable model for making this evaluation; this model was applied to the case history of the abandoned Cogne iron ore mining complex (Western Alps, Aosta Valley Region, NW Italy). The study focused on quantifying the amount of water infiltrating into the mine drifts, using a water balance model in a Geographic Information System (GIS) environment. In the model, five different infiltration scenarios were calculated, including a detailed analysis of rainfall data, snow density and thickness (Snow Water Equivalent calculation), and melting periods. The maximum water discharge that could affect the mine tunnels was, therefore, determined under several scenarios of normal precipitation conditions and during heavy rainfall, including the case of the Cogne valley flood in October 2000, used as a reference for the limit conditions. Taking into account the various approximations considered, the results can be considered a good indication of the magnitude of the total amount of water that should be drained out through abandoned mine drifts and in the drainage network during implementation of final closure of the mine.

Across most urban centres, adequate supply of water becomes essential for the urban poor. Attempts at ensuring water availability has been formulated using standard water models. This study developed a water balance model using Remote Sensing and GIS techniques during periods of water surplus and deficit across pre-millennium (1996 – 2000) and the post-millennium (2010 – 2014) periods within an urbanized Sub-Saharan city. Digital Elevation Model of the basin was prepared to generate flow direction and to delimit the watershed using ArcHydro tools in ArcGIS 10.3. Supervised classification was used to delineate the landuse classes from Landsat imageries. The result revealed that 1996, 1997, 1998, 2000, 2010 and 2013 were periods of water deficit, while 1999, 2011, 2012 and 2014 were periods of water surplus. This study affirmed the need to reactivate institutional frameworks that are responsible for assessment, conservation, management and planning of water resources for sustainable utilization.

This paper proposes an approach for the integration of land-surface evapotranspiration monitoring system. Two software of GIS and RS and two models of Energy and Water Balance were seamless integrated by component technology. An opening three-layer (data-layer, function-layer and application-layer from bottom to top) framework is constructed. The patterns of integrating GIS into RS and Energy and Water Balance model into the RS model are researched. Using C# mixed with ENVI/IDL and ArcGIS-Engine, the monitoring system has been developed, in which the functions of inversing the land-surface evapotranspiration(ET) parameters, model calculation, ET validation and 3D-ET visualization are established. The Sanhuajian basin in the middle of Yellow River was taken as a research watershed. The result shows that it is helpful to solve the shortage of complication processes in ET estimation using common softwares. The integration pattern can provide the effective technical support for the land-surface evapotranspiration monitoring system.

Observed differences between water balance for temperate and semiarid catchments can be attributed to variability in the primary controls of the soil profile (soil water storage capacity and permeability), vegetation (surface coverage and water use efficiency), and climate (rainfall and potential evaporation). Motivated by the downward development methodology mooted by Klemes [1983], this paper explores the underlying climate and landscape interactions that cause differences in water balance between a number of temperate and semiarid benchmark catchments around Australia. A systematic parsimonious analysis based around a simple water balance model whose parameters are estimated from available data is used to observe and interpret signatures of annual, intra‐annual, and daily water balance behavior and variability. The essence of the downward analysis is that model complexity evolves from the simplest form to its final quasi‐distributed framework with multiple storages and process interactions in response to inadequacies in signature prediction. Consequently, each step in the process develops insight into local climate and landscape influences and sensitivity to their temporal and spatial variabilities. The results affirm that the relative influence of climate and landscape properties upon catchment response is manifested in a systematic transformation of process sensitivity with increasing timescales (annual, monthly, daily, and hourly). Significantly, drier catchments are shown to be more sensitive to small‐scale perturbation than humid catchments. We conclude that the downward analysis approach is a means to better understand the significance of landscape variability upon catchment response. Consequential knowledge can begin to explain data requirements and requisite model complexity and to provide insight into the ability to reliably predict water balance behavior. These are important first steps toward the longer‐term objective of reliably predicting water balance variability across ungauged basins.

Irrigation plays a pivotal role in the hydrological cycle, representing about 70% of freshwater withdrawals. However, its representation in Earth System models is characterized by significant uncertainties in terms of amount, timing and spatial distribution. Earth Observation data offer a viable way to reduce this uncertainty thanks to their ability to sense the soil and vegetation in its real condition with few-days revisit timing and high spatial resolution (~ 10 m), e.g. with the new Sentinel missions.  In this contribution, we use remote sensing observations from the Sentinel-1 and Sentinel-2 satellite missions to constrain a simple Soil Water Balance (SWB) model coupled with the semi-empirical Water Cloud Model (WCM) and obtain irrigation estimates via an inverse modelling solution. The WCM, which is a model simulating backscatter observations (σ0) from soil moisture and a vegetation descriptor, is forced by vegetation indexes from Sentinel-2 data and soil moisture simulated by the SWB that includes a sprinkler irrigation scheme. The model outputs are then matched with Sentinel-1 observations to obtain irrigation estimates. The model is tested over an irrigated field of the Po River valley, one of the most intensively European irrigated areas. Results show that the model can capture the irrigation signal with relatively good accuracy. It also provides an estimate of soil moisture in the field.  Nonetheless the revisit time of the satellite platforms and the simplicity of backscatter model, especially in the representation of the vegetation component, constitute two main limitations of the model. This model is a viable tool that can be easily applied in the context of precision agriculture to optimize irrigation practices and conserve water resources even when in-situ soil moisture and irrigation measurements are not available.

It is well known that one of the major problems in the application of land surface models is the determination of the various model parameters. In most cases, only one or a limited number of variables are used to estimate these parameters. This study evaluates the use of two fundamentally different global optimization methods, multistart weight‐adaptive recursive parameter estimation (MWARPE) and particle swarm optimization (PSO), for the estimation of hydrologic model parameters on the basis of data for multiple variables. MWARPE iteratively uses the linear recursive filter equations in a Monte Carlo setting and therefore does not rely on the explicit minimization of an objective function. However, a major drawback of the MWARPE method is the high dimensionality, determined by the number of observations, of the matrix to be inverted. On the other hand, PSO is a stochastic optimization method based on the collective strength of a population of individuals with flocking or herding behavior, as observed in a wide number of biological systems. In situ observations of net radiation; latent, sensible, and ground heat fluxes; and the soil moisture profile are used to determine the parameters of a simplified water and energy balance model. Both optimization methods are analyzed in terms of model performance and computational efficiency. Comparable results, expressed in terms of the root mean square error values, were obtained for both methods. However, it was found that MWARPE tends to slightly overfit the data.