Search

Estimation of turbulent heat fluxes by assimilating sequences of land surface temperature (LST) measurements into variational data assimilation (VDA) frameworks has been the subject of several studies. The VDA approaches estimate turbulent heat fluxes by minimizing the difference between LST observations and estimations from the heat diffusion equation. The VDA methods have been tested only with high temporal resolution LST observations (e.g., from geostationary satellites) when applied at regional scales. Geostationary satellites can capture the diurnal cycle of LST, but they have a relatively low spatial resolution and mainly focus on low latitudes. To overcome these shortcomings, this study assimilates high spatial resolution LST data from polar orbiting satellites (e.g., Moderate Resolution Imaging Spectroradiometer, MODIS) into the combined-source (CS) and dual-source (DS) VDA schemes. An expression is developed to obtain an a priori evaporative fraction (EF) estimate from leaf area index (LAI) or apparent thermal inertia (ATI). The a priori EF estimate is used as an initial guess in the VDA approach. The results indicate that the VDA method is able to find the optimal value of EF by assimilating the low-temporal resolution MODIS LST data. The predicted turbulent heat fluxes from VDA are compared with the measurements from the large-aperture scintillometer at three sites (Arou, Daman, and Sidaoqiao) in the Heihe River Basin (located in northwest China). The findings indicate that the CS and DS VDA models perform well in various hydrological and vegetative conditions. The three-site-average root mean square errors (RMSEs) of sensible and latent heat fluxes estimates from the CS scheme are 37.44 W m−2 and 94.30 W m−2, respectively. The DS model reduces the abovementioned RMSEs by 19.82% and 21.37%, respectively. Overall, the results show that using the a priori EF estimate from the proposed expression in the VDA approach eliminates the need for the high resolution LST data from geostationary satellites, and allows the VDA method to estimate turbulent heat fluxes by assimilating LST data from polar orbiting satellites. Finally, several numerical tests are conducted to assess the effect of LST temporal sampling on the turbulent heat fluxes estimates. The results show that the LST measurement at 1400 Local Time (LT) has the most amount of information for partitioning the available energy into sensible and latent heat fluxes.

Emamgholizadeh S, Bateni SM, Nielson JR. 2018. Evaluation of different strategies for management of reservoir sedimentation in semi-arid regions: a case study (Dez Reservoir). Lake Reserv Manage. 0...

Estimation of turbulent heat fluxes by assimilating sequences of land surface temperature (LST) observations into a variational data assimilation (VDA) framework has been the subject of numerous studies. The VDA approaches are focused on the estimation of two key parameters that regulate the partitioning of available energy between sensible and latent heat fluxes. These parameters are neutral bulk heat transfer coefficient CHN and evaporative fraction (EF). The CHN mainly depends on the roughness of the surface and varies on the time scale of changing vegetation phenology. The existing VDA methods assumed that the variations in vegetation phenology over the period of one month are negligible and took CHN as a monthly constant parameter. However, during the growing season, bare soil may turn into a fully vegetated surface within a few weeks. Thus, assuming a constant CHN may result in a significant error in the estimation of surface fluxes, especially in regions with a high temporal variation in vegetation cover. In this study the VDA approach is advanced by taking CHN as a function of leaf area index (LAI). This allows the characterization of the dynamic effect of vegetation phenology on CHN. The performance of the new VDA model is tested over three sites in the United States and one site in China. The results show that the new model outperforms the previous one and reduces the root-mean-square error (and bias) in sensible and latent heat flux estimates across the four sites on average by 31% (61%) and 21% (37%), respectively.

The primary objective of this study is to assess the accuracy of the two-source variational data assimilation (TVDA) system for partitioning evapotranspiration (ET) into soil evaporation (ETS) and canopy transpiration (ETC). Its secondary aim is to compare performance of the TVDA system with the commonly used two-source surface energy balance (TSEB) method. A combination of eddy-covariance-based ET observations and stable-isotope-based measurements of the ratio of evaporation and transpiration to total evapotranspiration (ETS/ET and ETC/ET) over an irrigated cropland site (the so-called Daman site) in the middle reach of the Heihe River basin (northwestern China) was used to investigate these objectives. The results indicate that the TVDA method predicts ETS and ETC more accurately than TSEB. Root-mean-square errors (RMSEs) of midday (1300–1500 LT) averaged soil and canopy latent heat flux (LES and LEC) estimates from TVDA are 23.1 and 133.0 W m−2, respectively. Corresponding RMSE values from TSEB are 41.9 and 156.0 W m−2. Compared to TSEB, the TVDA method takes advantage of all of the information in land surface temperature observations in the estimation period by leveraging a dynamic model (the heat diffusion equation) and thus can generate more accurate LES and LEC estimates.

Summary When the number of computational grids increases, water quality simulation complexities arise. Therefore, using a reduced order framework to express the variations of the objective parameter may facilitate the simulation task and also the interpretation of computational results. In this regard, a new reduced-order approach was proposed to link a water quality simulator model (CE-QUAL-W2, W2) with a data reduction technique (proper orthogonal decomposition, POD). The W2 model simulated spatio-temporal variations of nitrate in the Karkheh Reservoir, Iran. Thereafter, the POD model reduced the dimensions of simulated nitrate in the computational grids. The performance of the developed reduced-order model (ROM) results was evaluated through the comparison of the regenerated nitrate data by the model, and the simulated ones by W2. Findings indicated that the first four modes among 1825 computed ones by ROM conserved approximately over 91% of the nitrate variations. It means that the ROM was capable of showing the spatio-temporal variations of nitrate in the reservoir using the first few modes. Finally, confirmation of ROM indicated that the error of order of magnitude was less than 0.001, for nitrate ROM to regenerate nitrate data using 100 basis functions.

Summary The soil cation exchange capacity (CEC) is one of the main soil chemical properties, which is required in various fields such as environmental and agricultural engineering as well as soil science. In situ measurement of CEC is time consuming and costly. Hence, numerous studies have used traditional regression-based techniques to estimate CEC from more easily measurable soil parameters (e.g., soil texture, organic matter (OM), and pH). However, these models may not be able to adequately capture the complex and highly nonlinear relationship between CEC and its influential soil variables. In this study, Genetic Expression Programming (GEP) and Multivariate Adaptive Regression Splines (MARS) were employed to estimate CEC from more readily measurable soil physical and chemical variables (e.g., OM, clay, and pH) by developing functional relations. The GEP- and MARS-based functional relations were tested at two field sites in Iran. Results showed that GEP and MARS can provide reliable estimates of CEC. Also, it was found that the MARS model (with root-mean-square-error (RMSE) of 0.318 Cmol + kg −1 and correlation coefficient ( R 2 ) of 0.864) generated slightly better results than the GEP model (with RMSE of 0.270 Cmol + kg −1 and R 2 of 0.807). The performance of GEP and MARS models was compared with two existing approaches, namely artificial neural network (ANN) and multiple linear regression (MLR). The comparison indicated that MARS and GEP outperformed the MLP model, but they did not perform as good as ANN. Finally, a sensitivity analysis was conducted to determine the most and the least influential variables affecting CEC. It was found that OM and pH have the most and least significant effect on CEC, respectively.

Comparison of time-drawdown plots from graphical, NLP, and ACO techniques for sample confined (datasets C3), unconfined (datasets U4), and leaky (dataset L1) aquifers. We proposed an Ant Colony Optimization (ACO) to estimate the aquifer hydraulic properties.ACO is efficient, and reliable for estimating various aquifer hydraulic parameters from the time-drawdown data.ACO provides better estimation of aquifer properties than the graphical and NLP techniques. An accurate estimation of aquifer hydraulic parameters is required for groundwater modeling and proper management of vital groundwater resources. In situ measurements of aquifer hydraulic parameters are expensive and difficult. Traditionally, these parameters have been estimated by graphical methods that are approximate and time-consuming. As a result, nonlinear programming (NLP) techniques have been used extensively to estimate them. Despite the outperformance of NLP approaches over graphical methods, they tend to converge to local minima and typically suffer from a convergence problem. In this study, Genetic Algorithm (GA) and Ant Colony Optimization (ACO) methods are used to identify hydraulic parameters (i.e., storage coefficient, hydraulic conductivity, transmissivity, specific yield, and leakage factor) of three types of aquifers namely, confined, unconfined, and leaky from real time-drawdown pumping test data. The performance of GA and ACO is also compared with that of graphical and NLP techniques. The results show that both GA and ACO are efficient, robust, and reliable for estimating various aquifer hydraulic parameters from the time-drawdown data and perform better than the graphical and NLP techniques. The outcomes also indicate that the accuracy of GA and ACO is comparable. Comparing the running time of various utilized methods illustrates that ACO converges to the optimal solution faster than other techniques, while the graphical method has the highest running time.

This study investigated the capability of M Model Tree M MT to predict reference evapotranspiration ET M MT was trained and tested with climatic data from eight weather stations located in coastal areas of Iran for the years It was validated with climatic data from seven California Irrigation Management Information System CIMIS weather stations for the year Four different data combinations were utilized to train test and validate the M MT model These were daily mean air temperature wind speed relative humidity and solar radiation configuration daily mean air temperature and solar radiation configuration daily mean air temperature and relative humidity configuration and daily maximum minimum and mean air temperature and extraterrestrial radiation configuration The Penman Monteith PM equation was used as a standard method to provide target ET values Mean absolute error MAE root mean square error RMSE and the coefficient of determination R were used to evaluate the performance of the M MT models developed with different input configurations Results indicated that M MT was able to successfully estimate ET Configuration provided the most accurate results Configuration showed to have the variables that have a greater influence on ET than configuration Configuration performed the worst MAE of ET estimates from M MT was respectively and lower than that of M MT M MT and M MT when the model is validated in California Also RMSE from M MT was and smaller than that of M MT M MT and M MT respectively

Remotely sensed Land Surface Temperature (LST) and Fraction of Photosynthetically Active Radiation absorbed by vegetation (FPAR) are assimilated, respectively, into the Surface Energy Balance (SEB) equation and a Vegetation Dynamics Model (VDM) in order to estimate surface fluxes and vegetation dynamics. The problem is posed in terms of three unknown and dimensionless parameters: (1) neutral bulk heat transfer coefficient, which scales the sum of turbulent heat fluxes, (2) soil and canopy evaporative fractions that characterize partitioning among the turbulent heat fluxes over soil and vegetation, and (3) specific leaf area, which captures seasonal phenology and vegetation dynamics. The model is applied over the Gourma site in Mali, the northern region of the West African Monsoon (WAM) domain. The application of the model over the Gourma site shows that spaceborne LST observations can be used to constrain the SEB equation and obtain its key two unknown parameters (i.e., neutral bulk heat transfer coefficient and evaporative fraction). We demonstrate that the spatial patterns of estimated neutral bulk heat transfer coefficient and evaporative fraction resemble, respectively, those of independently observed vegetation index and soil moisture. The framework also yields estimates of surface energy balance components. The daily sensible, latent, and ground heat flux estimates at the Agoufou site that is located in the south of the Gourma region have, respectively, a root-mean-square error (RMSE) of 53.6, 34.4, and 45.1 Wm−2. The daily sensible heat flux estimates at the Bamba site, which is located in the north of the Gourma domain, have a RMSE of 42.6 Wm−2. The results also show that remotely sensed FPAR observations can constrain the VDM and retrieve its main unknown parameter (specific leaf area) over large-scale domains without costly in situ measurements. The results indicate that the estimated specific leaf area values vary reasonably with the expected influential environmental variables such as precipitation, air temperature, and solar radiation. Assimilating FPAR observations into the VDM can also provide an estimate of Leaf Area Index (LAI) dynamics. The estimated LAI values are comparable in magnitude, spatial pattern and temporal evolution with satellite retrievals.

An accurate estimation of half-cone geometry (i.e., volume and length) created by pressure flushing operation in dam reservoirs is required for sediment management in the reservoir storage. In this study, two artificial intelligence techniques namely, Artificial Neural Network (ANN) and Adaptive Neuro-fuzzy Inference System (ANFIS) were utilized to estimate the volume and length of flushing half-cone based on influential variables, i.e., mean flow velocity through bottom outlet (u), water depth in reservoir (H"w), mean grain diameter of deposited sediments (d"5"0), thickness of deposited sediment (H"s) and bottom outlet diameter (D). Experimental data in both dimensional and non-dimensional forms were used to train and test ANN and ANFIS models. The results of the intelligence-based models were also compared with those of existing studies. The outcomes indicated that both ANN and ANFIS models predict the volume and length of flushing half-cone more accurately than existing studies. Also, it was found that the ANN model provides a better estimation of the geometry of flushing half-cone compared to the ANFIS model. Finally, sensitivity analysis was conducted to determine the most and the least influential variables affecting the flushing half-cone geometry. It was found that the sediment characteristics (H"s and d"5"0) and fluids properties (H"w and u) have respectively the most and the least effect on flushing half-cone volume and length.

An ensemble-based data assimilation approach is developed to characterize the snow water equivalent (SWE) and underlying soil freeze-thaw state (including the soil surface temperature and both soil ice and liquid water content) using multifrequency passive and active microwave remote-sensing measurements. Its feasibility was examined using a synthetic test where passive microwave (1.4, 18.7, and 36.5 GHz) and active microwave [L-band (1.4 GHz), C-band (5.4 GHz), and Ku-band (12 GHz)] measurements at the point scale were individually and simultaneously assimilated to estimate the SWE and soil freeze-thaw state using an Ensemble Batch Smoother framework. The contribution of each channel in retrieving the true SWE, soil surface temperature, soil liquid water and ice content was investigated at the local-scale observation site of the National Aeronautics and Space Administration Cold Land Processes Experiments Field Campaign in northern Colorado during both the snow accumulation (Fall 2002-Winter 2003) and melt (Spring 2003) periods. All of the utilized passive and active measurements were found to contain valuable and complementary information for characterizing the SWE and freeze-thaw state of the underlying soil. L-band measurements were most effective for soil freeze-thaw state estimation, whereas higher frequencies were more effective at SWE characterization. In addition, results from the simultaneous assimilation of passive and active microwave data were compared to those from a modeling approach without assimilating microwave data (open loop). It was found that assimilating both passive and active microwave data decreased the errors that are associated with the open-loop approach. Finally, passive and active measurements were undersampled as expected from the overpasses of current and future satellite platforms. It was observed that the developed method can reliably estimate the soil freeze-thaw state and SWE, even with measurement sequences anticipated from the temporal frequency of existing and future satellites such as the Special Sensor Microwave/Imager, Soil Moisture Active Passive Mission, and Cold Regions Hydrology High-Resolution Observatory.

[1] A variational data assimilation model is developed to estimate surface energy fluxes from remotely sensed land surface temperature (LST). Components of the surface energy balance (sensible, latent, and ground heat fluxes) have different degrees of efficiency in dissipating available energy. LST is the state variable of the surface energy balance (SEB). Land surface models that capture the exchange and storage of energy in the soil and vegetation media use LST as a prognostic variable. Sequences of LST measurements implicitly contain information on partitioning of available energy among the components of SEB. In this study, we focus on the estimation of the sum of the turbulent fluxes as well as the partitioning among them. Two dimensionless parameters are used to characterize the sum and the partitioning. Using LST observations from a constellation of satellites, these parameters are mapped over a large region. The remotely sensed LST is assimilated to the heat diffusion equation within the SEB framework. In addition, a model error term is added to the SEB equation such that the variational data assimilation scheme includes model uncertainty as well as observation error. The framework is tested over the Southern Great Plains region. The mapped results of the surface evaporation estimation are used to study the surface control on evaporation. Independent mapped soil moisture estimates from an airborne microwave campaign are used. The dependence of the evaporation control-soil moisture relationship on vegetation cover and plant functional types over large regions is examined in this first and exploratory study.

Recently, numerous studies have focused on the estimation of surface energy flux components, based on the assimilation of land surface temperature (LST) within a variational data assimilation (VDA) framework. Unlike the previous investigations based on the force-restore equation, in this study, the full heat diffusion equation is employed in the VDA scheme as an adjoint (constraint). In addition, a model error term is added to the surface energy balance (SEB) equation and the VDA scheme to include the model uncertainty. Both VDA schemes (with and without the model uncertainty) are tested over the First International Satellite Land Surface Climatology Project Field Experiment (FIFE) site. The comprehensive comparisons between the present model (with heat diffusion equation) and previous model (with the force-restore equation) demonstrate that the present model will decrease the phase error associated with the ground heat flux diurnal cycle, and improve the evaporative fraction and heat fluxes estimation. The numerical examples also conclude that the errors caused by model structures and noisy data in the SEB equation can be detected and quantified in the present model (with model uncertainty).

The accurate estimation of available soil phosphorus (ASP) is required for the effective management of environmental resources and improving agricultural productivity. Direct measurement of ASP at large scales is time-consuming and expensive. Therefore, many empirical models or pedotransfer functions (PTFs) have been developed to obtain ASP. This study introduces a new model that estimates ASP from easily measurable soil properties by using the multi-objective group method of data handling (mGMDH) approach. Organic carbon, clay content, calcium carbonate equivalent, pH and ASP were obtained from 100 soil samples, which were collected by the stratified random sampling strategy in Mashhad Plain, Khorasan-e-Razavi Province, Northeast Iran. The suggested model for the prediction of ASP generated satisfactory results in terms of the statistical performance criteria. The coefficient of determination (R2) and mean square error (MSE) values of the mGMDH model in training step were 0.606 and 4.9671 mg kg−1, respectively, whereas those were 0.504 and 5.6685 mg kg−1, in the testing step. It was concluded that only 50 % of variations of ASP could be described by selecting easy-to-measure soil variables and owing to the key role of phosphorus in soil fertility and precision agriculture, the other related factors such as terrain attributes and fertilizer management should also be considered as they influence the distribution of the soil phosphorus.

Most models used in land surface hydrology, vadose zone hydrology, and hydro-climatology require an accurate representation of soil thermal properties (soil thermal conductivity and volumetric heat capacity). Various empirical relations have been suggested to estimate soil thermal properties. However, they require many input parameters such as soil texture, mineralogical composition, porosity and water content, which are not always available from laboratory experiments and field measurements. In this paper, to overcome the above challenge, a hybrid numerical method, Genetic Algorithm-Finite Difference (GA-FD), is proposed to estimate soil thermal properties using land surface temperature (LST) as the only input. The genetic algorithm (GA) optimization method coupled with the finite difference (FD) modeling technique is a viable hybrid approach for estimating soil thermal properties. The finite difference method is employed to solve the heat diffusion equation and simulate LST, while a robust optimization technique (GA) is used to retrieve soil thermal properties by minimizing the difference between observed and simulated LST. Furthermore, a generalization of the hybrid model is developed for inhomogeneous soil, in which soil thermal properties are not constant throughout the soil slab. The proposed model is applied to the First International Satellite Land Surface Climatology Project (ISLSCP) Field Experiment (FIFE). The results show that the proposed hybrid numerical method is able to estimate soil thermal properties accurately, and therefore effectively eliminate the need for the unavailable soil parameters which are required by empirical methods for determining the soil thermal conductivity and volumetric heat capacity. Remarkably, the temporal variation of the retrieved soil thermal conductivity is consistent with the volumetric water content, even though no water content information is used in the model.

The Greenland ice sheet (GrIS) has been the focus of climate studies due to its considerable impact on sea level rise. Accurate estimates of surface mass fluxes would contribute to understanding the cause of its recent changes and would help to better estimate the past, current and future contribution of the GrIS to sea level rise. Though the estimates of the GrIS surface mass fluxes have improved significantly over the last decade, there is still considerable disparity between the results from different methodologies (e.g., Rae et al., 2012; Vernon et al., 2013). The data assimilation approach can merge information from different methodologies in a consistent way to improve the GrIS surface mass fluxes. In this study, an ensemble batch smoother data assimilation approach was developed to assess the feasibility of generating a reanalysis estimate of the GrIS surface mass fluxes via integrating remotely sensed ice surface temperature measurements with a regional climate model (a priori) estimate. The performance of the proposed methodology for generating an improved posterior estimate was investigated within an observing system simulation experiment (OSSE) framework using synthetically generated ice surface temperature measurements. The results showed that assimilation of ice surface temperature time series were able to overcome uncertainties in near-surface meteorological forcing variables that drive the GrIS surface processes. Our findings show that the proposed methodology is able to generate posterior reanalysis estimates of the surface mass fluxes that are in good agreement with the synthetic true estimates. The results also showed that the proposed data assimilation framework improves the root-mean-square error of the posterior estimates of runoff, sublimation/evaporation, surface condensation, and surface mass loss fluxes by 61, 64, 76, and 62 %, respectively, over the nominal a priori climate model estimates.

The phenomenon of the hydraulic jump is so complex that despite considerable laboratory and prototype studies, estimation of its main characteristics in a generalized and accurate form is still difficult. The Artificial Neural Network (ANN) approach aims at limiting the needs for costly and timeconsuming experiments. In this study, twoANN models, multi-layer perceptron using back propagation algorithm (MLP/BP) and radial basis function using orthogonal least-squares algorithm (RBF/OLS), were used to predict the roller length, sequent depth, and the relative energy loss of the B-jump. Based on a pre-specified range of jump parameters, the input vectors include: upstream bed slope (tan ?), inflow depth h1, approach velocity V1 and elevation of jump toe from the datum plane z1, generated from the experimental data of Hager (J. Hydraul. Res., IAHR, 26(5), 539–558, 1988) and Kawagoshi and Hager (J. Hydraul. Res., IAHR, 28(4), 461–480, 1990). Once the network is trained to an acceptable level of accuracy, it pr...

An accurate estimation of scour depth around piles is important for coastal and ocean engineers involved in the design of marine structures. Owing to the complexity of the problem, most conventional approaches are often unable to provide sufficiently accurate results. In this paper, an alternative attempt is made herein to develop adaptive neuro-fuzzy inference system (ANFIS) models for predicting scour depth as well as scour width for a group of piles supporting a pier. The ANFIS model provides the system identification and interpretability of the fuzzy models and the learning capability of neural networks in a single system. Two combinations of input data were used in the analyses to predict scour depth: the first input combination involves dimensional parameters such as wave height, wave period, and water depth, while the second combination contains nondimensional numbers including the Reynolds number, the Keulegan–Carpenter number, the Shields parameter and the sediment number. The test results show that ANFIS performs better than the existing empirical formulae. The ANFIS predicts scour depth better when it is trained with the original (dimensional) rather than the nondimensional data. The depth of scour was predicted more accurately than its width. A sensitivity analysis showed that scour depth is governed mainly by the Keulegan–Carpenter number, and wave height has a greater influence on scour depth than the other independent parameters.

The mechanism of flow around a pier structure is so complicated that it is difficult to establish a general empirical model to provide accurate estimation for scour. Interestingly, each of the proposed empirical formula yields good results for a particular data set. Hence, in this study, alternative approaches, artificial neural networks (ANNs) and adaptive neuro-fuzzy inference system (ANFIS), are proposed to estimate the equilibrium and time-dependent scour depth with numerous reliable data base. Two ANN models, multi-layer perception using back-propagation algorithm (MLP/BP) and radial basis using orthogonal least-squares algorithm (RBF/OLS), were used. The equilibrium scour depth was modeled as a function of five variables; flow depth, mean velocity, critical flow velocity, mean grain diameter and pier diameter. The time variation of scour depth was also modeled in terms of equilibrium scour depth, equilibrium scour time, scour time, mean flow velocity and critical flow velocity. The training and testing data are selected from the experimental data of several valuable references. Numerical tests indicate that MLP/BP model provide a better prediction of scour depth than RBF/OLS and ANFIS models as well as the previous empirical approaches. Finally, sensitivity analysis shows that pier diameter has a greater influence on equilibrium scour depth than the other independent parameters.

The Greenland ice sheet (GrIS) has been the focus of climate studies due to its considerable impact on sea level rise. Accurate estimates of surface mass fluxes would contribute to understanding the cause of its recent unprecedented changes and would help to better estimate the past, current and future contribution of the GrIS to sea level rise. Though the estimates of the GrIS surface mass balance have improved significantly over the last decade, there is considerable disparity between the results from different methodologies that need to be addressed. In this study, an Ensemble Batch Smoother data assimilation approach was developed to assess the feasibility of generating a reanalysis estimate of the GrIS surface mass fluxes via integrating remotely sensed ice surface temperature measurements with a regional climate model (a priori) estimate. The performance of the proposed methodology for generating an improved posterior estimate was investigated within an observing system simulation experiment (OSSE) framework using synthetically generated ice surface temperature measurements. The results showed that assimilation of ice surface temperature time series were able to overcome uncertainties in near-surface meteorological forcing variables that drive the GrIS surface processes. Our findings show that the proposed methodology is able to generate posterior reanalysis estimates of the surface mass fluxes that are in good agreement with the synthetic true estimates. The results also showed that the proposed data assimilation framework improves the root-mean-square-error (RMSE) of the posterior estimates of runoff, sublimation/evaporation, surface condensation and surface mass loss fluxes by 61, 64, 76, and 62 % respectively over the nominal a priori climate model estimates.

Recently, a number of studies have focused on estimating surface turbulent heat fluxes via assimilation of sequences of land surface temperature (LST) observations into variational data assimilation (VDA) schemes. Using the full heat diffusion equation as a constraint, the surface energy balance equation can be solved via assimilation of sequences of LST within a VDA framework. However, the VDA methods have been tested only in limited field sites that span only a few climate and land use types. Hence, in this study, combined-source (CS) and dual-source (DS) VDA schemes are tested extensively over six FluxNet sites with different vegetation covers (grassland, cropland, and forest) and climate conditions. The CS model groups the soil and canopy together as a single source and does not consider their different contributions to the total turbulent heat fluxes, while the DS model considers them to be different sources. LST data retrieved from the Geostationary Operational Environmental Satellites are assimilated into these two VDA schemes. Sensible and latent heat flux estimates from the CS and DS models are compared with the corresponding measurements from flux tower stations. The results indicate that the performance of both models at dry, lightly vegetated sites is better than that at wet, densely vegetated sites. Additionally, the DS model outperforms the CS model at all sites, implying that the DS scheme is more reliable and can characterize the underlying physics of the problem better.

[1] Recently, a number of studies have assimilated land surface temperature (LST) within a variational data assimilation (VDA) framework to estimate turbulent heat fluxes. These VDA models have mainly considered soil and vegetation as a combined source (CS) and have not accounted for the difference between soil and canopy temperatures and turbulent exchange rates, although soil and canopy can exhibit very different behaviors. Hence, in this study the contribution of soil and canopy to the LST and turbulent heat fluxes is taken into account separately by developing a dual-source (DS) VDA model. The unknown model parameters are the neutral bulk heat transfer coefficient (that scales the sum of turbulent heat fluxes) and the evaporative fractions for soil and canopy (which represent partitioning among the turbulent fluxes over soil and vegetation). The model as developed has been tested with area-averaged measurements of turbulent heat fluxes obtained from the First International Satellite Land Surface Climatology Project Field Experiment (FIFE) during the summers of 1987 and 1988. The results show that the predicted turbulent heat fluxes match well with observations. For FIFE 1987 (1988), the half-hourly latent heat flux estimates from the new model have a root-mean square-error (RMSE) of 57.4 Wm−2 (66.8 Wm−2), which represents a significant improvement over the previous study.

[1] The partitioning of available energy into dissipative fluxes over land surfaces is dependent on the state variable of the surface energy balance (land surface temperature) and the state variable of the surface water balance (soil moisture). The direct measurement of the turbulent fluxes is achieved with in situ instruments at tower sites. These point-scale measurements are sparsely distributed. Broader scale mapping of the turbulent fluxes is mostly dependent on land surface temperature (LST) and optical/infrared vegetation that can be sensed remotely. There are several data assimilation approaches currently in use that intake sequences of daytime LST that attain different diurnal amplitudes depending on available energy and the relative efficiency of surface energy balance to infer the magnitude of surface flux components such as latent and sensible heat flux. In this study we perform stability analysis on the evolution of LST in order to provide insights into the physical bases for why LST variations can be used to diagnose surface energy balance (SEB) components. The derived relative efficiencies of SEB components in dissipating available energy at the land surface are tested using two field experiment measurements. The results show that the theoretically derived relative efficiencies of SEB components agree well with field observations. The study provides insight into how LST sequences implicitly contain the signature of partitioning of available energy among SEB components and can be used to infer their magnitudes.