Forecasting Hourly Roadside Particulate Matter in Taipei County of Taiwan Based on First‐Order and One‐Variable Grey Model

In this study, seven types of first‐order and one‐variable grey differential equation model (abbreviated as GM (1, 1) model) were used to forecast hourly roadside particulate matter (PM) including PM₁₀ and PM_2.5 concentrations in Taipei County of Taiwan. Their forecasting performance was also compared. The results indicated that the minimum mean absolute percentage error (MAPE), mean squared error (MSE), root mean squared error (RMSE), and maximum correlation coefficient (R) was 11.70%, 60.06, 7.75, and 0.90%, respectively when forecasting PM₁₀. When forecasting PM_2.5, the minimum MAPE, MSE, RMSE, and maximum R‐value of 16.33%, 29.78, 5.46, and 0.90, respectively could be achieved. All statistical values revealed that the forecasting performance of GM (1, 1, x⁽⁰⁾), GM (1, 1, a), and GM (1, 1, b) outperformed other GM (1, 1) models. According to the results, it revealed that GM (1, 1) was an efficiently early warning tool for providing PM information to the roadside inhabitants.     

Modelling soil moisture in a high‐latitude landscape using LiDAR and soil data

Soil moisture has a pronounced effect on earth surface processes. Global soil moisture is strongly driven by climate, whereas at finer scales, the role of non‐climatic drivers becomes more important. We provide insights into the significance of soil and land surface properties in landscape‐scale soil moisture variation by utilizing high‐resolution light detection and ranging (LiDAR) data and extensive field investigations. The data consist of 1200 study plots located in a high‐latitude landscape of mountain tundra in north‐western Finland. We measured the plots three times during growing season 2016 with a hand‐held time‐domain reflectometry sensor. To model soil moisture and its temporal variation, we used four statistical modelling methods: generalized linear models, generalized additive models, boosted regression trees, and random forests. The model fit of the soil moisture models were R² = 0.60 and root mean square error (RMSE) 8.04 VWC% on average, while the temporal variation models showed a lower fit of R² = 0.25 and RMSE 13.11 CV%. The predictive performances for the former were R² = 0.47 and RMSE 9.34 VWC%, and for the latter R² = 0.01 and RMSE 15.29 CV%. Results were similar across the modelling methods, demonstrating a consistent pattern. Soil moisture and its temporal variation showed strong heterogeneity over short distances; therefore, soil moisture modelling benefits from high‐resolution predictors, such as LiDAR based variables. In the soil moisture models, the strongest predictor was SAGA (System for Automated Geoscientific Analyses) wetness index (SWI), based on a 1 m² digital terrain model derived from LiDAR data, which outperformed soil predictors. Thus, our study supports the use of LiDAR based SWI in explaining fine‐scale soil moisture variation. In the temporal variation models, the strongest predictor was the field‐quantified organic layer depth variable. Our results show that spatial soil moisture predictions can be based on soil and land surface properties, yet the temporal models require further investigation. Copyright © 2017 John Wiley & Sons, Ltd.     

Artificial Neural Network Prediction of Tropospheric Ozone Concentrations in Istanbul,Turkey

Tropospheric (ground‐level) ozone has adverse effects on human health and environment. In this study, next day's maximum 1‐h average ozone concentrations in Istanbul were predicted using multi‐layer perceptron (MLP) type artificial neural networks (ANNs). Nine meteorological parameters and nine air pollutant concentrations were utilized as inputs. The total 578 datasets were divided into three groups: training, cross‐validation, and testing. When all the 18 inputs were used, the best performance was obtained with a network containing one hidden layer with 24 neurons. The transfer function was hyperbolic tangent. The correlation coefficient (R), mean absolute error (MAE), root mean squared error (RMSE), and index of agreement or Willmott's Index (d₂) for the testing data were 0.90, 8.78 µg/m³, 11.15 µg/m³, and 0.95, respectively. Sensitivity analysis has indicated that the persistence information (current day's maximum and average ozone concentrations), NO concentration, average temperature, PM₁₀, maximum temperature, sunshine time, wind direction, and solar radiation were the most important input parameters. The values of R, MAE, RMSE, and d₂ did not change considerably for the MLP model using only these nine inputs. The performances of the MLP models were compared with those of regression models (i.e., multiple linear regression and multiple non‐linear regression). It has been found that there was no significant difference between the ANN and regression modeling techniques for the forecasting of ozone concentrations in Istanbul.     

Bi‐Solute Sorption of Estrogens at Low Concentration: Two‐ and Three‐Parametric Analysis

Competitive sorption of estriol (E3) and 17α‐ethinylestradiol (EE2) was studied on activated charcoal. Better sorption of E3 (88.9%) and EE2 (69.5%) was observed with single‐solute sorption system than with bi‐solute sorption system. Single‐solute sorption kinetics of E3 and EE2 were evaluated with two (Langmuir and Freundlich) and three (dual mode and Song) parameter models. Freundlich model (R², 0.9915 (E3); 0.9875 (EE2)) showed good prediction compared to other models for single‐solute sorption. Adsorption capacity documented reduced efficacy (86.4% (E3); 65.9% (EE2)) due to induced competitive behavior between the estrogens in aqueous phase. Bi‐solute adsorption kinetics of E3 and EE2 were described by IAST with two and three parameter models. Among these models, IAST‐Freundlich model (R², 0.9765 (E3); 0.9985 (EE2)) was best in predicting bi‐solute sorption of E3 and EE2 by activated charcoal. All these models showed favorable representation of both single‐ and bi‐solute sorption behaviors.     

Multi-step water quality forecasting using a boosting ensemble multi-wavelet extreme learning machine model

The use of electrical conductivity (EC) as a water quality indicator is useful for estimating the mineralization and salinity of water. The objectives of this study were to explore, for the first time, extreme learning machine (ELM) and wavelet-extreme learning machine hybrid (WA-ELM) models to forecast multi-step-ahead EC and to employ an integrated method to combine the advantages of WA-ELM models, which utilized the boosting ensemble method. For comparative purposes, an adaptive neuro-fuzzy inference system (ANFIS) model, and a WA-ANFIS model, were also developed. The study area was the Aji-Chay River at the Akhula hydrometric station in Northwestern Iran. A total of 315 monthly EC (µS/cm) datasets (1984–2011) were used, in which the first 284 datasets (90% of total datasets) were considered for training and the remaining 31 (10% of total datasets) were used for model testing. Autocorrelation function (ACF) and partial autocorrelation function (PACF) demonstrated that the 6-month lags were potential input time lags. The results illustrated that the single ELM and ANFIS models were unable to forecast the multi-step-ahead EC in terms of root mean square error (RMSE), coefficient of determination (R²) and Nash–Sutcliffe model efficiency coefficient (NSC). To develop the hybrid WA-ELM and WA-ANFIS models, the original time series of lags as inputs, and time series of 1, 2 and 3 month-step-ahead EC values as outputs, were decomposed into several sub-time series using different maximal overlap discrete wavelet transform (MODWT) functions, namely Daubechies, Symlet, Haar and Coiflet of different orders at level three. These sub-time series were then used in the ELM and ANFIS models as an input dataset to forecast the multi-step-ahead EC. The results indicated that single WA-ELM and WA-ANFIS models performed better than any ELM and ANFIS models. Also, WA-ELM models outperformed WA-ANFIS models. To develop the boosting multi-WA-ELM and multi-WA-ANFIS ensemble models, a least squares boosting (LSBoost) algorithm was used. The results showed that boosting multi-WA-ELM and multi-WA-ANFIS ensemble models outperformed the individual WA-ELM and WA-ANFIS models.     

Adsorption Kinetic Modeling of Safranin onto Rice Husk Biomatrix Using Pseudo‐first‐ and Pseudo‐second‐order Kinetic Models: Comparison of Linear and Non‐linear Methods

Batch kinetic studies were carried out for the removal of safranin from aqueous solution using a biomatrix prepared from rice husk. The adsorption kinetic data were modeled using the pseudo‐first‐order and pseudo‐second‐order kinetic equations. The linear and non‐linear forms of these two widely used kinetic models were compared in this study. In order to determine the best‐fitting equation, the coefficient of determination (r²), the sum of the squares of the errors (SSE), sum of the absolute errors (SAE), average relative error (ARE), hybrid fractional error function (HYBRID), Marquardt's percent standard deviation (MPSD), and the Chi‐squared test (χ²) were used as error analysis methods. Results showed that the non‐linear forms of pseudo‐first‐order and pseudo‐second‐order models were more suitable than the linear forms for fitting the experimental data. Non‐linear method is thus more appropriate for estimating the kinetic parameters and should primarily be used to describe adsorption kinetics.     

Evaluating best evaporation estimate model for water surface evaporation in semi‐arid region,India

Evaluating performances of four commonly used evaporation estimate methods, namely; Bowen ratio energy balance (BREB), mass transfer (MT), Priestley–Taylor (PT) and pan evaporation (PE), based on 4 years experimental data, the most effective and the reliable evaporation estimates model for the semi‐arid region of India has been derived. The various goodness‐of‐fit measures, such as; coefficient of determination (R²), index of agreement (D), root mean square error (RMSE), and relative bias (RB) have been chosen for the performance evaluation. Of these models, the PT model has been found most promising when the Bowen ratio, β is known a priori, and based on its limited data requirement. The responses of the BREB, the PT, and the PE models were found comparable to each other, while the response of the MT model differed to match with the responses of the other three models. The coefficients, β of the BREB, µ of the MT, α of the PT and K_P of the PE model were estimated as 0·07, 2·35, 1·31 and 0·65, respectively. The PT model can successfully be extended for free water surface evaporation estimates in semi‐arid India. A linear regression model depicting relationship between daily air and water temperature has been developed using the observed water temperatures and the corresponding air temperatures. The model helped to generate unrecorded water temperatures for the corresponding ambient air temperatures. Copyright © 2007 John Wiley & Sons, Ltd.     

Evaluating the SEBS‐estimated evaporative fraction from MODIS data for a complex underlying surface

With the complex nature of land surfaces, more attention should be paid to the performance of remotely sensed models to estimate evapotranspiration from moderate and low spatial resolution data. Taking into account the characteristic of a stable evaporative fraction (EF) in the daytime, this paper uses the surface energy balance system (SEBS) to estimate the EF from MODIS data for a subtropical evergreen coniferous plantation in southern China and evaluates the stability of the SEBS model in estimating the EF under complex surface conditions. The results show that the SEBS‐estimated EF is larger than the measured EF partly because of the serious lack of energy‐balance closure. This difference can be largely reduced by the residual energy correction method. More evaporative land cover within the MODIS pixel is a main reason for the overestimated EF. SEBS underestimates sensible heat flux, and the underestimation of surface available energy also contributes to the overestimation of the EF. The EF estimated from MODIS/Terra data is in agreement with that from MODIS/Aqua data with a coefficient of determination (R²) of 0.552, a mean bias error (BIAS) of 0.028, and a root mean square error (RMSE) of 0.079, which is consistent with the result from in situ measurements. In addition, the estimation of surface available energy from remotely sensed data is evaluated on this complex underlying surface. Compared with in situ measurements, the available energy is underestimated by 28 W m^?2 with an RMSE = 50 W m^?2 and an R² = 0.87. Copyright © 2012 John Wiley & Sons, Ltd.     

A comparison of performance of several artificial intelligence methods for forecasting monthly discharge time series

Developing a hydrological forecasting model based on past records is crucial to effective hydropower reservoir management and scheduling. Traditionally, time series analysis and modeling is used for building mathematical models to generate hydrologic records in hydrology and water resources. Artificial intelligence (AI), as a branch of computer science, is capable of analyzing long-series and large-scale hydrological data. In recent years, it is one of front issues to apply AI technology to the hydrological forecasting modeling. In this paper, autoregressive moving-average (ARMA) models, artificial neural networks (ANNs) approaches, adaptive neural-based fuzzy inference system (ANFIS) techniques, genetic programming (GP) models and support vector machine (SVM) method are examined using the long-term observations of monthly river flow discharges. The four quantitative standard statistical performance evaluation measures, the coefficient of correlation (R), Nash–Sutcliffe efficiency coefficient (E), root mean squared error (RMSE), mean absolute percentage error (MAPE), are employed to evaluate the performances of various models developed. Two case study river sites are also provided to illustrate their respective performances. The results indicate that the best performance can be obtained by ANFIS, GP and SVM, in terms of different evaluation criteria during the training and validation phases.     

Daily streamflow modelling and assessment based on the curve‐number technique

A cell‐based long‐term hydrological model (CELTHYM) that can be integrated with a geographical information system (GIS) was developed to predict continuous stream flow from small agricultural watersheds. The CELTHYM uses a cell‐by‐cell soil moisture balance approach. For surface runoff estimation, the curve number technique considering soil moisture on a daily basis was used, and release rate was used to estimate baseflow. Evapotranspiration was computed using the FAO modified Penman equation that considered land‐use‐based crop coefficients, soil moisture and the influence of topography on radiation. A rice paddy field water budget model was also adapted for the specific application of the model to East Asia. Model sensitivity analysis was conducted to obtain operational information about the model calibration parameters. The CELTHYM was calibrated and verified with measured runoff data from the WS#1 and WS#3 watersheds of the Seoul National University, Department of Agricultural Engineering, in Hwaseong County, Kyounggi Province, South Korea. The WS#1 watershed is comprised of about 35·4% rice paddy fields and 42·3% forest, whereas the WS#3 watershed is about 85·0% forest and 11·5% rice paddy fields. The CELTHYM was calibrated for the parameter release rate, K, and soil moisture storage coefficient, STC, and results were compared with the measured runoff data for 1986. The validation results for WS#1 considering all daily stream flow were poor with R², E² and RMSE having values of 0·40, ?6·63 and 9·69 (mm), respectively, but validation results for days without rainfall were statistically significant (R² = 0·66). Results for WS#3 showed good agreement with observed data for all days, and R², E² and RMSE were 0·92, 0·91 and 2·23 (mm), respectively, suggesting potential for CELTHYM application to other watersheds. The direct runoff and water balance components for watershed WS#1 with significant areas of paddy fields did not perform well, suggesting that additional study of these components is needed. Copyright © 2002 John Wiley & Sons, Ltd.     

Developing an empirical model of canopy water flux describing the common response of transpiration to solar radiation and VPD across five contrasting woodlands and forests

A modified Jarvis–Stewart model of canopy transpiration (E_c) was tested over five ecosystems differing in climate, soil type and species composition. The aims of this study were to investigate the model's applicability over multiple ecosystems; to determine whether the number of model parameters could be reduced by assuming that site‐specific responses of E_c to solar radiation, vapour pressure deficit and soil moisture content vary little between sites; and to examine convergence of behaviour of canopy water‐use across multiple sites. This was accomplished by the following: (i) calibrating the model for each site to determine a set of site‐specific (SS) parameters, and (ii) calibrating the model for all sites simultaneously to determine a set of combined sites (CS) parameters. The performance of both models was compared with measured E_c data and a statistical benchmark using an artificial neural network (ANN). Both the CS and SS models performed well, explaining hourly and daily variation in E_c. The SS model produced slightly better model statistics [R² = 0.75–0.91; model efficiency (ME) = 0.53–0.81; root mean square error (RMSE) = 0.0015–0.0280 mm h^‐1] than the CS model (R² = 0.68–0.87; ME = 0.45–0.72; RMSE = 0.0023–0.0164 mm h^‐1). Both were highly comparable with the ANN (R² = 0.77–0.90; ME = 0.58–0.80; RMSE = 0.0007–0.0122 mm h^‐1). These results indicate that the response of canopy water‐use to abiotic drivers displayed significant convergence across sites, but the absolute magnitude of E_c was site specific. Period totals estimated with the modified Jarvis–Stewart model provided close approximations of observed totals, demonstrating the effectiveness of this model as a tool aiding water resource management. Analysis of the measured diel patterns of water use revealed significant nocturnal transpiration (9–18% of total water use by the canopy), but no Jarvis–Stewart formulations are able to capture this because of the dependence of water‐use on solar radiation, which is zero at night. Copyright © 2012 John Wiley & Sons, Ltd.     

Start‐Up Study on Biohydrogen from Palm Oil Mill Effluent in a Pilot‐Scale Reactor

A start‐up study for biohydrogen production from palm oil mill effluent (POME) is carried out in a pilot‐scale up‐flow anaerobic sludge blanket fixed‐film reactor (UASFF). A substrate with a chemical oxygen demand (COD) of 30 g L^?1 is used, starting with molasses solution for 30 days and followed by a 10% v/v increment of POME/molasses ratio. At 100% POME, a hydrogen content of 80%, hydrogen production rate of 36 L H₂ per day, and maximum COD removal of 48.7% are achieved. Bio‐kinetic coefficients of Monod, first‐order, Grau second‐order, and Stover‐Kincannon kinetic models are calculated to describe the performance of the system. The steady‐state data with 100% POME shows that Monod and Stover‐Kincannon models with bio‐kinetic coefficients of half‐velocity constant (K_s) of 6000 mg COD L^?1, microbial decay rate (K_d) of 0.0015 per day, growth yield constant (Y) of 0.786 mg volatile suspended solids (VSS)/mg COD, specific biomass growth rate (μ_max) of 0.568 per day, and substrate consumption rate of (U_max) 3.98 g/L day could be considered as superior models with correlation coefficients (R²) of 0.918 and 0.989, respectively, compared to first‐order and Grau's second‐order models with coefficients of K₁ 1.08 per day, R² 0.739, and K_2s 1.69 per day, a = 7.0 per day, b = 0.847.     

Cr(VI) Immobilization Process in a Cr‐Spiked Soil by Zerovalent Iron Nanoparticles: Optimization Using Response Surface Methodology

The response surface methodology involving the five‐level central composite design (CCD) was employed to model and optimize the Cr(VI) immobilization process in a Cr‐spiked soil using starch‐stabilized zerovalent iron nanoparticles (ZVIn). ZVIn were synthesized via a borohydride reduction method and characterized by X‐ray diffraction (XRD) and scanning electron microscopy (SEM). All Cr(VI) immobilization experiments were conducted in a batch system. The variables for the CCD optimization were the ZVIn dosage (% w/w), reaction time (min), and initial Cr(VI) concentration in soil (mg/kg). The predicted response values by the second‐order polynomial model were found to be in good agreement with experimental values (R² = 0.968 and adj‐R² = 0.940). The optimization result showed that the Cr(VI) immobilization efficiency presented the maximal result (90.63%) at the following optimal conditions: ZVIn dosage of 1.5% w/w, reaction time of 60 min, and an initial Cr(VI) concentration of 400 mg/kg.     

Uncertainty analysis of CO<Subscript>2</Subscript> flux components in subtropical evergreen coniferous plantation

We present an uncertainty analysis of ecological process parameters and CO₂ flux components (R _eco, NEE and gross ecosystem exchange (GEE)) derived from 3 years’ continuous eddy covariance measurements of CO₂ fluxes at subtropical evergreen coniferous plantation, Qianyanzhou of ChinaFlux. Daily-differencing approach was used to analyze the random error of CO₂ fluxes measurements and bootstrapping method was used to quantify the uncertainties of three CO₂ flux components. In addition, we evaluated different models and optimization methods in influencing estimation of key parameters and CO₂ flux components. The results show that: (1) Random flux error more closely follows a double-exponential (Laplace), rather than a normal (Gaussian) distribution. (2) Different optimization methods result in different estimates of model parameters. Uncertainties of parameters estimated by the maximum likelihood estimation (MLE) are lower than those derived from ordinary least square method (OLS). (3) The differences between simulated R_eco, NEE and GEE derived from MLE and those derived from OLS are 12.18% (176 g C·m⁻²·a⁻¹), 34.33% (79 g C·m⁻²·a⁻¹) and 5.4% (92 g C·m⁻²·a⁻¹). However, for a given parameter optimization method, a temperature-dependent model (T_model) and the models derived from a temperature and water-dependent model (TW_model) are 1.31% (17.8 g C·m⁻²·a⁻¹), 2.1% (5.7 g C·m⁻²·a⁻¹), and 0.26% (4.3 g C·m⁻²·a⁻¹), respectively, which suggested that the optimization methods are more important than the ecological models in influencing uncertainty in estimated carbon fluxes. (4) The relative uncertainty of CO₂ flux derived from OLS is higher than that from MLE, and the uncertainty is related to timescale, that is, the larger the timescale, the smaller the uncertainty. The relative uncertainties of R_eco, NEE and GEE are 4%−8%, 7%−22% and 2%−4% respectively at annual timescale. Supported by the National Natural Science Foundation of China (Grant No. 30570347), Innovative Research International Partnership Project of the Chinese Academy of Sciences (Grant No. CXTD-Z2005-1) and National Basic Research Program of China (Grant No. 2002CB412502)     

Comparison of AnnAGNPS and SWAT model simulation results in USDA‐CEAP agricultural watersheds in south‐central Kansas

This study was conducted under the USDA‐Conservation Effects Assessment Project (CEAP) in the Cheney Lake watershed in south‐central Kansas. The Cheney Lake watershed has been identified as ‘impaired waters’ under Section 303(d) of the Federal Clean Water Act for sediments and total phosphorus. The USDA‐CEAP seeks to quantify environmental benefits of conservation programmes on water quality by monitoring and modelling. Two of the most widely used USDA watershed‐scale models are Annualized AGricultural Non‐Point Source (AnnAGNPS) and Soil and Water Assessment Tool (SWAT). The objectives of this study were to compare hydrology, sediment, and total phosphorus simulation results from AnnAGNPS and SWAT in separate calibration and validation watersheds. Models were calibrated in Red Rock Creek watershed and validated in Goose Creek watershed, both sub‐watersheds of the Cheney Lake watershed. Forty‐five months (January 1997 to September 2000) of monthly measured flow and water quality data were used to evaluate the two models. Both models generally provided from fair to very good correlation and model efficiency for simulating surface runoff and sediment yield during calibration and validation (correlation coefficient; R², from 0·50 to 0·89, Nash Sutcliffe efficiency index, E, from 0·47 to 0·73, root mean square error, RMSE, from 0·25 to 0·45 m³ s^?1 for flow, from 158 to 312 Mg for sediment yield). Total phosphorus predictions from calibration and validation of SWAT indicated good correlation and model efficiency (R² from 0·60 to 0·70, E from 0·63 to 0·68) while total phosphorus predictions from validation of AnnAGNPS were from unsatisfactory to very good (R² from 0·60 to 0·77, E from ? 2·38 to 0·32). The root mean square error–observations standard deviation ratio (RSR) was estimated as excellent (from 0·08 to 0·25) for the all model simulated parameters during the calibration and validation study. The percentage bias (PBIAS) of the model simulated parameters varied from unsatisfactory to excellent (from 128 to 3). This study determined SWAT to be the most appropriate model for this watershed based on calibration and validation results. Copyright © 2008 John Wiley & Sons, Ltd.     

Daily pan evaporation modelling using multi‐layer perceptrons and radial basis neural networks

This paper reports on investigations of the abilities of three different artificial neural network (ANN) techniques, multi‐layer perceptrons (MLP), radial basis neural networks (RBNN) and generalized regression neural networks (GRNN) to estimate daily pan evaporation. Different MLP models comprising various combinations of daily climatic variables, that is, air temperature, solar radiation, wind speed, pressure and humidity were developed to evaluate the effect of each of these variables on pan evaporation. The MLP estimates are compared with those of the RBNN and GRNN techniques. The Stephens‐Stewart (SS) method is also considered for the comparison. The performances of the models are evaluated using root mean square errors (RMSE), mean absolute error (MAE) and determination coefficient (R²) statistics. Based on the comparisons, it was found that the MLP and RBNN computing techniques could be employed successfully to model the evaporation process using the available climatic data. The GRNN was found to perform better than the SS method. Copyright © 2008 John Wiley & Sons, Ltd.     

Numerically modelling groundwater in an arid area with ANN‐generated dynamic boundary conditions

Groundwater is sensitive to the climate change and agricultural activities in arid and semi‐arid areas. Over the past several decades, human activities, such as groundwater extraction for irrigation, have resulted in aquifer overdraft and disrupted the natural equilibrium in these areas. Regional groundwater simulation is important to determine appropriate groundwater management policies, and numerical simulation has become the most popular method. However, most groundwater models were developed with static boundary conditions. In this research, the Minqin oasis, an arid region located in northwest China, was selected as the study area. An artificial neural network (ANN) was developed to simulate effects of weather conditions, agricultural activities and surface water on groundwater level in a dynamic boundary of the domain. Subsequently, a groundwater numerical model, named ANN‐FEFLOW model, was developed, with a dynamic boundary condition defined by the ANN model. The verifying results showed that the model has higher precision, with a root mean square error (RMSE) of 0·71 m, relative error (RE) of 17·96% and R² of 0·84 relative to the great groundwater change. Furthermore, the groundwater model has higher precision than the conventional groundwater model with static boundary condition, particularly in the area near the dynamic boundary. This study demonstrated that dynamic boundaries can improve the precision of the regional groundwater model in an arid area and that ANN can provide higher accuracy prediction capability for groundwater levels with dynamic boundary. Copyright © 2010 John Wiley & Sons, Ltd.     

Ground truthed performance of single‐and dual‐polarized radar rain rates at large ranges

Radar accuracy in estimating qualitative precipitation estimation at distances larger than 120 km degrades rapidly because of increased volume coverage and beam height. The performance of the recently upgraded dual‐polarized technology to the NEXRAD network and its capabilities are in need of further examination, as improved rainfall estimates at large distances would allow for significant hydrological modelling improvements. Parameter based methods were applied to radars from St. Louis (KLSX) and Kansas City (KEAX), Missouri, USA, to test the precision and accuracy of both dual‐ and single‐polarized parameter estimations of precipitation at large distances. Hourly aggregated precipitation data from terrestrial‐based tipping buckets provided ground‐truthed reference data. For all KLSX data tested, an R(Z,ZDR) algorithm provided the smallest absolute error (3.7 mm h^?1) and root‐mean‐square‐error (45%) values. For most KEAX data, R(ZDR,KDP) and R(KDP) algorithms performed best, with RMSE values of 37%. With approximately 100 h of precipitation data between April and October of 2014, nearly 800 and 400 mm of precipitation were estimated by radar precipitation algorithms but was not observed by terrestrial‐based precipitation gauges for KLSX and KEAX, respectively. Additionally, nearly 30 and 190 mm of measured precipitation observed by gauges were not detected by the radar rainfall estimates from KLSX and KEAX, respectively. Results improve understanding of radar based precipitation estimates from long ranges thereby advancing applications for hydrometeorological modelling and flood forecasting. Copyright © 2016 John Wiley & Sons, Ltd.     

Nested‐Scale Nutrient Flux in a Mixed‐Land‐Use Urbanizing Watershed

To improve quantitative understanding of mixed‐land‐use impacts on nutrient yields, a nested‐scale experimental watershed study design (n = 5) was applied in a 303(d), clean water act impaired urbanizing watershed of the lower Missouri River Basin, USA. From 2010 to 2013, water samples (n = 858 sample days per site) were analysed for total inorganic nitrogen (TIN‐N), nitrite (NO₂–N) nitrate (NO₃–N), ammonia (NH₃–N), and total phosphorus (TP‐P). Annual, seasonal, and monthly flow‐weighted concentrations (FWCs) and nutrient yields were estimated. Mean nutrient concentrations were highest where agricultural land use comprised 58% of the drainage area (NH₃ = 0.111 mg/l; NO₂ = 0.045 mg/l; NO₃ = 0.684 mg/l, TIN = 0.840 mg/l; TP = 0.127 mg/l). Average TP‐P increased by 15% with 20% increased urban land use area. Highly variable annual precipitation was observed during the study with highest nutrient yields during 2010 (record setting wet year) and lowest nutrient yields during 2012 (extreme drought year). Annual TIN‐N and TP‐P yields exceeded 10.3 and 2.04 kg ha^?1 yr^?1 from the agricultural dominated headwaters. Mean annual NH₃–N, NO₂–N, NO₃–N, TIN‐N, and TP‐P yields were 0.742, 0.400, 4.24, 5.38, and 0.979 kg ha^?1 yr^?1, respectively near the watershed outlet. Precipitation accounted for the majority of the explained variance in nutrient yields (R² values from 0.68 to 0.85). Nutrient yields were also dependent on annual precipitation of the preceding year (R² values from 0.87 to 0.91) thus enforcing the great complexity of variable mixed‐land‐use mediated source‐sink nutrient yield relationships. Study results better inform land managers and best management practices designed to mitigate nutrient pollution issues in mixed‐land‐use freshwater ecosystems. Copyright © 2015 John Wiley & Sons, Ltd.     

Estimation of monthly reference evapotranspiration using novel hybrid machine learning approaches

ABSTRACT

In this research, five hybrid novel machine learning approaches, artificial neural network (ANN)-embedded grey wolf optimizer (ANN-GWO), multi-verse optimizer (ANN-MVO), particle swarm optimizer (ANN-PSO), whale optimization algorithm (ANN-WOA) and ant lion optimizer (ANN-ALO), were applied for modelling monthly reference evapotranspiration (ET_o) at Ranichauri (India) and Dar El Beida (Algeria) stations. The estimates yielded by hybrid machine learning models were compared against three models, Valiantzas-1, 2 and 3 based on root mean square error (RMSE), Nash-Sutcliffe efficiency (NSE), Pearson correlation coefficient (PCC) and Willmott index (WI). The results of comparison show that the ANN-GWO-1 model with five input variables (T_min, T_max, RH, U_s, R_s) provides better estimates at both study stations (RMSE = 0.0592/0.0808, NSE = 0.9972/0.9956, PCC = 0.9986/0.9978, and WI = 0.9993/0.9989). Also, the adopted modelling strategy can build a truthful expert intelligent system for estimating the monthly ET_o at study stations.