Kling-Gupta Efficiency Research Articles

Groundwater level modeling using Augmented Artificial Ecosystem Optimization

Nature-inspired optimization is an active area of research in the artificial intelligence (AI) field and has recently been adopted in hydrology for the calibration (training) of both process-based and statistical models. This study proposes an improved AI model, Augmented Artificial Ecosystem Optimization-based Multi-Layer Perceptron (AAEO-MLP), to build a monthly groundwater level (GWL) forecasting model. AAEO-MLP model is built on the novel Augmented version of Artificial Ecosystem Optimization and traditional MLP network. In AAEO, Levy-flight trajectory and Gaussian random are utilized in exploration and exploitation to improve the optimizing ability. The AAEO-MLP model is tested on two time-series (1989–2012) datasets collected at two wells in India. Various explanatory variables such as monthly cumulative precipitation, mean temperature, tidal height, and previous measurements of GWL were considered for predicting 1-month ahead of GWL. The performance of AAEO-MLP was benchmarked against 17 different models (original AEO, 3 different variants of AEO, and 13 well-known models) in terms of forecasting accuracy based on six metrics (e.g., mean absolute error, root mean square error, Kling–Gupta efficiency, normalized Nash–Sutcliffe efficiency, Pearson’s correlation index, a20 index). Furthermore, convergence analysis and model stability are employed to indicate the effectiveness of AAEO-MLP. The compared results express that the AAEO-MLP is superior to other models in terms of prediction accuracy, convergence, and stability. Overall, the results depict that the AAEO is a promising approach for optimizing machine learning models (e.g., MLP) and should be explored for other hydrological forecasting applications (e.g., streamflow, rainfall) to further assess its strengths over existing methods.

Weather Impact on Solar Farm Performance: A Comparative Analysis of Machine Learning Techniques

Forecasting the performance and energy yield of photovoltaic (PV) farms is crucial for establishing the economic sustainability of a newly installed system. The present study aims to develop a prediction model to forecast an installed PV system’s annual power generation yield and performance ratio (PR) using three environmental input parameters: solar irradiance, wind speed, and ambient air temperature. Three data-based artificial intelligence (AI) techniques, namely, adaptive neuro-fuzzy inference system (ANFIS), response surface methodology (RSM), and artificial neural network (ANN), were employed. The models were developed using three years of data from an operational 2MWp Solar PV Project at Kuzhalmannam, Kerala state, India. Statistical indices such as Pearson’s R, coefficient of determination (R2), root-mean-squared error (RMSE), Nash-Sutcliffe efficiency (NSCE), mean absolute-percentage error (MAPE), Kling-Gupta efficiency (KGE), Taylor’s diagram, and correlation matrix were used to determine the most accurate prediction model. The results demonstrate that ANFIS was the most precise performance ratio prediction model, with an R2 value of 0.9830 and an RMSE of 0.6. It is envisaged that the forecast model would be a valuable tool for policymakers, solar energy researchers, and solar farm developers.

Long-term reconstruction of satellite-based precipitation, soil moisture, and snow water equivalent in China

Abstract. A long-term high-resolution national dataset of precipitation (P), soil moisture (SM), and snow water equivalent (SWE) is necessary for predicting floods and droughts and assessing the impacts of climate change on streamflow in China. Current long-term daily or sub-daily datasets of P, SM, and SWE are limited by a coarse spatial resolution or the lack of local correction. Although SM and SWE data derived from hydrological simulations at a national scale have fine spatial resolutions and take advantage of local forcing data, hydrological models are not directly calibrated with SM and SWE data. In this study, we produced a daily 0.1∘ dataset of P, SM, and SWE in 1981–2017 across China, using global background data and local on-site data as forcing input and satellite-based data as reconstruction benchmarks. Global 0.1∘ and local 0.25∘P data in 1981–2017 are merged to reconstruct the historical P of the 0.1∘ China Merged Precipitation Analysis (CMPA) available in 2008–2017 using a stacking machine learning model. The reconstructed P data are used to drive the HBV hydrological model to simulate SM and SWE data in 1981–2017. The SM simulation is calibrated by Soil Moisture Active Passive Level 4 (SMAP-L4) data. The SWE simulation is calibrated by the national satellite-based snow depth dataset in China (Che and Dai, 2015) and the Moderate Resolution Imaging Spectroradiometer (MODIS) snow cover data. Cross-validated by the spatial and temporal splitting of the CMPA data, the median Kling–Gupta efficiency (KGE) of the reconstructed P is 0.68 for all grids at a daily scale. The median KGE of SM in calibration is 0.61 for all grids at a daily scale. For grids in two snow-rich regions, the median KGEs of SWE in calibration are 0.55 and −2.41 in the Songhua and Liaohe basins and the northwest continental basin respectively at a daily scale. Generally, the reconstruction dataset performs better in southern and eastern China than in northern and western China for P and SM and performs better in northeast China than in other regions for SWE. As the first long-term 0.1∘ daily dataset of P, SM, and SWE that combines information from local observations and satellite-based data benchmarks, this reconstruction product is valuable for future national investigations of hydrological processes.

WaterBench-Iowa: a large-scale benchmark dataset for data-driven streamflow forecasting

Abstract. This study proposes a comprehensive benchmark dataset for streamflow forecasting, WaterBench-Iowa, that follows FAIR (findability, accessibility, interoperability, and reuse) data principles and is prepared with a focus on convenience for utilizing in data-driven and machine learning studies, and provides benchmark performance for state of art deep learning architectures on the dataset for comparative analysis. By aggregating the datasets of streamflow, precipitation, watershed area, slope, soil types, and evapotranspiration from federal agencies and state organizations (i.e., NASA, NOAA, USGS, and Iowa Flood Center), we provided the WaterBench-Iowa for hourly streamflow forecast studies. This dataset has a high temporal and spatial resolution with rich metadata and relational information, which can be used for a variety of deep learning and machine learning research. We defined a sample benchmark task of predicting the hourly streamflow for the next 5 d for future comparative studies, and provided benchmark results on this task with sample linear regression and deep learning models, including long short-term memory (LSTM), gated recurrent units (GRU), and sequence-to-sequence (S2S). Our benchmark model results show a median Nash-Sutcliffe efficiency (NSE) of 0.74 and a median Kling-Gupta efficiency (KGE) of 0.79 among 125 watersheds for the 120 h ahead streamflow prediction task. WaterBench-Iowa makes up for the lack of unified benchmarks in earth science research and can be accessed at Zenodo https://doi.org/10.5281/zenodo.7087806 (Demir et al., 2022a).

How well do the multi-satellite and atmospheric reanalysis products perform in hydrological modelling

Multi-satellite, gauge-based and atmospheric reanalysis precipitation and temperature datasets with increasing spatiotemporal resolution are proliferating in recent decades. Their estimation accuracy has been evaluated in several areas, but their applications in global hydrological monitoring and modelling have been poorly understood so far. Here, we focus on the performance of the recently released Multi-Source Weighted-Ensemble Precipitation, version 2 (MSWEP V2) and the atmospheric reanalysis ERA5 temperature in modelling daily river discharge over 10,596 global catchments. Four lumped models including XAJ, GR4J, HMETS, and HBV are employed to take different model conceptual schemes into consideration. Moreover, different streamflow conditions including the peak timing and volumes of both high and low flows are evaluated to reveal further dynamics. The results show that: (a) the MSWEP V2 and ERA5 combination yields satisfactory performance in global streamflow simulation with at least one candidate model exhibiting Kling-Gupta efficiency (KGE) > 0.5 at both calibration and validation periods for each catchment; (b) better simulation skills are achieved in temperate, cold, polar, and tropical Köppen-Geiger climate zones, whereas the arid zone has poorer performance; (c) the XAJ, GR4J, and HMETS models with different degree of realisms have their own merits in simulating both daily mean flow and wet flow conditions but perform less satisfactorily in low flow conditions, while the HBV exhibits consistently poor performance in various flow conditions over most catchments. Our findings highlight the importance of selecting appropriate hydrological models by considering both climatic and geophysical conditions and targeting streamflow metrics, and suggest the potential of using multi-satellite and atmospheric reanalysis datasets as observed alternatives for quasi-global hydrological modelling.

Hydrological modeling in a region with sparsely observed data in the eastern Central Andes of Peru, Amazon

The Amazon basin is presented in eight South American countries. It is a shelter of great biodiversity from its source in the Andes to its mouth in the Atlantic Ocean. Hydrological research in its headwater region in the Peruvian Andes is affected by the low availability of observed data. In this context, evaluating the applicability of remote sensing hydrological data is an essential alternative for providing background for water resource management. The objectives of this study were i) to evaluate the quality of the Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG) satellite precipitation, ii) to perform hydrological modeling of the Chicllarazo River Basin (CRB) using the Soil and Water Assessment Tool (SWAT) hydrological model based in three precipitation data (observed (OP), satellite (IMERG) and satellite with bias correction (IMERG*)), and iii) to study the hydrological processes in the CRB's sub-basins. The evaluation results of the IMERG dataset based on observed precipitation data showed an underestimation of 25.1%. Comparing the IMERG and IMERG* datasets, correlation coefficients (r) of 0.79 and 0.80, root means square errors (RMSEs) of 55.4 and 49.3 mm, and Kling-Gupta efficiency coefficients (KGEs) of 0.66 and 0.76 were observed, respectively. These results indicate that bias correction in the IMERG data before performing hydrological modeling at the CRB is relevant. In the hydrological modeling, better results were obtained with IMERG*, according to NSE, Pbias, and KGE. The evaluation of the hydrological modeling results in sub-basins was found to be feasible for the study of the hydrological behavior in scarcely monitored regions, like Andes Cordillera.

Daily precipitation performances of regression-based statistical downscaling models in a basin with mountain and semi-arid climates.

The online version contains supplementary material available at 10.1007/s00477-022-02345-5.

Tropospheric water vapor: a comprehensive high-resolution data collection for the transnational Upper Rhine Graben region

Abstract. Tropospheric water vapor is one of the most important trace gases of the Earth's climate system, and its temporal and spatial distribution is critical for the genesis of clouds and precipitation. Due to the pronounced dynamics of the atmosphere and the nonlinear relation of air temperature and saturated vapor pressure, it is highly variable, which hampers the development of high-resolution and three-dimensional maps of regional extent. With their complementary high temporal and spatial resolutions, Global Navigation Satellite Systems (GNSS) meteorology and Interferometric Synthetic Aperture Radar (InSAR) satellite remote sensing represent a significant alternative to generally sparsely distributed radio sounding observations. In addition, data fusion with collocation and tomographical methods enables the construction of detailed maps in either two or three dimensions. Finally, by assimilation of these observation-derived datasets with dynamical regional atmospheric models, tropospheric water vapor fields can be determined with high spatial and continuous temporal resolution. In the following, a collection of basic and processed datasets, obtained with the above-listed methods, is presented that describes the state and course of atmospheric water vapor for the extent of the GNSS Upper Rhine Graben Network (GURN) region. The dataset contains hourly 2D fields of integrated water vapor (IWV) and 3D fields of water vapor density (WVD) for four multi-week, variable season periods between April 2016 and October 2018 at a spatial resolution of (2.1 km)2. Zenith total delay (ZTD) from GNSS and collocation and refractivities are provided as intermediate products. InSAR (Sentinel-1A/B)-derived double differential slant total delay phases (ddSTDPs) and GNSS-based ZTDs are available for March 2015 to July 2019. The validation of data assimilation with five independent GNSS stations for IWV shows improving Kling–Gupta efficiency (KGE) scores for all seasons, most notably for summer, with collocation data assimilation (KGE = 0.92) versus the open-cycle simulation (KGE = 0.69). The full dataset can be obtained from https://doi.org/10.1594/PANGAEA.936447 (Fersch et al., 2021).

Predictability of daily streamflow for the great rivers of South America based on a simple metric

ABSTRACT In this work, the role of initial conditions (ICs) and boundary conditions (BCs) in the predictability of natural flows of large river basins (>1000 km2) in South America using the South American Large Basins Model (MGB-SA) was investigated. We proposed an analysis based on a simple metric: predictability was estimated by the number of days in which the streamflow forecasts performed above a given threshold. The role of ICs was assessed through null precipitation experiments and the role of BCs through climatological rainfall. For experiments with a Kling-Gupta efficiency metric and threshold of 0.95, the study shows that streamflow predictability from the ICs is relatively high (up to 40 days) in the main river reaches of flat-relief basins. The increase in predictability due to climatology-based BCs mostly occurs in areas with high predictability based on ICs. Basins with fast response present low streamflow predictability (up to three days).

Forecasting High-Flow Discharges in a Flashy Catchment Using Multiple Precipitation Estimates as Predictors in Machine Learning Models

The use of machine learning (ML) for predicting high river flow events is gaining prominence and among its non-trivial design decisions is the definition of the quantitative precipitation estimate (QPE) product included in the input dataset. This study proposes and evaluates the use of multiple concurrent QPEs to improve the performance of a ML model towards the forecasting of the discharge in a flashy urban catchment. Multiple extreme learning machine (ELM) models were trained with distinct combinations of QPEs from radar, reanalysis, and gauge datasets. Their performance was then assessed in terms of goodness of fit and contingency analysis for the prediction of high flows. It was found that multi-QPEs models overperformed the best of its single-QPE counterparts, with gains in Kling-Gupta efficiency (KGE) values up to 4.76% and increase of precision in detecting high flows up to 7.27% for the lead times in which forecasts were considered “useful”. The novelty of these results suggests that the implementation of ML models could achieve better performance if the predictive features related to rainfall data were more diverse in terms of data sources when compared with the currently predominant use of a single QPE product.

Enhancing the Australian Gridded Climate Dataset rainfall analysis using satellite data

Rainfall estimation over large areas is important for a thorough understanding of water availability, influencing societal decision-making, as well as being an input for scientific models. Traditionally, Australia utilizes a gauge-based analysis for rainfall estimation, but its performance can be severely limited over regions with low gauge density such as central parts of the continent. At the Australian Bureau of Meteorology, the current operational monthly rainfall component of the Australian Gridded Climate Dataset (AGCD) makes use of statistical interpolation (SI), also known as optimal interpolation (OI) to form an analysis from a background field of station climatology. In this study, satellite observations of rainfall were used as the background field instead of station climatology to produce improved monthly rainfall analyses. The performance of these monthly datasets was evaluated over the Australian domain from 2001 to 2020. Evaluated over the entire national domain, the satellite-based SI datasets had similar to slightly better performance than the station climatology-based SI datasets with some individual months being more realistically represented by the satellite-SI datasets. However, over gauge-sparse regions, there was a clear increase in performance. For a representative sub-domain, the Kling-Gupta Efficiency (KGE) value increased by + 8% (+ 12%) during the dry (wet) season. This study is an important step in enhancing operational rainfall analysis over Australia.

Accuracy of methods for simulating daily water surface evaporation evaluated by the eddy covariance measurement at boreal flux sites

Evaporation from the surface of water is a vital component of the hydrological cycle. Accurately simulating evaporation is crucial for modeling hydrological processes, ensuring ecosystem water efficiency, and managing water resources. Previous methods for evaluating evaporation have generally been limited to a single site and on a regional scale. Based on eddy covariance measurements of 14 boreal flux sites, the accuracy of 26 methods for simulating daily water surface evaporation was comprehensively evaluated in this study. Most methods accurately simulated daily evaporation at most sites with a KGE >0.60. The combination methods with simultaneous calibration of the energy and aerodynamic terms and the double-parameter aerodynamic methods based on the linear and exponential functions showed the best performance, with the median KGE of the evaluated sites ranging from 0.72 to 0.76. The Bowen ratio energy balance (BREB) method embedded with Tw, Rs-based two-variable empirical methods, double-parameter aerodynamic methods, and the combination method of the energy term based on net radiation (Rn) and the aerodynamic term based on the exponential function method performed better than the other methods of the same type. The relative error (RE) simulated by most methods was generally within ± 30 %, with the median RE of all sites within ± 10 % for each method. The combination methods tended to overestimate the level of evaporation, whereas the BREB-type and aerodynamic methods tended to underestimate the extent of evaporation. The simulation accuracy of the daily evaporation showed significant variance among the sites. It showed the best model performance at two FLUXNET and two sites taken from the literature, in which 75 % of the methods were able to accurately simulate daily evaporation with a Kling–Gupta efficiency (KGE) >0.80. All methods performed poorly at the three sites from the literature and were mainly due to the lack of measured variables with a strong correlation with the latent heat flux (LE). The FLUXNET sites generally showed better performance. The observed LE of the sites with the best model performance generally showed a unimodal distribution throughout the year. The simulation accuracy of non-unimodal distribution sites mainly depended on the correlation between the LE and related variables. The combination of the energy term based on Rn and the aerodynamic term based on the power function method showed the best model stability. This finding indicated that the calibrated method was robust and showed high stability in simulating daily evaporation, and it was recommended for application in other sites or regions. The optimization of parameters calibrated in this study may improve the accuracy of simulating the daily evaporation for the application of the Penman and Priestley–Taylor equations when the water heat storage measurement is missing.

Spatio-temporal performance evaluation of 14 global precipitation estimation products across river basins in southwest Iran

Study regionKarkheh and Karun basins (29°−35°N, 46°−52°E) are two large river basins (area 51,000 and 67,000 km2, respectively) with complex topography in southwest Iran. Study focusAccess to spatio-temporally consistent precipitation data is a key requirement for hydrological studies, especially in data-scarce regions. This study evaluated 14 global precipitation products against gauge observations from 2003 to 2012 in Karun and Karkheh basins, southwest Iran. Different categorical and statistical indices at varying spatial and temporal resolution, including Kling-Gupta Efficiency (KGE), bias, correlation coefficient, and variability ratio, were used to evaluate the products. New hydrological insights for the regionFor daily time steps, TMPA-3B42V7.0, MERRA-2, and CMORPH-BLDV1.0 outperformed all other products, with KGE > 0.3. TMPA-3B42V7.0, MERRA-2, and PERSIANN-CDR were the best-performing products for monthly time steps, with KGE> 0.5. ERA5-Land showed the highest positive bias (bias>1.5) compared with in-situ observations, particularly for mountainous southeastern parts of Karun basin. Overall, bias-adjusted products obtained by merging ground-based observations in the estimations outperformed the unadjusted versions. The spatial distribution of statistical error metrics indicated that almost all products showed their greatest uncertainties for mountainous regions due to complex precipitation processes in these regions. These results can contribute significantly to hydrological and water resources planning measures in the study region, including early flood warning systems, drought monitoring, and optimization of dam operation.

Improving long-term streamflow prediction in a poorly gauged basin using geo-spatiotemporal mesoscale data and attention-based deep learning: A comparative study

Precise long-term streamflow prediction has always been important in the hydrology field, and has provided essential information for efficient water-resource management and disaster prevention. Attention to this field has increased recently owing to water and climate crises. Despite the remarkable improvements in existing data-driven models, they still have weaknesses, particularly for multistep predictions in poorly gauged basins. The purpose of this study is to improve the multistep ahead prediction ability by considering mesoscale hydroclimate data as booster predictors and employing attention-based deep learning. Therefore, a meta-algorithm is proposed for analyzing time-series data in complex geo-spatiotemporal environments to assess the superiority of prediction with booster predictors through direct and direct-recursive hybrid strategies. In the subsequent stage, a novel, integrated neural network architecture is demonstrated that couples a deep convolutional neural network (CNN) with a deep attention network. In particular, the former network performs automatic feature engineering, and the latter focuses on lengthy sequence complexities. Four state-of-the-art combinations for 12-month ahead prediction are introduced, including pairs of TimeDistributed-CNN (TD-CNN) and 3D-CNN along with a Long- and Short-term Time-series network (LSTNet) or a Transformer network. Moreover, a base architecture was employed for model comparison that contains the ConvLSTM2D layer that is compatible with multidimensional time-series data. For this, all models were applied to the Karkheh River basin in the northeast of the Persian Gulf, where monthly historical records of streamflow are available from 1955 to 2021. The results revealed that the application of hydroclimate sea surface temperature data and mean surface level pressure along with local data increased the prediction accuracy. Moreover, the proposed integrated networks delivered more accurate long-term streamflow predictions than the base models through various evaluation criteria, including r, R2, mean absolute error, root-mean-square error, Kling–Gupta efficiency, Willmott's Index, Legates–McCabe's, and the Akaike information criterion. In summary, the 3D-CNN–Transformer achieved the best performance, followed by the TD-CNN–Transformer, TD-CNN–LSTNet, and 3D-CNN–LSTNet with R2 values equal to 0.952, 0.930, 0.900, and 0.837, respectively. This study demonstrates that the application of hydroclimate data with proposed integrated networks are particularly useful for poorly gauged basin. Thus, the proposed models can potentially improve multistep ahead streamflow prediction compared to univariate and equation-based models.

Modeling monthly reference evapotranspiration process in Turkey: application of machine learning methods.

In this study, the predictive power of three different machine learning (ML)-based approaches, namely, multi-gene genetic programming (MGGP), M5 model trees (M5Tree), and K-nearest neighbor algorithm (KNN), for long-term monthly reference evapotranspiration (ET0) prediction were investigated. The input data consist of monthly solar radiation (Rs), maximum air temperature (Tmax), and wind speed (Ws) derived from 163 meteorological stations in Turkey. Different input combinations were created and analyzed. The model's performance was evaluated using criteria such as Nash-Sutcliffe efficiency, Kling-Gupta efficiency, relative root mean squared error, mean absolute percentage error, and determination coefficient. Moreover, Taylor, radar, and boxplot diagrams were created. It was determined that the MGGP model outperformed both the M5Tree and the KNN models. The equation obtained from the MGGP model, for the best-performed combination of Rs-Tmax-Ws, was presented. The best weather conditions were obtained as 0.029 to 31.814MJ/m2, - 5.8 to 45.7°C, and 0.140 to 5.086m/s for Rs, Tmax, and Ws, respectively. It was also found that the Rs was the most potent input variable for ET0 estimation while Ws was the weakest.

Long-Term Streamflow Prediction Using Hybrid SVR-ANN Based on Bayesian Model Averaging

The Volga River, as the primary supplier of the Caspian Sea, plays a huge role in its ecosystem sustainability. In this study, we analyze its runoff predictability for different monthly forecast horizons. Additionally, the meteorological and hydrological variables affecting runoff in each month are identified. A wide range of the potential variables was first collected and the Boruta variable preprocessing method was employed to select the important ones. Then the hybrid models were created by combining the selected variables and the data-driven models [i.e., support vector regression (SVR), artificial neural network (ANN), and multiple linear regression (MLR)]. To postprocess the predicted data, the Bayesian model averaging (BMA) method was employed using the combination of the Boruta–artificial neural network (B-ANN) and the Boruta–support vector regression (B-SVR) models. Finally, the Kling-Gupta efficiency (KGE) and the continuous ranked probability skill score (CRPSS) probabilistic evaluation criteria were applied to evaluate the hybrid models. The results showed that the streamflow of the previous steps is the most crucial variable in predicting the streamflow of all next horizons, while its significance decreases as the forecast time horizon increases. Moreover, the temperature variables have the unlike effect on the streamflow prediction and the minimum temperature for winter and spring and the maximum and the average temperatures for summer and autumn are the most effective ones. Correspondingly, BMA-B-ANN-SVR presents the best performance among the hybrid models (for example, the median of 0.8 and 0.91 for CRPSS and KGE in the first horizon, respectively); the reliability of its predicted runoff for different forecast horizons is much better than other hybrid models (B-ANN and B-SVR).

Advancing AI-based pan-European groundwater monitoring

The main challenge of pan-European groundwater (GW) monitoring is the sparsity of collated water table depth (wtd) observations. The wtd anomaly (wtda ) is a measure of the increased wtd due to droughts. Combining long short-term memory (LSTM) networks and transfer learning (TL), we propose an AI-based methodology LSTM-TL to produce reliable wtda estimates at the European scale in the absence of consistent wtd observational data sets. The core idea of LSTM-TL is to transfer the modeled relationship between wtda and input hydrometeorological forcings to the observation-based estimation, in order to provide reliable wtda estimates for regions with no or sparse wtd observations. With substantially reduced computational cost compared to physically-based numerical models, LSTM-TL obtained wtda estimates in good agreement with in-situ wtda measurements from 2569 European GW monitoring wells, showing r ⩾ 0.5, root-mean-square error ⩽1.0 and Kling-Gupta efficiency ⩾0.3 at about or more than half of the pixels. Based on the reconstructed long-term European monthly wtda data from the early 1980s to the near present, we provide the first estimate of seasonal wtda trends in different European regions, that is, significant drying trends in central and eastern Europe, which facilitates the understanding of historical GW dynamics in Europe. The success of LSTM-TL in estimating wtda also highlights the advantage of combining AI techniques with knowledge contained in physically-based numerical models in hydrological studies.

A guideline for spatio‐temporal consistency in water quality modelling in rural areas

A guideline for spatio‐temporal consistency in water quality modelling in rural areas

Karst spring recession and classification: efficient, automated methods for both fast- and slow-flow components

Abstract. Analysis of karst spring recession hydrographs is essential for determining hydraulic parameters, geometric characteristics, and transfer mechanisms that describe the dynamic nature of karst aquifer systems. The extraction and separation of different fast- and slow-flow components constituting a karst spring recession hydrograph typically involve manual and subjective procedures. This subjectivity introduces a bias that exists, while manual procedures can introduce errors into the derived parameters representing the system. To provide an alternative recession extraction procedure that is automated, fully objective, and easy to apply, we modified traditional streamflow extraction methods to identify components relevant for karst spring recession analysis. Mangin's karst-specific recession analysis model was fitted to individual extracted recession segments to determine matrix and conduit recession parameters. We introduced different parameter optimization approaches into Mangin's model to increase the degree of freedom, thereby allowing for more parameter interaction. The modified recession extraction and parameter optimization approaches were tested on three karst springs under different climate conditions. Our results showed that the modified extraction methods are capable of distinguishing different recession components and derived parameters that reasonably represent the analyzed karst systems. We recorded an average Kling–Gupta efficiency KGE > 0.85 among all recession events simulated by the recession parameters derived from all combinations of recession extraction methods and parameter optimization approaches. While there are variabilities among parameters estimated by different combinations of extraction methods, optimization approaches, and seasons, we found much higher variability among individual recession events. We provided suggestions to reduce the uncertainty among individual recession events and raised questions about how to improve confidence in the system's attributes derived from recession parameters.

Application and evaluation of SWAT-C model to predict TOC loading in the Hwangryong River Watershed

Objectives : This study aims to assess the applicability of the SWAT-C water quality model recently developed to predict in-stream Total Organic Carbon (TOC) in a watershed within South Korea.Methods : The SWAT-C model was tested in the Hwangryong River Watershed. SWAT-C is an advanced version of Soil and Water Assessment Tool (SWAT) to simulate carbon cycling at the watershed scale. The model was simulated for 11 years (2010-2020) consisting of 2-year warm-up (2010-2011), 6-year calibration (2012-2017), and 3-year validation (2018-2020) periods. SWAT-C was calibrated and validated against monthly streamflow and TOC loads. The model performance was evaluated using NSE (Nash-Sutcliffe efficiency), KGE (Kling-Gupta efficiency), and PBIAS (percent bias) Simulated Dissolve Organic Carbon (DOC) and Particulate Organic Carbon (POC) loads were also presented to see the major caron type loaded to the downstream.Results and Discussion : The calibration results showed that SWAT-C could depict streamflow and TOC loads with acceptance model performances (NSE > 0.5, KGE > 0.5 and PBIAS =< 25%). Especially, the seasonal variations of TOC loads were well simulated by SWAT-C. The TOC load was predicted to increase in the summer season with high precipitation and slightly lower in the other seasons. This seasonal pattern is likely caused by frequent heavy rainfall events in summer.Conclusion : This study demonstrated that SWAT-C model is applicable in South Korea with sufficient prediction accuracy. Thus, SWAT-C would serve as an efficient tool to monitor TOC loads in South Korea.