RMSE-observations Standard Deviation Ratio Research Articles

Prediction of carbon dioxide emissions from Atlantic Canadian potato fields using advanced hybridized machine learning algorithms – Nexus of field data and modelling

In this study, three novel machine learning algorithms of additive regression-random forest (AR-RF), Iterative Classifier Optimizer (ICO-AR-RF), and multi-scheme (MS-RF) were explored for carbon dioxide (CO2) flux rate prediction from three agricultural fields. To build the dataset, 401 samples were collected from two fields in Prince Edward Island (PEI) and 122 samples from the New Brunswick (NB), Canada. In addition, soil moisture (SM), temperature (ST), and electrical conductivity (EC), alongside eight climatic variables including wind speed (WS), solar radiation (SR), relative humidity (RH), precipitation (PCP), air temperature (AT), dew point (DP), vapour pressure difference (VPD) and reference evapotranspiration (ETo) were also collected. Greedy stepwise (GS) approach was implemented for feature selection. Finally, different qualitative (scatter plot, line graph, Taylor diagram, box plot, and Rug plot), and quantitative (uncertainty analysis, root mean square error (RMSE), percent of BIAS (PBIAS), Nash Sutcliff efficiency (NSE) and RMSE-observations standard deviation ratio (RSR)) techniques were used for model evaluation and comparison. Results of feature selection approaches revealed that DP, AT, SM, and ST are the four most effective variables at CO2 prediction in PEI, while AT, RH, DP, and ST are the most effective in the NB study area. For optimum input scenario, the GS algorithm was applied, and results showed that a combination of DP, AT, ST, SM, and ETo was the best for the PEI study area, while for NB, all input variables should be involved. Our analysis, for prediction of CO2 fluxes, confirmed that the ICO-AR-RF model performed the best at both PEI (RMSE=0.70, NSE=0.76, PBIAS=-5.11, RSR=0.48) and NB (RMSE=0.74, NSE=0.75, PBIAS=3.23, RSR=0.50), followed by MS-RF and AR-RF. Uncertainty analysis showed that CO2 prediction is more sensitive to input scenario selection than models in both study areas. Results revealed that climatic variables are more effective in CO2 prediction than soil characteristics and the developed hybrid model ICO-AR-RF can be a promising tool for decision-makers and beneficial for stakeholders.

A hybrid CNN–RNN model for rainfall–runoff modeling in the Potteruvagu watershed of India

AbstractAccurate rainfall‐runoff analysis is essential for water resource management, with artificial intelligence (AI) increasingly used in this and other hydrological areas. The need for precise modelling has driven substantial advancements in recent decades. This study employed six AI models. These were the support vector regression model (SVR), the multilinear regression model (MLR), the extreme gradient boosting model (XGBoost), the long‐short‐term memory (LSTM) model, the convolutional neural network (CNN) model, and the convolutional recurrent neural network (CNN‐RNN) hybrid model. It covered 1998–2006, with 1998–2004 for calibration/training and 2005–2006 for validation/testing. Five metrics were used to measure model performance: coefficient of determination (R2), Nash‐Sutcliffe efficiency (NSE), mean absolute error (MAE), root‐mean square error (RMSE), and RMSE‐observations standard deviation ratio (RSR). The hybrid CNN‐RNN model performed best in both training and testing periods (training: R2 is 0.92, NSE is 0.91, MAE is 10.37 m3s−1, RMSE is 13.13 m3s−1, and RSR is 0.30; testing: R2 is 0.95, NSE is 0.94, MAE is 12.18 m3s−1, RMSE is 15.86 m3s−1, and RSR is 0.25). These results suggest the hybrid CNN‐RNN model is highly effective for rainfall‐runoff analysis in the Potteruvagu watershed.

Statistical Performance of Gridded Rainfall Datasets Over Ungauged Jalaur River Basin, Philippines

The study presented aims to find the most appropriate climate dataset for the data-scarce Jalaur River Basin (JRB), Iloilo, Philippines, by evaluating the statistical performance of five rainfall datasets (APHRODITE, CPC NOAA, ERA5, SA-OBS, and PGF-V3) with resolutions of 0.25° and 0.5° having a time domain of 1981 to 2005. Bilinear interpolation implemented through Climate Data Operator (CDO) was used to extract and process grid climate datasets with Linear scaling as bias correction to minimize product simulation uncertainties. The datasets were compared to the lone meteorological station nearest to JRB investigated at monthly and annual timescales using six statistical metrics, namely, Pearson’s correlation coefficient (r), coefficient of determination (R2), modified index of agreement (d1), Kling-Gupta efficiency, Nash-Sutcliffe efficiency (NSE), and RMSE-observations standard deviation ratio (RSR). The results indicate a strong positive correlation with the observed data for both rainfall and temperature (r > 0.8; R2, d1 > 0.80). Although graphical observation shows an underestimation of rainfall, goodness-of-fit values indicate very good model performance (NSE, KGE > 0.75; RSR < 0.50). In terms of temperature, variable responses are observed with significant overestimation for maximum temperature and underestimation for minimum temperature. SA-OBS proved to be the best-performing dataset, followed by ERA5 and PGF-V3. These key findings supply useful information in deciding the most appropriate gridded climate dataset for hydrometeorological investigation in the JRB and could enhance the regional representation of global datasets.

Comparative analysis of HEC-HMS and SWAT hydrological models for simulating the streamflow in sub-humid tropical region in India.

Assessment of water availability in sub-humid regions is important due to distinct climatic and environmental conditions. In this study, Soil and Water Assessment Tool (SWAT) and Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS) models have been assessed in simulating streamflows in the sub-humid tropical Kabini basin in Kerala, India, spanning 1260 km2. Calibration and validation utilized daily weather data from 1997 to 2015 from the Muthankera gauging station. The study investigated the impact of routing methods on runoff simulation in the ArcSWAT, exploring Muskingum and Variable Storage methods. Evaluation metrics encompassed Nash-Sutcliffe Efïciency (NSE), Coefficient of Determination (R2), Percent bias (PBIAS), RMSE-observations standard deviation ratio (RSR), and Peak Percent Threshold Statistics (PPTS) approach for high-flow values. The result indicates that HEC-HMS outperforms SWAT concerning R2 and NSE values during daily calibration and validation. Monthly simulations showed HEC-HMS closely aligning with SWAT (Variable storage), outperforming SWAT (Muskingum). The PPTS approach proved effective in simulating high-flow values. Both models exhibited proficiency in streamflow analysis within the study area, promising predictive potential for future hydrological studies in sub-humid regions.

Vehicle Delay Model for Signalized Midblock Crosswalk Under Mixed Traffic Conditions

For signalized locations, delay is one of the key parameters in assessing the level of service and redesigning the traffic signals. Around the world, Highway Capacity Manual (HCM) and Webster delay models are commonly used in calculating the vehicle delay at signalized intersections, notwithstanding that these models assume homogeneous traffic conditions and lane discipline. Furthermore, it is worth noting that previous studies have mostly focused on developing delay models for signalized intersections, whereas signalized midblock crosswalks were overlooked by past researchers. In view of this, the current study presents a modified vehicle delay model for signalized midblock crosswalks (SMCs) under mixed traffic conditions. Mixed traffic conditions refers to diverse vehicles, complex interactions, and no lane discipline on the road—conditions which are usually noticed in developing nations. The HCM vehicle delay model is modified in the current study by incorporating factors related to mixed traffic conditions and SMCs. To test the accuracy of the proposed model, it was validated with the field delay using data collected at other SMCs in Hyderabad, India. Field delay was calculated by plotting the measured queue length against the signal cycle length. The performance metrics of root mean square error (RMSE), mean absolute percentage error, Nash–Sutcliffe efficiency, RMSE-observations standard deviation ratio, and percent bias were used to test the accuracy of the proposed model. Results revealed that the proposed model yielded the lowest error rates and the highest correlation coefficients between modeled and actual vehicle delay. The maximum error between the field delay and the modeled delay was found to be less than 5%.

Comprehensive propagation characteristics between paired meteorological and hydrological drought events: Insights from various underlying surfaces

Propagation from meteorological to hydrological drought events is crucial for understanding drought variation mechanisms and their consequences. Most previous studies were implemented using various standardized drought indices at month scale, which were difficult to reveal the physical processes and insufficient to characterize the comprehensive propagation characteristics (e.g., propagation time, duration, and severity) from meteorological to hydrological drought events. Furthermore, the impact mechanisms of underlying surfaces on drought propagation remained unclear. This study identified the paired relationships between meteorological and hydrological drought events and investigated their comprehensive propagation characteristics at the hydrologic response unit scale using the Soil and Water Assessment Tool (SWAT), run theory and paired analysis. The impact of underlying surfaces (e.g., land use, soil type, and slope) on drought propagation was further assessed using the Student's t-test at the 95% confident level. The upper middle Yuanjiang River Basin in China, which is prone to drought and exhibits diverse underlying surface conditions, was selected as the study area. Results indicate that the SWAT model optimally simulates the monthly streamflow with the Nash-Sutcliffe efficiency being ≥0.75 and the RMSE-observations standard deviation ratio being ≤0.50. Four main propagation types are identified, i.e., one-to-one, many-to-one, one-to-many, and many-to-many, in which the first two types are the dominant ones, accounting for 86.7% and 13.3% of total paired events, respectively. More than 85% of the study area exhibits a propagation time of ≤2 months, especially in the headwater area and the right bank of Lvzhijiang River. Drought duration and severity are amplified during the propagation process in 60.8% and 60.3% of the study area, respectively. Drought propagation time, duration, and severity are mainly controlled by soil and slope, and are slightly affected by land use type. The propagation time is quite short in areas with a high soil infiltration rate and steep slope, and the drought duration and severity are amplified in areas with a low soil infiltration rate and gentle slope. This study aims to provide deep insights on propagation from meteorological to hydrological drought considering various underlying surfaces, and further provides a scientific basis for the drought propagation mechanism and for drought disaster prevention and adaptation.

Climate change and its effects on the streamflow of an Andean river basin with volcanic activity

Abstract We use a Generalized Watershed Loading Function (GWLF) model to simulate streamflow in the Gualí River Basin. The model's performance is assessed using metrics such as percentage of bias (PBIAS), Nash–Sutcliffe efficiency (NSE), RMSE–observations standard deviation ratio (RSR), and Kling–Gupta efficiency (KGE) to indicate good performance. Furthermore, we analyze projections of precipitation and streamflow using several global climate models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6) and three shared socioeconomic pathways (SSP1-2.6, SSP2-4.5, SSP5-8.5). Despite the uncertainties and coarse resolution, our results show that increases in the mean streamflow and significantly decreasing trends in projected precipitation and streamflow are observed from 2015 to 2099 under the SSP5-8.5 scenario. Furthermore, our findings suggest an increase in long-term mid-flow and low-flow. Moreover, this work provides a methodological framework for hydrological modeling in small tropical river basins, by incorporating data from GCMs while raising concerns and caveats. This study offers valuable insights into the potential effects of climate change on streamflow in an Andean river basin characterized by volcanic activity and significant human impacts. The findings reported here provide useful information for future decisions related to water supply for the social, environmental, and productive sectors in the seven towns within the catchment.

Predicting the compressive strength of High-performance concrete by using Radial basis function with optimization Improved Grey Wolf optimizer and Dragonfly algorithm

This paper uses two optimizers (Improved Gray Wolf Optimizer (I_GWO) and Dragonfly Optimization Algorithm (DA)) for the sensitivity and robustness of artificial intelligence (AI) techniques, namely radial basis functions (RBFs). The purpose is to evaluate and analyze the predictive strength of high-performance concrete (HPC). 170 samples were collected for this purpose. This includes eight input parameters, cement, silica fume, fly ash, water, coarse aggregate, total aggregate, high water reducing agent, concrete age, and one output parameter, the compressive strength, to produce Increase learning and validation data sets. The proposed AI model was validated against several standard criteria: coefficient of determination (R2), root mean square error (RMSE), scatter index (SI), RMSE-observations standard deviation ratio (RSR), and coefficient of persistence (CP), n10_index. Many runs were performed to analyze the sensitivity and robustness of the model. The results show that I_GWO using RBF performs better than DA. Furthermore, sensitivity analysis indicated that cement content and HPC test age are the most essential and sensitive factors for predicting the compressive strength of HPC, according to the evaluations performed on the models, it was seen that the IGWO_RBF model provided better results compared to other models and can be introduced as the practical model for the prediction of HPC’s CS. In conclusion, this study can help to select appropriate AI models and suitable input parameters to accurately and quickly estimate the compressive strength of HPC.

Hydrological responses to land use and land cover change and climate dynamics in the Rift Valley Lakes Basin, Ethiopia

Abstract Due to the rapid socio-economic development in the Ethiopian Rift Valley basin, the pressures on water resources are increasing. To understand the change of spatio-temporal water fluxes, the hydrologic model SWAT+ (Soil and Water Assessment Tool+) was applied to five selected watersheds within the basin. With regards to the objective functions, Kling–Gupta efficiency (KGE: 0.68–0.84), the Nash–Sutcliffe efficiency (NSE: 0.61–0.73), percent bias (PBIAS: −3.4 to 1.4), and RMSE-observations standard deviation ratio (RSR: 0.52–0.69), the SWAT+ model performed very well for daily streamflow in all watersheds. The change in water balance components indicated a considerable spatial variation of water fluxes in the watersheds. Precipitation, evapotranspiration, and infiltration have generally decreased, but surface runoff has increased in the interference period compared to the baseline period. The spatial distribution of rainfall (−40 to 10%), evapotranspiration (−20 to 5%), surface runoff (7.8–13.1%), lateral flow (4.47 to −16.5%), and percolation (−3.3 to −10.2%) varied. The changes in the hydrologic system within the basin are greatly attributed to the combination of land use and land cover change due to rapid population growth and climate variability.

Deep Learning-Based Univariate Prediction of Daily Rainfall: Application to a Flood-Prone, Data-Deficient Country

There are several attempts to model rainfall time series which have been explored by members of the hydrological research communities. Rainfall, being one of the defining factors for a flooding event, is rarely modeled singularly in deep learning, as it is usually performed in multivariate analysis. This study will attempt to explore a time series modeling method in four subcatchments located in Samar, Philippines. In this study, the rainfall time series was treated as a signal and was reconstructed into a combination of a ‘smoothened’ or ‘denoised’ signal, and a ‘detailed’ or noise signal. The discrete wavelet transform (DWT) method was used as a reconstruction technique, in combination with the univariate long short-term memory (LSTM) network method. The combination of the two methods showed consistently high values of performance indicators, such as Nash–Sutcliffe efficiency (NSE), correlation coefficient (CC), Kling–Gupta efficiency (KGE), index of agreement (IA), and Legates–McCabe index (LMI), with mean average percentage error (MAPE) values at almost zero, and consistently low values for both residual mean square error (RMSE) and RMSE-observations standard deviation ratio (RSR). The authors believe that the proposed method can give efficient, time-bound results to flood-prone countries such as the Philippines, where hydrological data are deficient.

Comparison of performance of SWAT and SIMHYD models in simulation of stream flow from Hidkal dam catchment area of India under present and future scenarios

A study was conducted to compare the performance of SWAT and SIMHYD models in prediction of stream flow from the catchment of Hidkal dam of Karnataka in India for present and future scenarios. SWAT and SIMHYD models were used for the present study and calibrated with SWAT-CUP and generic algorithm respectively. The performance of models were tested with Nash Sutcliffe Efficiency (NSE), RMSE observations standard deviation ratio (RSR), Coefficient of Determination (R2) and Per cent Bias (PBIAS). The SWAT model performed well and very good agreement between monthly observed and simulated stream flows with R2, NSE, RSR and PBIAS as 0.84, 0.76, 0.49 and 23.8% respectively during calibration periods and 0.93, 0.83, 0.41 and 22.9 % respectively during validation period. SimHYD model also performed well and the NSE and correlation coefficient was found to be 0.77 and 0.93 during calibration period and 0.90 and 0.95 during validation period respectively. This study revealed that both models performed well in simulation of monthly stream flow during calibration and validation periods. The stream flows simulated with SimHYD were compared with stream flows simulated with SWAT model and found that very good agreement between two model outputs. This study concluded that SimHYD model could be used to simulate the stream flow when data limitation.

Estimation of Runoff and Sediment Yield in Response to Temporal Land Cover Change in Kentucky, USA

Land cover change is prevalent in the eastern Kentucky Appalachian region, mainly due to increased surface mining activities. This study explored the potential change in land cover and its relationship with stream discharge and sediment yield in a watershed of the Cumberland River near Harlan, Kentucky, between 2001 and 2016, using the Soil and Water Assessment Tool (SWAT). Two land cover scenarios for the years 2001 and 2016 were used separately to simulate the surface runoff and sediment yield at the outlet of the Cumberland River near Harlan. Land cover datasets from the National Land Cover Database (NLCD) were used to reclassify the land cover type into the following classes: water, developed, forest, barren, shrubland, and pasture/grassland. Evaluation of the relationship between the land cover change on discharge and sediment was performed by comparing the average annual basin values of streamflow and sediment from each of the land cover scenarios. The SWAT model output was evaluated based on several statistical parameters, including the Nash–Sutcliffe efficiency coefficient (NSE), RMSE-observations standard deviation ratio (RSR), percent bias (PBIAS), and the coefficient of determination (R²). Moreover, P-factor and R-factor indices were used to measure prediction uncertainty. The model showed an acceptable range of agreement for both calibration and validation between observed and simulated values. The temporal land cover change showed a decrease in forest area by 2.42% and an increase in developed, barren, shrubland, and grassland by 0.11%, 0.34%, 0.53%, and 1.44%, respectively. The discharge increased from 92.34 mm/year to 104.7 mm/year, and sediment increased from 0.83 t/ha to 1.63 t/ha from 2001 to 2016, respectively. Based on results from the model, this study concluded that the conversion of forest land into other land types could contribute to increased surface runoff and sediment transport detached from the soil along with runoff water. The research provides a robust approach to evaluating the effect of temporal land cover change on Appalachian streams and rivers. Such information can be useful for designing land management practices to conserve water and control soil erosion in the Appalachian region of eastern Kentucky.

Evaluating Monthly Flow Prediction Based on SWAT and Support Vector Regression Coupled with Discrete Wavelet Transform

Reliable and accurate streamflow prediction plays a critical role in watershed water resources planning and management. We developed a new hybrid SWAT-WSVR model based on 12 hydrological sites in the Illinois River watershed (IRW), U.S., that integrated the Soil and Water Assessment Tool (SWAT) model with a Support Vector Regression (SVR) calibration method coupled with discrete wavelet transforms (DWT) to better support modeling watersheds with limited data availability. Wavelet components of the simulated streamflow from the SWAT-Calibration Uncertainty Procedure (SWAT-CUP) and precipitation time series were used as inputs to SVR to build a hybrid SWAT-WSVR. We examined the performance and potential of the SWAT-WSVR model and compared it with observations, SWAT-CUP, and SWAT-SVR using statistical metrics, Taylor diagrams, and hydrography. The results showed that the average of RMSE-observation’s standard deviation ratio (RSR), Nash–Sutcliffe efficiency (NSE), percent bias (PBIAS), and root mean square error (RMSE) from SWAT-WSVR is 0.02, 1.00, −0.15, and 0.27 m3 s−1 in calibration and 0.14, 0.98, −1.88, and 2.91 m3 s−1 in validation on 12 sites, respectively. Compared with the other two models, the proposed SWAT-WSVR model possessed lower discrepancy and higher accuracy. The rank of the overall performance of the three SWAT-based models during the whole study period was SWAT-WSVR > SWAT-SVR > SWAT-CUP. The developed SWAT-WSVR model supplies an additional calibration approach that can improve the accuracy of the SWAT streamflow simulation of watersheds with limited data.

An integrated framework for simultaneously modeling primary and secondary salinity at a watershed scale

Salinity is a major environmental phenomenon that affects watersheds and irrigated areas worldwide. The watershed-scale salinity modeling requires a watershed-scale model like the Soil and Water Assessment Tool (SWAT) that simulates hydrologic processes throughout a watershed. The Richards equation can assist in addressing very complex processes that affect water flow in the unsaturated root zone influenced by gravity and capillarity at a field scale. The SWAT model does not use the Richards equation for modeling flows in the root zone, resulting in less accuracy in simulating hydrologic processes related to agriculture. This study presents an integrated framework for the watershed-scale salinity modeling by modifying the water-salt balance (SWAT-S) model linked to the Soil Water Atmosphere Plant (SWAP) model. The integrated framework was applied to the Mond River Basin of Iran. The model was calibrated and validated with historical records of the river discharge and salt concentration using the SUFI-2 optimization algorithm to evaluate the model's strength and to verify the accuracy of simulations. The model performance was satisfactory in calibration and validation periods with the coefficient of determination (R2), Nash–Sutcliffe efficiency (NSE), Percent bias (|PBIAS|), and RMSE-observations standard deviation ratio (RSR) ranging from 0.56 to 0.87, 0.52 to 0.87, 0.2 to 22.2, and 0.37 to 0.69, respectively. Also, P‐factor values were more than 0.65, while R‐factor values were<1.40.

A Comparative Evaluation of Lumped and Semi-Distributed Conceptual Hydrological Models: Does Model Complexity Enhance Hydrograph Prediction?

The prediction of hydrological phenomena using simpler hydrological models requires less computing power and input data compared to the more complex models. Ordinarily, a more complex, white-box model would be expected to have better predictive capabilities than a simple grey box or black-box model. But complexity may not necessarily translate to better prediction accuracy or might be unfeasible in data scarce areas or when computer power is limited. Therefore, the shift of hydrological science towards the more process-based models needs to be justified. To answer this, the paper compares 2 hydrological models: (a) the simpler tank model; and (b) the more complex TOPMODEL. More precisely, the difference in performance between tank model as a lumped model and the TOPMODEL concept as a semi-distributed model in Atari River catchment, in Eastern Uganda was conducted. The objectives were: (1) To calibrate tank model and TOPMODEL; (2) To validate tank model and TOPMODEL; and (3) To compare the performance of tank model and TOPMODEL. During calibration, both models exhibited equifinality, with many parameter sets equally likely to make acceptable hydrological simulations. In calibration, the tank model and TOPMODEL performances were close in terms of ‘Nash-Sutcliffe efficiency’ and ‘RMSE-observations standard deviation ratio’ indices. However, during the validation period, TOPMODEL performed much better than tank model. Owing to TOPMODEL’s better performance during model validation, it was judged to be better suited for making runoff forecasts in Atari River catchment.

HEC-HMS based rainfall-runoff model for Punpun river basin

Abstract Various hydrological models were used in different river basins to simulate the runoff on available rainfall, land use and soil property data. The HEC-HMS model is used by several researchers to estimate the water potential of the basin through rainfall-runoff modeling. In this study, a rainfall-runoff model for the Punpun river basin has been developed using HEC-HMS. Daily rainfall and runoff data from the years 2005 to 2017 were used for the development of model. ArcGIS has been used to analyze the hydrological parameters, preparation of LULC, soil and slope maps for the computation of curve number as input into the HEC-HMS model. Daily, monthly and monsoonal rainfall-runoff models have been developed. The performance of all the models has been evaluated using statistical indices–coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), percent bias (PBIAS) and RMSE-observations standard deviation ratio (RSR). R2 and NSE values for all the models are greater than 0.75 and PBIAS is less than 10, which shows very good results from all the models except the daily model, in which NSE values are less than 0.75. Based on statistical indices, the monthly model performs better than the daily and monsoonal models.

Assessing the Effect of Land-Use and Land-Cover Changes on Discharge and Sediment Yield in a Rural Coal-Mine Dominated Watershed in Kentucky, USA

The Appalachian Mountain region of eastern Kentucky is unique and contains high proportions of forestland along with coal and natural gas depositaries. Landscape changes due to extreme mining activities can eventually threaten the downstream ecosystems, including soil and water quality, resulting in excessive runoff and sedimentation. The purpose of this study is to assess the impacts of land-use and land-cover (LULC) changes in streamflow and sediment yield in Yellow Creek Watershed, Kentucky, USA, between 1992 and 2016. LULC, digital elevation model, soil, and weather data were inputted into the Soil and Water Assessment Tool (SWAT) to simulate discharge and sediment yield. The model output was evaluated on several statistical parameters, such as the Nash-Sutcliffe efficiency coefficient (NSE), RMSE-observations standard deviation ratio (RSR), percent bias (PBIAS), and the coefficient of determination (R2). In addition, two indices, P-factor and R-factor, were used to measure the prediction uncertainty. The calibrated model showed an increase in surface runoff and sediment yield due to changes in LULC in the Yellow Creek Watershed. The results provided important insights for studying water management strategies to make more informed land management decisions and adaptive practices.

Forecast of rainfall distribution based on fixed sliding window long short-term memory

Applying data mining techniques for rainfall modeling because of a lack of sufficient memory components may increase uncertainty in rainfall forecasting. To solve this issue, in this research, a deep-learning-based long short-term memory (LSTM) model is developed for the first time for forecasting monthly rainfall data, and its capability is compared with a random forest (RF) data-driven model. To this end, monthly rainfall data for a period of 41 years (1980–2020) from two meteorological stations in Turkey, namely Rize and Konya, with different climatic conditions, are used. The analysis is carried out using optimum window sizes for determining the optimum lag times of rainfall time series. The performance of the models is evaluated using five statistical measures, namely root mean square error (RMSE), RMSE-observations standard deviation ratio (RSR), Legate and McCabe’s index (LMI), correlation coefficient (R) and Nash–Sutcliffe efficiency (NSE), and also using two visual means, namely Taylor and violin diagrams. The results reveal that the LSTM model, as a more efficient tool, outperforms the RF model in forecasting rainfall at both stations, with improved RMSE of 12.2–14.9%, RSR of 12.3–14.8%, R of 9.4–13.5% and NSE of 32.9–33.2%. The LSTM-based approach proposed herein could be adopted over any global climatic conditions to forecast the monthly rainfall with reasonable accuracy.

Insights into the Pollutant Removal Performance of Stormwater Green Infrastructures: A Case Study of Detention Basins and Retention Ponds.

The quality of water has deteriorated due to urbanization and the occurrence of urban stormwater runoff. To solve this problem, this study investigated the pollutant reduction effects from the geometric and hydrological factors of green infrastructures (GIs) to more accurately design GI models, and evaluated the factors that are required for such a design. Among several GIs, detention basins and retention ponds were evaluated. This study chose the inflow, outflow, total suspended solids (TSS), total phosphorus (TP), watershed area, GI area (bottom area in detention basins and permanent pool surface area in retention ponds), and GI volume (in both detention basins and retention ponds) for analysis and applied both ordinary least squares (OLS) regression and multiple linear regression (MLR). The geometric factors do not vary within each GI, but there may be a bias due to the number of stormwater events. To solve this problem, three methods that involved randomly extracting data with a certain range and excluding outliers were applied to the models. The accuracies of these OLS and MLR models were analyzed through the percentage bias (PBIAS), Nash-Sutcliffe efficiency (NSE), and RMSE-observations standard deviation ratio (RSR). The results of this study suggest that models which consider the influent concentration combined with the hydrological and GI geometric parameters have better correlations than models that consider only a single parameter.

Simulation of runoff in Baitarani basin using composite and distributed curve number approaches in HEC-HMS model

The present study was conducted in Baitarani basin up to Anandapur gauging station of Odisha covering an area of 8603.7 km2. Pre-processing of basin from digital elevation model (DEM) was done using HEC-Geo-HMS extension and spatial analyst tool in ArcGIS. These pre-processed files were then imported to HEC-HMS for simulating runoff. In this study, runoff simulation was done using two methods, viz., composite and distributed curve number (CN) approaches. SCS curve number method was used for computation of runoff volume, SCS UH method for direct runoff, constant- monthly varying base flow method for base flow and Muskingum method for flow routing. The model was calibrated and validated using both composite and distributed CN approaches. Data from 1st January, 2007 to 31st December, 2013 were used for calibration and 1st January, 2014 to 31st December, 2016 were used for validation. During the calibration period of composite CN approach, the statistical parameters like Nash-Sutcliffe efficiency (NSE), Coefficient of determination (R2), Percent bias (PBIAS) and RMSE-observations standard deviation ratio (RSR) were found to be 0.51, 0.63, 12.82 and 0.7, respectively and during the validation period they were found to be 0.53, 0.54, -19.73 and 0.7, respectively. In case of distributed CN approach, the statistical parameters like NSE, R2, PBIAS and RSR were found to be 0.62, 0.63, -8.64 and 0.6, respectively during the calibration period and 0.67, 0.66, -2.25 and 0.6, respectively during the validation period. The study indicated that distributed CN approach is more accurate than composite CN approach in simulation of runoff using HEC-HMS model.