A new interpretable streamflow prediction approach based on SWAT-BiLSTM and SHAP

Environmental Science and Pollution Research, Год журнала: 2024, Номер 31(16), С. 23896 - 23908

Опубликована: Март 2, 2024

Язык: Английский

---

Optimizing Spatial Distribution of Retail Shops against Neighborhood Tree Canopy Shade Using Big Data Extracted from Streetscape

Land, Год журнала: 2024, Номер 13(8), С. 1249 - 1249

Опубликована: Авг. 9, 2024

The visibility of retail frontages is critical for earning profits from spontaneous traffic visits to shops located along a street. urban tree canopy plays crucial role in enhancing the street-side environment, yet more not always better when considering placement behind trees with big canopies. Related evidence literature rarely provided, and an unclear relationship has been reported exist between number specific type quantified ratio shade street view. In this study, both data crawling deep learning were employed unravel Changchun, Northeast China. entire study area was divided into 6037 grid cells side length ~0.6 km, wherein five types (food beverage, shopping, life services, entertainment, hotel) by computer counting their points interest (POIs). evaluated using green view index (GVI) through pixels all image obtained online map API. A neighboring road network categorized four classes: class I density mainly reduced shops, densities classes III IV accounted shops. GVI could be fitted positive skewness curves II roads, where peak estimated about 3.27%. optimization scheme indicated that should placed roads. conclusion, food services landscape

Язык: Английский

---

Coupling SWAT+ with LSTM for enhanced and interpretable streamflow estimation in arid and semi-arid watersheds, a case study of the Tagus Headwaters River Basin, Spain

Environmental Modelling & Software, Год журнала: 2025, Номер unknown, С. 106360 - 106360

Опубликована: Фев. 1, 2025

Язык: Английский

---

Utilizing Deep Learning Models to Predict Streamflow

Water, Год журнала: 2025, Номер 17(5), С. 756 - 756

Опубликована: Март 5, 2025

This study employs convolutional neural network (CNN), long short-term memory (LSTM), bidirectional (BiLSTM), and gated recurrent unit (GRU) deep learning models to simulate daily streamflow using precipitation data. Two approaches were explored: one without dimension reduction another incorporating dimensionality technique. Principal component analysis (PCA) was employed for reduction, partial autocorrelation function (PACF) used determine time lags. An augmented Dickey–Fuller (ADF) test utilized ascertain the stationarity of data, ensuring optimal model performance. The data normalized then partitioned into features target variables, before being split training, validation, sets. developed tested their performance, robustness, stability at three locations along Neuse River, which is in River Basin, North Carolina, USA, covering an area about 14,500 km2. Furthermore, model’s performance during peak flood events assess ability capture temporal resolution streamflow. results revealed that CNN could variability prediction, as evidenced by excellent statistical measures, including mean absolute error, root square Nush–Sutcliffe efficiency. also found significantly improved

Язык: Английский

---

Exploring alternate coupling inputs of a data-driven model for optimum daily streamflow prediction in calibrated SWAT-BiLSTM rainfall-runoff modeling

Frontiers in Water, Год журнала: 2025, Номер 7

Опубликована: Апрель 2, 2025

Accurate streamflow prediction in mountainous regions is vital for sustaining water resources downstream areas, ensuring reliable availability agriculture, energy, and consumption. However, physically based models are prone to substantial uncertainties due complex processes the inherent variability model parameters parameterization. This study addresses these challenges by exploring alternative coupling inputs data-driven (DD) optimize daily a calibrated SWAT-BiLSTM rainfall-runoff within Astore sub-basin of Upper Indus Basin (UIB), Pakistan. The research explores two standalone (SWAT BiLSTM) three inputs: conventional climatic variables (precipitation temperature), cross-correlation selected inputs, exclusion direct model. spans calibration, validation, periods from 2007 2011, 2012 2015, 2017 2019, respectively. Based on compromise programing (CP) ranking, SWAT-C-BiLSTM (Q P ) (T 1 Q showed competent performances followed BiLSTM, (PTQ ), SWAT. These findings highlight that excluding enhances couple model’s accuracy sufficiently underscores potential this approach contribute sustainable resource management.

Язык: Английский

---

Stormwater characterisation and modelling for Sungai Air Hitam in Selangor, Malaysia using model for urban stormwater improvement conceptualisation (music)

Ecological Indicators, Год журнала: 2024, Номер 160, С. 111850 - 111850

Опубликована: Март 1, 2024

The aim of this study is to evaluate the current water quality status one urban rivers in Malaysia, called Sungai Air Hitam. river's supply not only unsuitable for inhabitants but also hazardous aquatic species that depend on it. In order simulate formulation river, Model Urban Stormwater Improvement Conceptualization (MUSIC) was used. effects various best management practices (BMPs) components have been examined improve quality. This investigated different scenarios expected future changes land cover and river. As proportion impervious surfaces increases, hydrology cycle can be significantly altered, resulting an increase volumes peak flows, a decrease storage, infiltration, interception. MUSIC results shown significant reductions biochemical oxygen demand (BOD), total suspended solids (TSS), phosphorus (TP), nitrogen (TN) after introducing BMPs. It noticed prediction pollutants falls within acceptable range set by Management Manual Malaysia (MSMA) 2nd edition. For cover, it found reduction BOD, TSS, TP, TN existing use 92.5 %, 94.5 90.7 % 91.9 %. Meanwhile, 81.6 86.2 80.9 TP 80.8 TN. From simulation results, observed application BMPs has successfully reduced mean BOD concentration from 92.38 mg/L (Class V) 6.93 IV) national standards, NWQS, index. result, index overall catchment improved Class IV III (WQ1, WQ3, WQ4) V (WQ2) with assessment aims raise awareness Hitam community regarding importance preserving river cleanliness understanding long-term environmental impact These findings underscore integrated system managing systems, which offer valuable insight decision-makers.

Язык: Английский

---

Farmer adoption-based prompt networking and modeling for targeting optimal agro-conservation practices

Environmental Modelling & Software, Год журнала: 2024, Номер 177, С. 106060 - 106060

Опубликована: Апрель 27, 2024

Язык: Английский

---

A coupled SWATPlus and BiLSTM tuning model for improved daily scale hydroclimate simulation in typical loess hilly areas of China

Natural Hazards, Год журнала: 2024, Номер unknown

Опубликована: Июль 25, 2024

Язык: Английский

---

Assessing Water Resource Vulnerability in an Agricultural Basin for Climate Change Adaptation

Water Resources Management, Год журнала: 2024, Номер unknown

Опубликована: Сен. 18, 2024

Язык: Английский

---