Improved monthly streamflow prediction using integrated multivariate adaptive regression spline with K-means clustering: implementation of reanalyzed remote sensing data

Stochastic Environmental Research and Risk Assessment, Journal Year: 2024, Volume and Issue: 38(6), P. 2489 - 2519

Published: March 16, 2024

Abstract This study investigates monthly streamflow modeling at Kale and Durucasu stations in the Black Sea Region of Turkey using remote sensing data. The analysis incorporates key meteorological variables, including air temperature, relative humidity, soil wetness, wind speed, precipitation. also accuracy multivariate adaptive regression (MARS) with Kmeans clustering (MARS-Kmeans) by comparing it single MARS, M5 model tree (M5Tree), random forest (RF), multilayer perceptron neural network (MLP). In first stage, principal component is applied to diverse input combinations, both without lagged (Q), resulting twenty-three twenty respectively. Results demonstrate critical role Q for improved accuracy, as models exhibit significant performance degradation. second stage involves a comparative MARS-Kmeans other machine-learning models, utilizing best-input combination. MARS-Kmeans, incorporating three clusters, consistently outperforms showcasing superior predicting streamflow.

Language: Английский

---

Simulation and Reconstruction of Runoff in the High-Cold Mountains Area Based on Multiple Machine Learning Models

Water, Journal Year: 2023, Volume and Issue: 15(18), P. 3222 - 3222

Published: Sept. 10, 2023

Runoff from the high-cold mountains area (HCMA) is most important water resource in arid zone, and its accurate forecasting key to scientific management of resources downstream basin. Constrained by scarcity meteorological hydrological stations HCMA inconsistency observed time series, simulation reconstruction mountain runoff have always been a focus cold region research. Based on observations Yurungkash Kalakash Rivers, upstream tributaries Hotan River northern slope Kunlun Mountains at different periods, atmospheric circulation indices, we used feature analysis machine learning methods select input elements, train, simulate, preferences models runoffs two watersheds, reconstruct missing series River. The results show following. (1) Air temperature driver variability mountainous areas River, had strongest performance terms Pearson correlation coefficient (ρXY) random forest importance (FI) (ρXY = 0.63, FI 0.723), followed soil 0.043), precipitation, hours sunshine, wind speed, relative humidity, were weakly correlated. A total 12 elements selected as data. (2) Comparing simulated eight methods, found that gradient boosting performed best, AdaBoost Bagging with Nash–Sutcliffe efficiency coefficients (NSE) 0.84, 0.82, 0.78, while support vector regression (NSE 0.68), ridge 0.53), K-nearest neighbor 0.56), linear 0.51) poorly. (3) application four boosting, forest, AdaBoost, bagging, simulate for 1978–1998 was generally outstanding, NSE exceeding 0.75, reconstructing data period (1999–2019) could well reflect characteristics intra-annual inter-annual changes runoff.

Language: Английский

---

Review: Hydroinformatics: A Review and Future Outlook — R0/PR3

Published: Aug. 4, 2023

Hydroinformatics is a technology that combines information and communications technologies together with various disciplinary optimization simulation models focus on the management of water. This paper reviews historical development hydroinformatics summarizes current state this technology. It describes range modeling tools applications currently described in literature. The concludes some speculations about possible future developments hydroinformatics.

Language: Английский

---

Hydroinformatics: A review and future outlook

Cambridge Prisms Water, Journal Year: 2023, Volume and Issue: 1

Published: Jan. 1, 2023

Abstract Hydroinformatics is a technology that combines information and communications technologies together with various disciplinary optimization simulation models focus on the management of water. This paper reviews historical development hydroinformatics summarizes current state this technology. It describes range modeling tools applications currently described in literature. The concludes some speculations about possible future developments hydroinformatics.

Language: Английский

---

Research on modification and noise reduction optimization of Electric Multiple Units traction gear under multiple working conditions

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(2), P. e0298785 - e0298785

Published: Feb. 14, 2024

The vibration and radiation noise characteristics of the gear transmission system are different under traction conditions, modification optimization scheme based on a single working condition is not suitable for operating environment all conditions. To modify high-speed EMU, an optimized design reduction multiple conditions proposed. A plan tooth direction in conjunction with shape was devised using parametric model EMU’s system. after solved acoustic boundary element method prediction random forest proposed, parameter combination constructed to minimize noise. Then, optimal multi-condition parameters obtained weight running time contribution grey correlation degree evaluation established verify that can make EMU obtain satisfactory performance effect

Language: Английский

---

Review of flood monitoring and prevention approaches: a data analytic perspective

Natural Hazards, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 2, 2024

Language: Английский

---

Application of a Multi-Model Fusion Forecasting Approach in Runoff Prediction: A Case Study of the Yangtze River Source Region

Sustainability, Journal Year: 2024, Volume and Issue: 16(14), P. 5964 - 5964

Published: July 12, 2024

Runoff forecasting is crucial for sustainable water resource management. Despite the widespread application of deep learning methods in this field, there still a need improvement modeling and utilization multi-scale information. For first time, we introduce Neural Basis Expansion Analysis with Exogenous Variable (NBEATSx) model to perform runoff prediction full exploration rich temporal characteristics sequences. To harness wavelet transform (WT) information capabilities, developed WT-NBEATSx model, integrating WT NBEATSx. This was further enhanced by incorporating Long Short-Term Memory (LSTM) superior long-term dependency detection Random Forest (RF) as meta-model. The result advanced multi-model fusion WT-NBEATSx-LSTM-RF (WNLR). approach significantly enhances performance prediction. Utilizing daily scale meteorological dataset from Yangtze River Source region China 2006 2018, systematically evaluated WNLR tasks. Compared LSTM, Gated Recurrent Units (GRUs), NBEATSx models, not only outperforms original but also surpasses other comparison particularly accurately extracting cyclical change patterns, NSE scores 0.986, 0.974, 0.973 5-, 10-, 15-day forecasts, respectively. Additionally, compared standalone LSTM GRU introduction transforms form WT-LSTM WT-GRU notably improved robustness, especially where increased 32% 1.5%, study preliminarily proves effectiveness combining creatively proposes new RF, providing insights considering features complex time series, thereby enhancing effectiveness.

Language: Английский

---

Remote sensing of climate variability and flooding

Elsevier eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 183 - 204

Published: Jan. 1, 2024

Language: Английский

---

Modeling an evaluation framework for adding IoT water-level sensors based on ANN-derived 2D inundation simulations

Journal of Hydroinformatics, Journal Year: 2024, Volume and Issue: 26(9), P. 2261 - 2288

Published: Aug. 20, 2024

ABSTRACT This study aims to develop a smart model for evaluating the spatial density of added IoT sensors (called AIOT grids) optimize their amount and placements, named SM_ESD_AIOT model; proposed mainly collaborates cluster analysis with Akaike information criterion (AIC) based on resulting 2D inundation simulations from ANN-derived in comparison those physically hydrodynamic (SOBEK) under various sets AIOT-based sensor networks. Miaoli City northern Taiwan is selected as three practical sensors; also, 1,939 electrical poles are treated potential grids grouped 5, 10, 15, 20 clusters. Using simulated rainfall-induced flood event 51 h, five sets, consisting sensors, could be optimal one minimum AIC (around 1.45). Also, average, simulation indices networks 0.7 better than results (about 0.495). As result, shown efficiently placements enhance reliability accuracy simulation.

Language: Английский

---

Dynamic flood risk prediction in Houston: a multi-model machine learning approach

Geocarto International, Journal Year: 2024, Volume and Issue: 39(1)

Published: Jan. 1, 2024

Language: Английский

---