Association between forecasting models’ precision and nonlinear patterns of daily river flow time series

Rahmani, Farhang; Fattahi, Mohammad Hadi

doi:10.1007/s40808-022-01351-4

Cited by 6 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Forecasting models use previous data to make accurate predictions regardless of their structural types, such as nonlinear, linear, short, and extended memory. Previous studies' findings have shown that the results produced by the models such as Autoregressive Integrated Moving Average (ARIMA), Radial Basis Function (RBF) [10], Adaptive Network-based Fuzzy Inference System (ANFIS), and Artificial Neural Network (ANN) were sufficiently accurate and were suitable for forecasting hydrological time series [11][12][13][14][15][16]. However, in recent studies, gradient boosting-based regression algorithms such as extreme Gradient Boosting (XGBoost) and Light Gradient Boosting (Lightgbm) showed satisfactory results in forecasting problems.…”

Section: Plos Onementioning

confidence: 99%

Water level prediction using soft computing techniques: A case study in the Malwathu Oya, Sri Lanka

Rathnayake

Dang

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

Hydrologic models to simulate river flows are computationally costly. In addition to the precipitation and other meteorological time series, catchment characteristics, including soil data, land use, land cover, and roughness, are essential in most hydrologic models. The unavailability of these data series challenged the accuracy of simulations. However, recent advances in soft computing techniques offer better approaches and solutions at less computational complexity. These require a minimum amount of data, while they reach higher accuracies depending on the quality of data sets. The Gradient Boosting Algorithms and Adaptive Network-based Fuzzy Inference System (ANFIS) are two such systems that can be used in simulating river flows based on the catchment rainfall. In this paper, the computational capabilities of these two systems were tested in simulated river flows by developing the prediction models for Malwathu Oya in Sri Lanka. The simulated flows were then compared with the ground-measured river flows for accuracy. Correlation of coefficient (R), Per cent-Bias (bias), Nash Sutcliffe Model efficiency (NSE), Mean Absolute Relative Error (MARE), Kling-Gupta Efficiency (KGE), and Root mean square error (RMSE) were used as the comparative indices between Gradient Boosting Algorithms and Adaptive Network-based Fuzzy Inference Systems. Results of the study showcased that both systems can simulate river flows as a function of catchment rainfalls; however, the Cat gradient Boosting algorithm (CatBoost) has a computational edge over the Adaptive Network Based Fuzzy Inference System (ANFIS). The CatBoost algorithm outperformed other algorithms used in this study, with the best correlation score for the testing dataset having 0.9934. The extreme gradient boosting (XGBoost), Light gradient boosting (LightGBM), and Ensemble models scored 0.9283, 0.9253, and 0.9109, respectively. However, more applications should be investigated for sound conclusions.

show abstract

Section: Plos Onementioning

confidence: 99%

Water level prediction using soft computing techniques: A case study in the Malwathu Oya, Sri Lanka

Rathnayake

Dang

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…(2021); Legates and Outcalt (2022); Pal et al. (2020); Rahmani and Fattahi (2021, 2022c, 2022a, 2022b). The hypothesis of long‐term persistence (LTP) in annual precipitation has been explored in a number of studies of point and grid scale data.…”

Section: Introductionmentioning

confidence: 99%

“…The work of Hurst has been used to characterize LTP across multiple disciplines, ranging from climate science to the analysis of internet traffic and the flow of blood in human arteries (O'Connell et al, 2016). Recent research on Hurst behavior in the climate and hydrology fields is reported by Adarsh et al (2020); Adarsh and Priya (2021); Benavides-Bravo et al (2021); ; ; Legates and Outcalt (2022); Pal et al (2020); Rahmani and Fattahi (2021, 2022c, 2022a, 2022b. The hypothesis of long-term persistence (LTP) in annual precipitation has been explored in a number of studies of point and grid scale data.…”

mentioning

confidence: 99%

On the Spatial Scale Dependence of Long‐Term Persistence in Global Annual Precipitation Data and the Hurst Phenomenon

O’Connell

O’Donnell

Koutsoyiannis

2023

Water Resources Research

View full text Add to dashboard Cite

Long-term persistence (LTP) in geophysical time series, or the tendency of above or below average runs of years to be unusually long, was first quantified by Hurst (Hurst, 1951(Hurst, , 1956 using a coefficient H which characterizes LTP in the range 0.5 < H < 1, with H = 0.5 corresponding to the independent (white noise) case, that is, no persistence. Hurst analyzed a wide set of geophysical times series and found an average value of H = 0.73, with a standard deviation of 0.09. In particular, he reported a value of H = 0.9 for annual river Nile flows at Aswan, reflecting strong LTP. The disparity between these results, and the then current theory that predicted H = 0.5 based on the increments of Brownian motion, has come to be known as the Hurst Phenomenon. Over the years, a number of stochastic approaches to modeling LTP have emerged (e.g., fractional Gaussian noise (Mandelbrot & Wallis, 1968, 1969, ARMA models (O'Connell, 1974a(O'Connell, , 1974b, shifting mean models (Boes & Salas, 1978) and fractionally differenced models (Hosking, 1984)). Koutsoyiannis (2011a) has shown that those models exhibiting Hurst behavior asymptotically can be encapsulated within a Hurst -Kolmogorov (HK) stochastic dynamics framework characterized by a simple scaling law, acknowledging the contribution

show abstract

On the use of temporal evolution of persistence for change point detection of streamflow datasets

Sankaran,

Anilkumar,

Shajudeen

et al. 2024

Environ Earth Sci

View full text Add to dashboard Cite

Association between forecasting models’ precision and nonlinear patterns of daily river flow time series

Cited by 6 publications

References 39 publications

Water level prediction using soft computing techniques: A case study in the Malwathu Oya, Sri Lanka

Water level prediction using soft computing techniques: A case study in the Malwathu Oya, Sri Lanka

On the Spatial Scale Dependence of Long‐Term Persistence in Global Annual Precipitation Data and the Hurst Phenomenon

On the use of temporal evolution of persistence for change point detection of streamflow datasets

Contact Info

Product

Resources

About