Peak discharge is an essential element of hydrological forecasting. Due to rapid outbreaks of flash floods in hilly areas and the lack of measured data, the fast and accurate estimation of peak discharge is crucial for flash flood hazard management. Three machine learning algorithms were applied to estimate peak discharge; this estimation was compared with the results of hydrological–hydraulic models, and the results were verified with measured watershed data. In this paper, 10 hydrological and geomorphological parameters were selected to predict the flood peak discharge in 103 watersheds in Taiyi Mountain North District. The results show that the particle swarm optimization backpropagation (PSO-BP) neural network model outperforms the BP neural network and random forest regression in prediction performance. PSO-BP has a lower mean absolute error (2.51%), root mean square error (3.74%), and mean absolute percentage error (2.74%) than the other models, which indicates that PSO-BP has high prediction accuracy. Importance analysis revealed that rainfall, early impact rainfall, catchment area, and rain intensity are the key input parameters of PSO-BP. The proposed method was confirmed to be a fast and relatively accurate algorithm for estimating the peak discharge of flash floods in ungauged basins.

  • This paper applied three machine learning algorithms to estimate the peak discharge, comparing it with hydrological–hydraulic models and verifying the results by a watershed with measured data.

Floods can cause serious harm to the environment, economy, infrastructure, humans, and animals (Karami et al. 2024). The frequent occurrence of extreme rainfall scenarios due to climate change in recent years has placed greater demands on the accuracy and immediacy of hydrological forecasting and warning (Zsoter et al. 2020). However, floods in hilly areas are characterized by short ephemeral times and heavy rainfall, and many factors that affect flooding via complex processes. Thus, the quick and accurate prediction of the characteristic parameters of floods remains an urgent scientific problem.

Existing flood peak discharge prediction methods mainly include mechanism-based hydrological–hydraulic methods and data-driven machine learning-based prediction methods. Hydrological–hydraulic methods include methods such as building hydrological models and reasoning equation methods, antecedent precipitation index (API) (Yao et al. 2019) methods, and flood peak modulus calculations. Among the hydrological models, the Soil and Water Assessment Tool (SWAT) model (Karami et al. 2024; Yalcin 2024), the Nedbør-Afstrømnings-Model (NAM) model (Sun et al. 2020; Parvaze et al. 2022), the Variable Infiltration Capacity model (Meresa et al. 2022), and the Xinanjiang model (Gui et al. 2024; Liu et al. 2024) have more often been applied in hydrological forecasting and have achieved better results in practice. However, the premise of building hydrological models is to have a long series of measured flow and water level data for rate validation, which is more suitable for watersheds that have measured data and are large in size. In addition, methods such as the inference formula (Li et al. 2016), API (Ye et al. 2013), and flood peak modulus (Liu et al. 2022) are more widely used in the calculation of flood peaks in ungauged basins. These methods are mainly based on the probabilistic distribution modelling method (Meresa et al. 2022) and empirical formulas for flood calculation, which are characterized by low computational efficiency, mostly rely on manual experience, and neglect the mechanism of hydrological processes. In contrast, the machine learning method, as a data-driven model, overcomes the conditional limitations of the modelling process and can directly explore the intrinsic relationship of the data to achieve efficient prediction. Backpropagation (BP; Han et al. 2022), PSO-BP (Zhang et al. 2017), random forest regression (RFR) (Zhao et al. 2022), SVM (Rahimzad et al. 2021), long short-term memory (LSTM) (Rasheed et al. 2022), and other machine learning methods have been shown by many scholars to have excellent performance in hydrological forecasting and warning, but they are mostly applied to large watersheds in areas with large catchment areas, and LSTM has mostly been applied to time-series flood field prediction (Zhang et al. 2022). It is difficult to apply this method to ungauged basins. Thus, this paper selects the BP, PSO-BP, and RFR models as machine learning models.

Most of the current literature on flood flow prediction has focused on watersheds with large watershed areas where measured data are available (Tsakiri et al. 2018; Bernard & Gregoretti 2021), and less research has been carried out in watersheds with small natural watershed areas in ungauged basins. Hydrological–hydraulic modelling, a mechanism-based computational approach, suffers from the disadvantage of computational inefficiency, while machine learning methods, as a data-driven approach, require a large number of samples for the model to cover as large a distribution of variables as possible. Therefore, the coupling of hydrological–hydraulic methods with machine learning methods involves the coupling of model mechanisms with intelligent algorithmic tools, which improves not only the superb learning ability of machine learning to achieve fast and accurate predictions but also the physical interpretability and applicability of the model. In ungauged basins, the combination of hydrological–hydraulic methods and machine learning methods, such as the calculation of flood-forming flow (Han et al. 2022), the calculation of critical rainfall (Zhao et al. 2022), and the prediction of watershed runoff (Hughes et al. 2023), which have achieved better results. The combination of hydrological–hydraulic methods and machine learning methods has become a new research hotspot in flood prediction and early warning, but the prediction of the peak discharge of natural floods in ungauged basins is still relatively weak.

The paper uses a hydrological–hydraulic method combined with machine learning methods to predict natural flood peak discharge prediction in mountain flood watersheds. This method has been applied to 103 watersheds in the Taiyi Mountain North District. This is one of the best choices for natural flood parameterization of watersheds in a hilly area because of the distribution of 103 watersheds in the Taiyi Mountain North District, which is located in the highest topography of Shandong Province, has a small watershed area, lacks reservoirs in the watersheds, and is basically unaffected by human activities. First, the hydrological–hydraulic method was used to calculate and obtain the flood peak discharge results of 103 small watersheds, which served as the input data for the BP neural network, PSO-BP neural network, and RFR models for the prediction of flood peak discharge in ungauged basins. Second, the hydrological–hydraulic method and the machine learning method were validated in selected watersheds where measurement information was available. Last, input variable importance calculations were carried out to discuss in depth the contributions of parameters affecting the flood peak discharge in ungauged basins and to provide a new tool for exploring the natural flooding process in ungauged basins.

The manuscript is organized as follows: in Section 2, the study area and data source are described; in Section 3, the methodologies, BP neural network, PSO-BP neural network, RFR prediction model, and evaluation criteria of the prediction model are introduced; in Section 4 and Section 5, the results and discussion are reported; and in Section 6, the conclusions are drawn.

Study area

The Taiyi Mountain North District (Figure 1(a) is located in the central area of Shandong Province and includes Jinan City, Tai'an City, Zibo City, Weifang City, and Qingdao City. The elevation difference, total area, and ratio of hilly areas to plain areas are 1,342.82 m, 154,068.87 km2, and 0.55/0.45, respectively. The basin has a warm, temperate, semi-humid, continental monsoon climate. Rainfall is concentrated from early June to late September, with an average annual rainfall of 700 mm. It is simple to start a flood disaster when it arises in a mountain river since the flood is highly explosive and has a short duration.
Figure 1

An overview map of the study area. (a) Study area and (b) verification area.

Figure 1

An overview map of the study area. (a) Study area and (b) verification area.

Close modal

A total of 103 small watersheds in the Taiyi Mountain North District of Shandong Province were chosen as the study area for the following reasons: (1) there are no small reservoirs in the watershed, which avoids the influence of small reservoir storage on the flooding process and (2) no rainfall or hydrological stations belong to the watershed, and no measurements are available.

Dataset

Watershed parameter extraction

The ArcSWAT tool was used to extract information about 103 watersheds from 12.5 m Digital Elevation Model data, including variables such as watershed area, watershed slope, longest river confluence path, and river gradient. The DEM data were obtained from the ALOS (Advanced Land Observing Satellite, launched in 2006), which utilizes a phased array L-band synthetic aperture radar (PALSAR) (https://search.asf.alaska.edu/#/). The land use data and subsurface soil data are derived from the Resources and Environmental Science Data Center (http://www.resdc.cn/).

Rainfall data extraction

The 103 watersheds had no actual rainfall data. The rainfall in the watersheds was calculated using the spatial interpolation of actual rainfall measurements from stations near the watersheds. The watershed characteristic value adopts the extracted centroid point data of the watershed. The rainfall data were obtained from the Shandong Hydrology Bureau.

The range of the 10 meteorological and geomorphological parameters for the 103 watersheds is shown in Table S1 and Figure S3.

Machine learning methods

BP neural network

The BP neural network, which was proposed by Rumelhart and McClelland, is a multilayer feedforward neural network based on an error BP algorithm (Ervine et al. 2000). The BP neural network has become one of the most well-liked neural network models because of its strong nonlinear mapping and its capacity to accurately approximate any function. The BP neural network reduces the output error by understanding the training data using the gradient descent algorithm, and it trains the bias and node weights of the network via error backpropagation (Figure S1(a))).

PSO-BP neural network

Particle swarm optimization (PSO) is primarily utilized in the modified BP neural network algorithm model to optimize initial weights and thresholds, which are subsequently substituted into the BP neural network model for training and prediction (Hosseini et al. 2016). The PSO method is used to optimize the weights and thresholds, and the BP neural network model with optimized parameters is used to estimate the peak discharge of the small watershed by inputting parameters in hilly areas with ungauged basins (Figure S1(b)).

RFR model

As shown in Figure S2, the RFR model is a mixture of multiple regression decision trees with maximum growth connected to the independent distribution of random vectors , where x is an independent variable and T is the number of decision trees (Song 2015; Davis & Nielsen 2020; Lee & Kim 2020). Based on the idea of ensemble learning, the mean value of each regression decision subtree is used as the regression prediction result:
(1)
Where represents output based on x and .

Hydrological–hydraulic modelling to calculate flood peak discharge

The surface rainfall of the watershed was calculated using the point-rainfall conversion coefficient based on the spatially interpolated rainfall data, as well as the rainfall time distribution. The effective precipitation was then calculated using the rainfall-runoff relationship curve, and the flood discharge within the confluence time was calculated using the instantaneous unit line. In this study, the flood peak discharges of 15 different early impact rainfall events at five return periods (5, 10, 20, 50 and 100 years) were calculated for 103 small watersheds. The calculated 7,210 sets of data served as the actual values for the machine learning prediction model; 7,010 sets were used as the training set, and 200 sets were used as the test set.

Evaluation criteria

The mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE) (Yan et al. 2023), and R2 (Ayus et al. 2023) were applied to evaluate the performance of the model.

The absolute discrepancy between the predicted value and the actual value is measured as the MAE. The predicted value error is more accurately represented by the MAE. The smaller the MAE, the better the simulation prediction.
(2)
where represents the predicted value, and denotes the observed value.
The sample standard deviation between the expected value and the observed value is represented by the RMSE. The RMSE is used to indicate the degree of dispersion of the sample, and a smaller RMSE provides better results for a nonlinear fit.
(3)
The accuracy of various time-series model predictions is typically compared using the MAPE, a relative error measure that utilizes observed values to prevent positive and negative errors from cancelling one another out.
(4)
R2 is a measure of the estimated goodness of fit of the regression equation; the closer it is to 1, the better the fit of the model. R2 is the ratio of the sum of squares of the regression to the sum of squares of the total deviations in linear regression.
(5)

Results of machine learning models

Figure 2 shows the results of the BP neural network, the PSO-BP neural network, and the RFR. Figure 2(a) shows that the RMSE of the training data is the lowest and that the training impact is at its optimum when the number of hidden layers is set to 12. Figure 2(c) shows that the particle swarm stabilized at approximately 30 generations due to the large number of training datasets and that the optimization impact has reached the desired optimization value. As shown in Figure 2(e), the explained variance regression score stabilized after 50 decisions, indicating that the RFR model is optimized optimally.
Figure 2

Results of machine learning models. (a) The RMSE of different hidden layers. (b) The BP predicted value and the actual value. (c) PSO evolution convergence curve. (d) The improved PSO-BP predicted value and the actual value. (e) The number of decision tree and the explained variance regression score. (f) The RFR predicted value and the actual value.

Figure 2

Results of machine learning models. (a) The RMSE of different hidden layers. (b) The BP predicted value and the actual value. (c) PSO evolution convergence curve. (d) The improved PSO-BP predicted value and the actual value. (e) The number of decision tree and the explained variance regression score. (f) The RFR predicted value and the actual value.

Close modal

To verify the consistency and robustness of the BP neural network, the PSO-BP neural network, and the RFR model, Figure 2(b), 2(d), and 2(f) shows the scatter plots of the measured and simulated values for the 95% confidence scenario. The high overlap between the predicted trend line and the diagonal line (Figure 2(b)) proves that the predicted and actual values show agreement, indicating that the model has better prediction results. Figure 2(d) shows that the actual value of the test set and the anticipated value of the PSO-BP are dispersed around the diagonal line, proving that the prediction of the model is more accurate. The actual and predicted RFR values of the test set are dispersed along the diagonal line, as shown in Figure 2(f), demonstrating the broad applicability of the model. The correlation coefficients (r) of the BP neural network, the PSO-BP neural network, and RFR with the measured values are 0.99891, 0.99922, and 0.99114, respectively, again proving the high prediction accuracy of the three machine learning models, and the PSO-BP neural network has better robustness and consistency.

Comparative analysis of the results of different prediction models

A comparison between the predicted values and the actual values of the test set of three machine learning algorithms is provided in Figure 3(a), which provides a true picture of the trend and variability of the actual values. Figure 3(b) illustrates the error comparison between the test set's predicted value and the actual value for machine learning methods, with the deviations almost always falling between −20 and 20 m3/s.
Figure 3

Comparison of three machine learning prediction models. (a) The predicted value and the actual value. (b) The error between the predicted value and the actual value.

Figure 3

Comparison of three machine learning prediction models. (a) The predicted value and the actual value. (b) The error between the predicted value and the actual value.

Close modal
Figure 4 illustrates the evaluation metrics of the BP neural network, the PSO-BP neural network, and RFR machine learning models. The MAE is used to evaluate the actual prediction error. Figure 4(a) reveals that the PSO-BP neural network (2.506%) has a smaller prediction error, while Figure 4(c) shows that the PSO-BP neural network (2.737%) has a smaller average error. The RMSE amplifies and penalizes the larger error and is used to evaluate the degree of variability of the data. Figure 4(b) demonstrates that the PSO-BP neural network (3.742%) data have less variability. The r reflects the degree of correlation between the predicted value and the actual value, and the R2 reflects the metric for evaluating the goodness of fit. Figure 2(b), 2(d), and 2(e) confirm that the PSO-BP neural network has a better correlation, and Figure 4(d) confirms that the PSO-BP neural network achieves a better fit (0.99844). In summary, the PSO-BP neural network is more effective in reducing the prediction error and prediction model variability, and compared with BP and RFR, the MAE is reduced by 0.21 and 0.74, respectively, the RMSE is reduced by 0.16 and 0.63, respectively, the MAPE is reduced by 0.16 and 0.73, respectively, and the R2 is improved by 0.06 and 1.6%, respectively.
Figure 4

Machine learning algorithms predict flood peak evaluation parameters. (a) MAE, (b) RMSE, (c) MAPE, and (d) R2.

Figure 4

Machine learning algorithms predict flood peak evaluation parameters. (a) MAE, (b) RMSE, (c) MAPE, and (d) R2.

Close modal

Model verification

A watershed of a hilly area with measured peak discharge data was selected, and BP neural network, RFR, and PSO-BP neural network machine learning models and a hydrological–hydraulic model were used to calculate the peak discharge and compare it with the measured data to validate the model accuracy.

There are hydrological stations (Wohushan, Beifeng, Huangtaiqiao, and Gushan) within the selected watershed (Figure 1(b)). There should be no upstream reservoir as the selection criterion to guarantee the correctness of data verification, but this is challenging. Fortunately, a reservoir can be found in the upper reaches of the basin, which is located above the Beifeng Hydrological Station. We believe that the regulating and storage effects of the reservoir have little effect on the peak discharge at the station site when considering 100-year-return periods because of the peak-shaving and flood-regulating effects of the reservoir. Therefore, the once-in-100-year flood peak discharge of the Beifeng Hydrological Station was selected to verify the accuracy of the data.

Table 1 shows the comparison of the values of the measured, hydrological–hydraulic, BP neural network, RFR, and PSO-BP neural network models. The values calculated by the PSO-BP neural network model and the hydrological–hydraulic model are very similar, with a difference of 0.09%. Compared with those of the other two machine learning models, the predicted values of the PSO-BP neural network model are closer to the measured values, which once again confirm the excellent accuracy of the PSO-BP neural network model. The PSO-BP neural network model can be used for flood peak discharge prediction in hilly areas.

Table 1

Discharge data verification form

ModelValue (m3/s)Relative error (%)
Measured 1,723.00 – 
Hydrological–hydraulic model 1,766.36 2.52 
BP neural network model 1,798.81 4.40 
RFR model 1,788.99 3.83 
PSO-BP neural network model 1,767.97 2.61 
ModelValue (m3/s)Relative error (%)
Measured 1,723.00 – 
Hydrological–hydraulic model 1,766.36 2.52 
BP neural network model 1,798.81 4.40 
RFR model 1,788.99 3.83 
PSO-BP neural network model 1,767.97 2.61 

The performances of the BP, PSO-BP, and RFR models are further discussed. The BP neural network is adaptable to a large number of training samples, whereas the RFR model has superior application, primarily for small sample sets. The particle swarm algorithm has a significant advantage in parameter optimization, and based on the findings displayed in Figure 4, the PSO-BP has a prediction simulation accuracy that is superior to that of the BP neural network model (Zhang et al. 2017, 2023). The PSO-BP neural network algorithm addresses the issue that the gradient descent approach is prone to falling into local tiny values and significantly enhances the prediction performance of the entire model (Zhang et al. 2019). The advantage of the PSO-BP neural network algorithm over the RFR and BP neural network models is its excellent prediction accuracy. The experiments show that the PSO-BP neural network model is more suitable for predicting flood flow in hilly areas.

To further discuss the degree of influence of input factors on flood peak discharge, decision tree-based importance (Zhao et al. 2022) is introduced for visualization. The importance of rainfall, frequency, confluence time, early impact rainfall, catchment area, longest catchment path length, stream gradient, rain intensity, catchment slope, and shape factor was analysed, and rainfall, early impact rainfall, catchment area, and rain intensity accounted for 0.81, as shown in Figure 5(a). The results show that rainfall, early impact rainfall, catchment area, and rain intensity are more important. Based on the results of the importance analysis, the importance parameter was used as a model input, and the results are shown in Figure 5(b), with an r-value of 0.9963 and an R2-value of 0.99262, which further confirms the scientific validity of the importance analysis. The highest relevance of rainfall (Band et al. 2020; Rasheed et al. 2022) attests to the fact that rainfall reflects the amount of rainfall, which, when combined with earlier impacts of rainfall, defines the amount of runoff. The watershed area, on the other hand, ranked second in importance, which also proves that the watershed area has an important influence on the formation of flood peak discharge (Potdar et al. 2021). An early effect of precipitation indicates that the amount of runoff produced by rainfall is greatly influenced by the moisture and aridity of the soil in the basin (Han et al. 2022). The study showed that the accumulated rainfall in the early period will affect the runoff of the later rainfall. Due to the varying early-stage wetting conditions and the significance of early impact rainfall, when the rainfall intensity is the same, the abortion results will vary. The top three positions in Figure 5(a) illustrate the function that rainfall intensity plays in the prediction model. The rain intensity represents the amount of rainfall in a unit period. The rainfall duration is typically short, and the rainfall is large in hilly places, which is also a key signal in hydrological forecasting.
Figure 5

Characteristic parameter importance results and validation plot. (a) Feature parameter importance result graph and (b) scatterplot of the model for importance parameters.

Figure 5

Characteristic parameter importance results and validation plot. (a) Feature parameter importance result graph and (b) scatterplot of the model for importance parameters.

Close modal

Predictions in ungauged basins are considered another milestone that has significant implications for the advancement of hydrology. An accurate forecast of flood peak discharge is crucial in the flood forecasting process. Currently, the primary indicator for forecasting flash floods is a rainfall warning. Although they are also significant indicators, discharge and water level warnings are rarely used. Discharge forecasting plays an important role as an intermediary between critical rainfall and hazardous water levels. However, it is challenging to calibrate and test hydrological models in hilly watersheds because of a lack of rainfall and hydrological data.

This paper compares the results of hydrological–hydraulic methods and three machine learning models for computationally predicting flood peak discharges and uses a watershed for data validation. The relative errors ranged from 2.52 to 4.40%, confirming the accuracy and suitability of the hydrological–hydraulic and machine learning models. Compared with hydrological–hydraulic models, machine learning models have the advantages of high modelling efficiency. They can deeply mine the relationships among the data, and the results are reliable. This is a way to solve the problem of predicting flood peak discharge in hilly areas with ungauged basins. The values of 0.99922 for r and of 0.99844 for R2 confirm the superiority of the PSO-BP neural network over the BP neural network and RFR models. The importance analysis revealed that rainfall, early impact rainfall, watershed area, and rainfall intensity are more important in calculating flood peak discharge. The innovation of these results lies in the parallel application of hydrological–hydraulic models and machine learning models, which is a combination of traditional computation and new tools. These models can be applied to future flood forecasting in hilly areas with ungauged basins. This outcome can help hydrologists and local administrators provide important guidance for flash flood warning and scheduling policies and aid in smart water network construction.

The authors wish to gratefully acknowledge the financial assistance from the Natural Science Foundation of Shandong Province (ZR2020ME249), the Natural Science Foundation of Shandong Province (NSFS) (No. ZR2020QE282), the National Natural Science Foundation of China (42301046), and the other anonymous reviewer whose comments significantly improved the quality of this paper.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Ayus
I.
,
Natarajan
N.
&
Gupta
D.
2023
Prediction of water level using machine learning and deep learning techniques
.
Iranian Journal of Science and Technology, Transactions of Civil Engineering
47
(
4
),
2437
2447
.
https://doi.org/10.1007/s40996-023-01053-6
.
Band
S. S.
,
Janizadeh
S.
,
Chandra Pal
S.
,
Saha
A.
,
Chakrabortty
R.
,
Melesse
A. M.
&
Mosavi
A.
2020
Flash flood susceptibility modeling using new approaches of hybrid and ensemble tree-based machine learning algorithms
.
Remote Sensing
12
(
21
),
3568
.
Bernard
M.
&
Gregoretti
C.
2021
The use of rain gauge measurements and radar data for the model-based prediction of runoff-generated debris-flow occurrence in early warning systems
.
Water Resources Research
57
(
3
),
e2020WR027893
.
https://doi.org/10.1029/2020WR027893
.
Davis
R. A.
&
Nielsen
M. S.
2020
Modeling of time series using random forests: Theoretical developments
.
Electronic Journal of Statistics
14
(
2
),
3644
3671
.
https://doi.org/10.1214/20-EJS1758.full
.
Ervine
D. A.
,
Babaeyan-Koopaei
K.
&
Sellin
R. H. J.
2000
Two-dimensional solution for straight and meandering overbank flows
.
Journal of Hydraulic Engineering
126
(
9
),
653
669
.
Han
J. T.
,
Wang
Z. R.
&
Yang
Y. T.
2022
Flash flood grading and warning based on dynamic rainfall thresholds
.
Journal of Hydroelectric Engineering
41
(
9
),
67
76
.
Hosseini
Z.
,
Gharechelou
S.
,
Nakhaei
M.
&
Gharechelou
S.
2016
Optimal design of BP algorithm by ACOR model for groundwater-level forecasting: A case study on Shabestar plain, Iran
.
Arabian Journal of Geosciences
9
(
6
),
436
.
Hughes
D. A.
,
Read
L.
,
Jeuland
M.
,
Kapangaziwiri
E.
,
Elkurdy
M.
,
Lambl
D.
,
Hale
E.
&
Opperman
J. J.
2023
Insights from a comparison of two hydrological modelling approaches in the Kwando (Cuando) River and the western tributaries of the Zambezi River basin
.
Journal of Hydrology: Regional Studies
48
,
101482
.
Karami
M.
,
Abedi Koupai
J.
&
Gohari
S. A.
2024
Integration of SWAT, SDSM, AHP, and TOPSIS to detect flood-prone areas
.
Natural Hazards
120
(
7
),
6307
6325
.
Li
L.
,
Zhu
Y. N.
&
Gu
H. X.
2016
Adaptability study on rational formula for Turkish small watershed design flood calculation
.
Journal of China Hydrology
36
(
2
),
41
45 + 6
.
Liu
J.
,
Hong
H.
,
Fan
W.
&
Chen
Y.
2024
A novel fast and efficient adaptive shuffled complex evolution algorithm for model parameter calibration
.
Frontiers in Environmental Science
11
,
1341017
.
Potdar
A.S.
,
Kirstetter
P.-E.
,
Woods
D.
&
Saharia
M.
2021
Towards predicting flood event peak discharge in ungauged basins by learning universal hydrological behaviors with machine learning
.
Journal of Hydrometeorology
,
22
(
11
),
2971
2982
.
Rahimzad
M.
,
Moghaddam Nia
A.
,
Zolfonoon
H.
,
Soltani
J.
,
Danandeh Mehr
A.
&
Kwon
H.-H.
2021
Performance comparison of an LSTM-based deep learning model versus conventional machine learning algorithms for streamflow forecasting
.
Water Resources Management
35
(
12
),
4167
4187
.
Rasheed
Z.
,
Aravamudan
A.
,
Gorji Sefidmazgi
A.
,
Anagnostopoulos
G. C.
&
Nikolopoulos
E. I.
2022
Advancing flood warning procedures in ungauged basins with machine learning
.
Journal of Hydrology
609
,
127736
.
Song
J.
2015
Bias corrections for random forest in regression using residual rotation
.
Journal of the Korean Statistical Society
44
(
2
),
321
326
.
Sun
J.
,
Sang
G. Q.
,
Wang
W. L.
,
Liu
C. J.
&
Zheng
C. Q.
2020
Calculation on the critical rainfall indexes of the small watershed based on semi-distributed hydrological model
.
Journal of China Hydrology
40
(
4
),
1
6
.
Yan
P.
,
Zhang
Z.
,
Hou
Q.
,
Lei
X.
,
Liu
Y.
&
Wang
H.
2023
A novel IBAS-ELM model for prediction of water levels in front of pumping stations
.
Journal of Hydrology
616
,
128810
.
Yao
C.
,
Qiu
Z. Y.
,
Li
Z. J.
,
Hu
W. D.
&
Xu
J.
2019
Parameter regionalization study and application of API model and Xin'anjiang model
.
Journal of Hohai University (Natural Sciences)
47
(
3
),
189
194
.
Ye
J. Y.
,
Li
Z. J.
&
Wu
Y. T.
2013
Study and application of flash flood warning method for ungauged basins
.
Journal of Hydroelectric Engineering
32
(
3
),
15
19 + 33
.
Zhang
Y.
,
Cui
N.
,
Feng
Y.
,
Gong
D.
&
Hu
X.
2019
Comparison of BP, PSO-BP and statistical models for predicting daily global solar radiation in arid Northwest China
.
Computers and Electronics in Agriculture
164
,
104905
.
Zhang
Y.
,
Ragettli
S.
,
Molnar
P.
,
Fink
O.
&
Peleg
N.
2022
Generalization of an encoder-decoder LSTM model for flood prediction in ungauged catchments
.
Journal of Hydrology
614
,
128577
.
Zhang
J.
,
Wang
Y.
,
Niu
F.
,
Zhang
H.
,
Li
S.
&
Wang
Y.
2023
Parameterization of multi-angle shaker based on PSO-BP neural network
.
Minerals
13
(
7
),
929
Zhao
L.
,
Sang
G. Q.
,
Wu
W.
,
Liu
C. J.
&
Wang
J. N.
2022
Critical rainfall prediction model for Mountain Torrent disaster based on random forest regression algorithm
.
Journal of University of Jinan (Science and Technology)
36
(
4
),
404
411 + 423
.
Zsoter
E.
,
Prudhomme
C.
,
Stephens
E.
,
Pappenberger
F.
&
Cloke
H.
2020
Using ensemble reforecasts to generate flood thresholds for improved global flood forecasting
.
Journal of Flood Risk Management
13
(
4
),
e12658
.
https://doi.org/10.1111/jfr3.12658
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Supplementary data