Evaluating the performance of random forest for large-scale flood discharge simulation

Schoppa, Lukas; Disse, Markus; Bachmair, Sophie

doi:10.1016/j.jhydrol.2020.125531

Cited by 110 publications

(45 citation statements)

References 80 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The accuracy of Random Forest is similar to (or even better than) other studies [33,42,43]. The glacier movement provided important information for understanding glacial changes.…”

Section: Comparison With Previous Glacier Classification Methods and supporting

confidence: 72%

Glacier Mapping Based on Random Forest Algorithm: A Case Study over the Eastern Pamir

Zhang

Huang

2020

Water

View full text Add to dashboard Cite

Debris-covered glaciers are common features on the eastern Pamir and serve as important indicators of climate change promptly. However, mapping of debris-covered glaciers in alpine regions is still challenging due to many factors including the spectral similarity between debris and the adjacent bedrock, shadows cast from mountains and clouds, and seasonal snow cover. Considering that few studies have added movement velocity features when extracting glacier boundaries, we innovatively developed an automatic algorithm consisting of rule-based image segmentation and Random Forest to extract information about debris-covered glaciers with Landsat-8 OLI/TIRS data for spectral, texture and temperature features, multi-digital elevation models (DEMs) for elevation and topographic features, and the Inter-mission Time Series of Land Ice Velocity and Elevation (ITS_LIVE) for movement velocity features, and accuracy evaluation was performed to determine the optimal feature combination extraction of debris-covered glaciers. The study found that the overall accuracy of extracting debris-covered glaciers using combined movement velocity features is 97.60%, and the Kappa coefficient is 0.9624, which is better than the extraction results using other schemes. The high classification accuracy obtained using our method overcomes most of the above-mentioned challenges and can detect debris-covered glaciers, illustrating that this method can be executed efficiently, which will further help water resources management.

show abstract

“…The accuracy of Random Forest is similar to (or even better than) other studies [33,42,43]. The glacier movement provided important information for understanding glacial changes.…”

Section: Comparison With Previous Glacier Classification Methods and supporting

confidence: 72%

Glacier Mapping Based on Random Forest Algorithm: A Case Study over the Eastern Pamir

Zhang

Huang

2020

Water

View full text Add to dashboard Cite

show abstract

“…This ensemble, tree-based method has seen previous applications in this field of study. In particular, RF models in the field of hydrology have proven useful in flood risk analysis and susceptibility mapping (Zhao et al 2018), rainfall forecasting (Taksande and Mohod 2015) with performance close to that of Support Vector Machines (Yu et al 2017;Mosavi et al 2018), and, in recent studies, seen as advantageous in large-scale flood discharge simulations (Schoppa et al 2020). These models are less prone to overfitting since an increasing number of base learners leads to a converging generalization error (see Theorem 1.2 in Breiman 2001).…”

Section: Development Of Predictive Modelsmentioning

confidence: 99%

“…The potential for ML-based approaches towards simulating streamflow among other hydrologic applications (Elshorbagy et al 2010;Kasiviswanathan et al 2016;Schoppa et al 2020) has been noted for well over a decade now, but broader exploration was attempted only recently. Over the last few years, the availability of large-sample, high-resolution, observed and simulated hydrometeorological datasets have enabled the analysis of various flood generation processes at the catchment scale along with their drivers.…”

Section: Introductionmentioning

confidence: 99%

Advancing flood warning procedures in ungauged basins with machine learning.

Rasheed¹,

Aravamudan²,

Sefidmazgi³

et al. 2021

Preprint

View full text Add to dashboard Cite

Flood prediction across scales and more specifically in ungauged areas remains still a great challenge that limits the efficiency of flood risk mitigation strategies and disaster preparedness.Building upon the recent success of Machine Learning (ML) models on streamflow prediction, this work presents a prototype ML-based framework for flood warning and flood peak prediction. The fundamental elements of the proposed system consist of a) a LSTM model for classifying storm events to threat/no-threat given a threshold based on the 90th flow percentile and b) the flood peak prediction models. The selected ML-models for flood peak prediction are the Histogram based Gradient Boosting Regressor and the Random Forest. One of the strengths, and reason for selection, of these decision-tree models is their degree of interpretability. This is exploited in the study to help us spatially disentangle the role of both the static and dynamic drivers of flood peak response. Our analysis is presented for 18 distinct hydroclimatic regions across the contiguous US. Results reveal a significant regional dependence on both predictive performance and dominant flood predictors, which emphasize the variability in the complexity of a catchment's hydrologic behavior as well as its impact on modeling flood response. Evaluation of the drivers of flood peaks noted distinct dependencies among the dynamic and static predictors considered in our models for flood peaks of different severity. Specifically, low-moderate flood events showed a clear preponderance for the static catchment attributes over dynamic predictors like precipitation whereas precipitation was the dominant driver for the high severity flood peaks. The proposed flood peak prediction models were compared against a state-of-the-art LSTM model and were shown to consistently outperform in ungauged basins. Overall, the proposed system classified storms correctly for >85% in all cases and exhibited a percent relative difference in flood peak estimates of <30% in most cases.

show abstract

“…However, the frequency peak in the training was concentrated close to 0.020 m 3 s -1 km -2 , while in validation was close to 0.025 m 3 s -1 km -2 in the Qm prediction by the RF model. Thus, the training in samples with a greater peak in lower flow rates could explain the overestimation in the validation prediction, as the predictive power of RF completely depends on the training experience (Schoppa et al, 2020).…”

Section: Figure 13mentioning

confidence: 99%

“…The use of RF in water resources is recent (Tyralis et al, 2019). The RF has been using on water price prediction (Xu et al, 2019), regionalization of hourly hydrological model parameters (Saadi et al, 2019), large-scale flood discharge simulation (Schoppa et al, 2020), and several hydrological parameters and signatures (Booker and Woods, 2014;Addor et al, 2018;Booker and Snelder, 2012). However, few studies focused on applying RF to predict specific quantiles along the flow duration curve (for example, Schnier and Cai, 2014).…”

Section: Introductionmentioning

confidence: 99%

Regionalization of hydrological variables for the Paraná state, Brazil

Nascimento¹

View full text Add to dashboard Cite

show abstract

Evaluating the performance of random forest for large-scale flood discharge simulation

Cited by 110 publications

References 80 publications

Glacier Mapping Based on Random Forest Algorithm: A Case Study over the Eastern Pamir

Glacier Mapping Based on Random Forest Algorithm: A Case Study over the Eastern Pamir

Advancing flood warning procedures in ungauged basins with machine learning.

Regionalization of hydrological variables for the Paraná state, Brazil

Contact Info

Product

Resources

About