<abstract xmlns="http://www.w3.org/1999/xhtml"><p>The event runoff coefficient (<italic>Rc</italic>) and the recession coefficient (<italic>tc</italic>) are of theoretical importance for understanding catchment response and of practical importance in hydrological design. We analyse 57 event periods in the period 2013 to 2015 in the 66 ha Austrian Hydrological Open Air Laboratory (HOAL), where the seven subcatchments are stratified by runoff generation types into wetlands, tile drainage and natural drainage. Three machine learning algorithms (Random forest (RF), Gradient Boost Decision Tree (GBDT) and Support vector machine (SVM)) are used to estimate <italic>Rc</italic> and <italic>tc</italic> from 22 event based explanatory variables representing precipitation, soil moisture, groundwater level and season. The model performance of the SVM algorithm in estimating <italic>Rc</italic> and <italic>tc</italic> is generally higher than that of the other two methods, measured by the coefficient of determination <italic>R<sup>2</sup></italic>, and the performance for <italic>Rc</italic> is higher than that for <italic>tc</italic>. The relative importance of the explanatory variables for the predictions, assessed by a heatmap, suggests that <italic>Rc</italic> of the tile drainage systems is more strongly controlled by the weather conditions than by the catchment state, while the opposite is true for natural drainage systems. Overall, model performance strongly depends on the runoff generation type.</p></abstract>

The event runoff coefficient (Rc) and the recession coefficient (tc) are of theoretical importance for understanding catchment response and of practical importance in hydrological design. We analyse 57 event periods in the period 2013 to 2015 in the 66 ha Austrian Hydrological Open Air Laboratory (HOAL), where the seven subcatchments are stratified by runoff generation types into wetlands, tile drainage and natural drainage. Three machine learning algorithms (Random forest (RF), Gradient Boost Decision Tree (GBDT) and Support vector machine (SVM)) are used to estimate Rc and tc from 22 event based explanatory variables representing precipitation, soil moisture, groundwater level and season. The model performance of the SVM algorithm in estimating Rc and tc is generally higher than that of the other two methods, measured by the coefficient of determination R2, and the performance for Rc is higher than that for tc. The relative importance of the explanatory variables for the predictions, assessed by a heatmap, suggests that Rc of the tile drainage systems is more strongly controlled by the weather conditions than by the catchment state, while the opposite is true for natural drainage systems. Overall, model performance strongly depends on the runoff generation type.

Controls on event runoff coefficients and recession coefficients for different runoff generation mechanisms identified by three regression methods

Institute of Hydraulic Engineering and Water Resources Management

Federal Agency for Water Management, Institute for Land and Water Management Research

Journal of Hydrology and Hydromechanics

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

The event runoff coefficient (Rc) and the recession coefficient (tc) are of theoretical importance for understanding catchment response and of practical...

Controls on event runoff coefficients and recession coefficients for different runoff generation mechanisms identified by three regression methods

Publié en ligne: 26 mai 2020

Pages: 155 - 169

Reçu: 09 mai 2019

Accepté: 30 janv. 2020

DOI: https://doi.org/10.2478/johh-2020-0008

Mots clés
Machine learning, Event runoff analyses, Event runoff coefficient, Recession coefficient, Runoff generation

© 2020 Xiaofei Chen et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Controls on event runoff coefficients and recession coefficients for different runoff generation mechanisms identified by three regression methods

Publié en ligne: 26 mai 2020

Pages: 155 - 169

Reçu: 09 mai 2019

Accepté: 30 janv. 2020

DOI: https://doi.org/10.2478/johh-2020-0008

Mots clésMachine learning, Event runoff analyses, Event runoff coefficient, Recession coefficient, Runoff generation

© 2020 Xiaofei Chen et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Mots clés
Machine learning, Event runoff analyses, Event runoff coefficient, Recession coefficient, Runoff generation