16 Jan 2023
 | 16 Jan 2023

Understanding hydrologic controls of slope response to precipitations through Machine Learning analysis applied to synthetic data

Daniel Camilo Roman Quintero, Pasquale Marino, Giovanni Francesco Santonastaso, and Roberto Greco

Abstract. The assessment of the response of slopes to precipitation is important for various applications, from water supply management to hazard assessment due to extreme rainfall events. It is well known that the underground conditions prior to the initiation of rainfall events control the hydrological processes that occur in slopes, affecting the water exchange through their boundaries. The present study aims at identifying suitable variables to be monitored and modelled to predict the response of the slope to precipitations. A case study consisting of a loose pyroclastic coarse grained soil cover overlaying a karstic bedrock located in the southern Apennines (Italy) is described, where field monitoring has been carried out, comprising stream level recordings, meteorological recordings, and soil water content among others. Nevertheless, to enhance the field dataset, the slope hydraulic behaviour of the case study has been simulated with a physically based model linked to a synthetic rainfall time series, getting a consistent hourly timeseries dataset of 1000 years, containing information on rainfall, aquifer water level and soil volumetric water content at different depths. Machine Learning techniques have been used to unwrap the relationships amongst the studied variables, which relations are commonly non-linear. The Random Forest technique has been used to assess the way the slope response could be addressed and the importance of each variable on the slope response and the k-means clustering technique has been used to explore the geometrical disposition of data, and so the identification of seasonally recurrent different scenarios linked to the slope response. It has been shown that the slope response in terms of the rainwater being stored in the soil cover is naturally highly dependent on the rainfall amount, but water drainage and storage processes can be identified by normalizing the change in water storage with the rainfall depth. Indeed, with the methodology presented here, different hydrometeorological scenarios controlling major hydrological processes have been identified not only from the meteorological and seasonal behaviour but also from the underground conditions prior to the rainfall initiation, weighting the role, on one hand, of the field capacity value on the ease of the water to flow in and out of the soil cover and, on the other hand, of the ground water level, the increase of which gives evidence of the activation of slope drainage even during relatively intense rainfall events.

Daniel Camilo Roman Quintero et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2022-1078', Anonymous Referee #1, 15 Mar 2023
    • AC1: 'Reply on RC1', Daniel Camilo Roman Quintero, 28 Apr 2023
  • RC2: 'Comment on egusphere-2022-1078', Anonymous Referee #2, 20 Mar 2023
    • AC2: 'Reply on RC2', Daniel Camilo Roman Quintero, 28 Apr 2023
  • RC3: 'Comment on egusphere-2022-1078', Anonymous Referee #3, 27 Mar 2023
    • AC3: 'Reply on RC3', Daniel Camilo Roman Quintero, 28 Apr 2023

Daniel Camilo Roman Quintero et al.

Daniel Camilo Roman Quintero et al.


Total article views: 572 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
426 120 26 572 10 9
  • HTML: 426
  • PDF: 120
  • XML: 26
  • Total: 572
  • BibTeX: 10
  • EndNote: 9
Views and downloads (calculated since 16 Jan 2023)
Cumulative views and downloads (calculated since 16 Jan 2023)

Viewed (geographical distribution)

Total article views: 563 (including HTML, PDF, and XML) Thereof 563 with geography defined and 0 with unknown origin.
Country # Views %
  • 1


Latest update: 26 Sep 2023
Short summary
This study shows a methodological approach using Machine Learning techniques to disentangle the relationships among the variables in a synthetic dataset, to identify the suitable variables that control the hydrologic response of the slopes. It has been found that not only the rainfall is responsible for the water accumulation in the slope; the underground conditions (soil water content and aquifer water level) indicate the activation of natural slope drainage mechanisms.