Integrated surface-subsurface model to investigate the role of groundwater in headwater catchment runoff generation: a minimalist approach to parameterisation

Pertti Ala-aho, Chris Soulsby, Hailong Wang, Doerthe Tetzlaff

Research output: Contribution to journalArticle

25 Citations (Scopus)
3 Downloads (Pure)

Abstract

Understanding the role of groundwater for runoff generation in headwater catchments is a challenge in hydrology, particularly so in data-scarce areas. Fully-integrated surface-subsurface modelling has shown potential in increasing process understanding for runoff generation, but high data requirements and difficulties in model calibration are typically assumed to preclude their use in catchment-scale studies. We used a fully integrated surface-subsurface hydrological simulator to enhance groundwater-related process understanding in a headwater catchment with a rich background in empirical data. To set up the model we used minimal data that could be reasonably expected to exist for any experimental catchment. A novel aspect of our approach was in using simplified model parameterisation and including parameters from all model domains (surface, subsurface, evapotranspiration) in automated model calibration. Calibration aimed not only to improve model fit, but also to test the information content of the observations (streamflow, remotely sensed evapotranspiration, median groundwater level) used in calibration objective functions. We identified sensitive parameters in all model domains (subsurface, surface, evapotranspiration), demonstrating that model calibration should be inclusive of parameters from these different model domains. Incorporating groundwater data in calibration objectives improved the model fit for groundwater levels, but simulations did not reproduce well the remotely sensed evapotranspiration time series even after calibration. Spatially explicit model output improved our understanding of how groundwater functions in maintaining streamflow generation primarily via saturation excess overland flow. Steady groundwater inputs created saturated conditions in the valley bottom riparian peatlands, leading to overland flow even during dry periods. Groundwater on the hillslopes was more dynamic in its response to rainfall, acting to expand the saturated area extent and thereby promoting saturation excess overland flow during rainstorms. Our work shows the potential of using integrated surface-subsurface modelling alongside with rigorous model calibration to better understand and visualise the role of groundwater in runoff generation even with limited datasets.
Original languageEnglish
Pages (from-to)664-677
Number of pages14
JournalJournal of Hydrology
Volume547
Early online date20 Feb 2017
DOIs
Publication statusPublished - Apr 2017

Fingerprint

headwater
evapotranspiration
parameterization
catchment
runoff
groundwater
calibration
overland flow
streamflow
saturation
rainstorm
hillslope
peatland
modeling
simulator
hydrology
time series
valley
rainfall

Keywords

  • integrated modelling
  • streamflow generation
  • groundwater
  • catchment modelling
  • model calibration

Cite this

@article{56dac0ab764c473ebc52623255749ffd,
title = "Integrated surface-subsurface model to investigate the role of groundwater in headwater catchment runoff generation: a minimalist approach to parameterisation",
abstract = "Understanding the role of groundwater for runoff generation in headwater catchments is a challenge in hydrology, particularly so in data-scarce areas. Fully-integrated surface-subsurface modelling has shown potential in increasing process understanding for runoff generation, but high data requirements and difficulties in model calibration are typically assumed to preclude their use in catchment-scale studies. We used a fully integrated surface-subsurface hydrological simulator to enhance groundwater-related process understanding in a headwater catchment with a rich background in empirical data. To set up the model we used minimal data that could be reasonably expected to exist for any experimental catchment. A novel aspect of our approach was in using simplified model parameterisation and including parameters from all model domains (surface, subsurface, evapotranspiration) in automated model calibration. Calibration aimed not only to improve model fit, but also to test the information content of the observations (streamflow, remotely sensed evapotranspiration, median groundwater level) used in calibration objective functions. We identified sensitive parameters in all model domains (subsurface, surface, evapotranspiration), demonstrating that model calibration should be inclusive of parameters from these different model domains. Incorporating groundwater data in calibration objectives improved the model fit for groundwater levels, but simulations did not reproduce well the remotely sensed evapotranspiration time series even after calibration. Spatially explicit model output improved our understanding of how groundwater functions in maintaining streamflow generation primarily via saturation excess overland flow. Steady groundwater inputs created saturated conditions in the valley bottom riparian peatlands, leading to overland flow even during dry periods. Groundwater on the hillslopes was more dynamic in its response to rainfall, acting to expand the saturated area extent and thereby promoting saturation excess overland flow during rainstorms. Our work shows the potential of using integrated surface-subsurface modelling alongside with rigorous model calibration to better understand and visualise the role of groundwater in runoff generation even with limited datasets.",
keywords = "integrated modelling, streamflow generation, groundwater, catchment modelling, model calibration",
author = "Pertti Ala-aho and Chris Soulsby and Hailong Wang and Doerthe Tetzlaff",
note = "This work was funded by NERC/JPI SIWA project (NE/M019896/1) and the European Research Council ERC (project GA 335910 VeWa). Numerical simulations were performed using the Maxwell High Performance Computing Cluster of the University of Aberdeen IT Service, provided by Dell Inc. and supported by Alces Software. Aquanty Inc. is acknowledged for support in providing HGS simulation software compatible with the Maxwell High Performance Computing Cluster. We would also like to thank the anonymous reviewers for their constructive comments that improved the manuscript.",
year = "2017",
month = "4",
doi = "10.1016/j.jhydrol.2017.02.023",
language = "English",
volume = "547",
pages = "664--677",
journal = "Journal of Hydrology",
issn = "0022-1694",
publisher = "Elsevier Science B. V.",

}

TY - JOUR

T1 - Integrated surface-subsurface model to investigate the role of groundwater in headwater catchment runoff generation

T2 - a minimalist approach to parameterisation

AU - Ala-aho, Pertti

AU - Soulsby, Chris

AU - Wang, Hailong

AU - Tetzlaff, Doerthe

N1 - This work was funded by NERC/JPI SIWA project (NE/M019896/1) and the European Research Council ERC (project GA 335910 VeWa). Numerical simulations were performed using the Maxwell High Performance Computing Cluster of the University of Aberdeen IT Service, provided by Dell Inc. and supported by Alces Software. Aquanty Inc. is acknowledged for support in providing HGS simulation software compatible with the Maxwell High Performance Computing Cluster. We would also like to thank the anonymous reviewers for their constructive comments that improved the manuscript.

PY - 2017/4

Y1 - 2017/4

N2 - Understanding the role of groundwater for runoff generation in headwater catchments is a challenge in hydrology, particularly so in data-scarce areas. Fully-integrated surface-subsurface modelling has shown potential in increasing process understanding for runoff generation, but high data requirements and difficulties in model calibration are typically assumed to preclude their use in catchment-scale studies. We used a fully integrated surface-subsurface hydrological simulator to enhance groundwater-related process understanding in a headwater catchment with a rich background in empirical data. To set up the model we used minimal data that could be reasonably expected to exist for any experimental catchment. A novel aspect of our approach was in using simplified model parameterisation and including parameters from all model domains (surface, subsurface, evapotranspiration) in automated model calibration. Calibration aimed not only to improve model fit, but also to test the information content of the observations (streamflow, remotely sensed evapotranspiration, median groundwater level) used in calibration objective functions. We identified sensitive parameters in all model domains (subsurface, surface, evapotranspiration), demonstrating that model calibration should be inclusive of parameters from these different model domains. Incorporating groundwater data in calibration objectives improved the model fit for groundwater levels, but simulations did not reproduce well the remotely sensed evapotranspiration time series even after calibration. Spatially explicit model output improved our understanding of how groundwater functions in maintaining streamflow generation primarily via saturation excess overland flow. Steady groundwater inputs created saturated conditions in the valley bottom riparian peatlands, leading to overland flow even during dry periods. Groundwater on the hillslopes was more dynamic in its response to rainfall, acting to expand the saturated area extent and thereby promoting saturation excess overland flow during rainstorms. Our work shows the potential of using integrated surface-subsurface modelling alongside with rigorous model calibration to better understand and visualise the role of groundwater in runoff generation even with limited datasets.

AB - Understanding the role of groundwater for runoff generation in headwater catchments is a challenge in hydrology, particularly so in data-scarce areas. Fully-integrated surface-subsurface modelling has shown potential in increasing process understanding for runoff generation, but high data requirements and difficulties in model calibration are typically assumed to preclude their use in catchment-scale studies. We used a fully integrated surface-subsurface hydrological simulator to enhance groundwater-related process understanding in a headwater catchment with a rich background in empirical data. To set up the model we used minimal data that could be reasonably expected to exist for any experimental catchment. A novel aspect of our approach was in using simplified model parameterisation and including parameters from all model domains (surface, subsurface, evapotranspiration) in automated model calibration. Calibration aimed not only to improve model fit, but also to test the information content of the observations (streamflow, remotely sensed evapotranspiration, median groundwater level) used in calibration objective functions. We identified sensitive parameters in all model domains (subsurface, surface, evapotranspiration), demonstrating that model calibration should be inclusive of parameters from these different model domains. Incorporating groundwater data in calibration objectives improved the model fit for groundwater levels, but simulations did not reproduce well the remotely sensed evapotranspiration time series even after calibration. Spatially explicit model output improved our understanding of how groundwater functions in maintaining streamflow generation primarily via saturation excess overland flow. Steady groundwater inputs created saturated conditions in the valley bottom riparian peatlands, leading to overland flow even during dry periods. Groundwater on the hillslopes was more dynamic in its response to rainfall, acting to expand the saturated area extent and thereby promoting saturation excess overland flow during rainstorms. Our work shows the potential of using integrated surface-subsurface modelling alongside with rigorous model calibration to better understand and visualise the role of groundwater in runoff generation even with limited datasets.

KW - integrated modelling

KW - streamflow generation

KW - groundwater

KW - catchment modelling

KW - model calibration

U2 - 10.1016/j.jhydrol.2017.02.023

DO - 10.1016/j.jhydrol.2017.02.023

M3 - Article

VL - 547

SP - 664

EP - 677

JO - Journal of Hydrology

JF - Journal of Hydrology

SN - 0022-1694

ER -