Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning
The process of evapotranspiration transfers liquid water from vegetation and soil surfaces to the atmosphere, the so-called latent heat flux ( ${Q_{{\text{LE}}}}$ ), and modulates the Earth’s energy, water, and carbon cycle. Vegetation controls ${Q_{{\text{LE}}}}$ by regulating leaf stomata opening...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IOP Publishing
2023-01-01
|
Series: | Environmental Research Letters |
Subjects: | |
Online Access: | https://doi.org/10.1088/1748-9326/acbbe0 |
_version_ | 1797747394839838720 |
---|---|
author | Reda ElGhawi Basil Kraft Christian Reimers Markus Reichstein Marco Körner Pierre Gentine Alexander J Winkler |
author_facet | Reda ElGhawi Basil Kraft Christian Reimers Markus Reichstein Marco Körner Pierre Gentine Alexander J Winkler |
author_sort | Reda ElGhawi |
collection | DOAJ |
description | The process of evapotranspiration transfers liquid water from vegetation and soil surfaces to the atmosphere, the so-called latent heat flux ( ${Q_{{\text{LE}}}}$ ), and modulates the Earth’s energy, water, and carbon cycle. Vegetation controls ${Q_{{\text{LE}}}}$ by regulating leaf stomata opening (surface resistance ${r_{\text{s}}}$ in the Big Leaf approach) and by altering surface roughness (aerodynamic resistance ${r_{\text{a}}}$ ). Estimating ${r_{\text{s}}}$ and ${r_{\text{a}}}$ across different vegetation types is a key challenge in predicting ${Q_{{\text{LE}}}}$ . We propose a hybrid approach that combines mechanistic modeling and machine learning for modeling ${Q_{{\text{LE}}}}$ . The hybrid model combines a feed-forward neural network which estimates the resistances from observations as intermediate variables and a mechanistic model in an end-to-end setting. In the hybrid modeling setup, we make use of the Penman–Monteith equation in conjunction with multi-year flux measurements across different forest and grassland sites from the FLUXNET database. This hybrid model setup is successful in predicting ${Q_{{\text{LE}}}}$ , however, this approach leads to equifinal solutions in terms of estimated physical parameters. We follow two different strategies to constrain the hybrid model and therefore control for the equifinality that arises when the two resistances are estimated simultaneously. One strategy is to impose an a priori constraint on ${r_{\text{a}}}$ based on mechanistic assumptions (theory-driven strategy), while the other strategy makes use of more observational data and adds a constraint in predicting ${r_{\text{a}}}$ through multi-task learning of both latent and sensible heat flux ( ${Q_{\text{H}}}$ ; data-driven strategy) together. Our results show that all hybrid models predict the target variables with a high degree of success, with ${R^2}$ = 0.82–0.89 for grasslands and ${R^2}$ = 0.70–0.80 for forest sites at the mean diurnal scale. The predicted ${r_{\text{s}}}$ and ${r_{\text{a}}}$ show strong physical consistency across the two regularized hybrid models, but are physically implausible in the under-constrained hybrid model. The hybrid models are robust in reproducing consistent results for energy fluxes and resistances across different scales (diurnal, seasonal, and interannual), reflecting their ability to learn the physical dependence of the target variables on the meteorological inputs. As a next step, we propose to test these heavily observation-informed parameterizations derived through hybrid modeling as a substitute for ad hoc formulations in Earth system models. |
first_indexed | 2024-03-12T15:50:59Z |
format | Article |
id | doaj.art-792cd987f1fa44188abc41e155309956 |
institution | Directory Open Access Journal |
issn | 1748-9326 |
language | English |
last_indexed | 2024-03-12T15:50:59Z |
publishDate | 2023-01-01 |
publisher | IOP Publishing |
record_format | Article |
series | Environmental Research Letters |
spelling | doaj.art-792cd987f1fa44188abc41e1553099562023-08-09T15:13:56ZengIOP PublishingEnvironmental Research Letters1748-93262023-01-0118303403910.1088/1748-9326/acbbe0Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learningReda ElGhawi0https://orcid.org/0000-0003-2930-4537Basil Kraft1Christian Reimers2Markus Reichstein3https://orcid.org/0000-0001-5736-1112Marco Körner4https://orcid.org/0000-0002-9186-4175Pierre Gentine5https://orcid.org/0000-0002-0845-8345Alexander J Winkler6https://orcid.org/0000-0001-6574-4471Max Planck Institute for Biogeochemistry, Biogeochemical Integration , Jena, Germany; International Max Planck Research School for Global Biogeochemical Cycles, Max Planck Institute for Biogeochemistry , Jena, Germany; Department of Aerospace and Geodesy, School of Engineering and Design, Technical University of Munich , Munich, GermanyMax Planck Institute for Biogeochemistry, Biogeochemical Integration , Jena, GermanyMax Planck Institute for Biogeochemistry, Biogeochemical Integration , Jena, GermanyMax Planck Institute for Biogeochemistry, Biogeochemical Integration , Jena, Germany; ELLIS Unit Jena, Michael-Stifel-Center, University of Jena , Jena, GermanyDepartment of Aerospace and Geodesy, School of Engineering and Design, Technical University of Munich , Munich, GermanyDepartment of Earth and Environmental Engineering, Columbia University , NY, New York, 10027, United States of AmericaMax Planck Institute for Biogeochemistry, Biogeochemical Integration , Jena, GermanyThe process of evapotranspiration transfers liquid water from vegetation and soil surfaces to the atmosphere, the so-called latent heat flux ( ${Q_{{\text{LE}}}}$ ), and modulates the Earth’s energy, water, and carbon cycle. Vegetation controls ${Q_{{\text{LE}}}}$ by regulating leaf stomata opening (surface resistance ${r_{\text{s}}}$ in the Big Leaf approach) and by altering surface roughness (aerodynamic resistance ${r_{\text{a}}}$ ). Estimating ${r_{\text{s}}}$ and ${r_{\text{a}}}$ across different vegetation types is a key challenge in predicting ${Q_{{\text{LE}}}}$ . We propose a hybrid approach that combines mechanistic modeling and machine learning for modeling ${Q_{{\text{LE}}}}$ . The hybrid model combines a feed-forward neural network which estimates the resistances from observations as intermediate variables and a mechanistic model in an end-to-end setting. In the hybrid modeling setup, we make use of the Penman–Monteith equation in conjunction with multi-year flux measurements across different forest and grassland sites from the FLUXNET database. This hybrid model setup is successful in predicting ${Q_{{\text{LE}}}}$ , however, this approach leads to equifinal solutions in terms of estimated physical parameters. We follow two different strategies to constrain the hybrid model and therefore control for the equifinality that arises when the two resistances are estimated simultaneously. One strategy is to impose an a priori constraint on ${r_{\text{a}}}$ based on mechanistic assumptions (theory-driven strategy), while the other strategy makes use of more observational data and adds a constraint in predicting ${r_{\text{a}}}$ through multi-task learning of both latent and sensible heat flux ( ${Q_{\text{H}}}$ ; data-driven strategy) together. Our results show that all hybrid models predict the target variables with a high degree of success, with ${R^2}$ = 0.82–0.89 for grasslands and ${R^2}$ = 0.70–0.80 for forest sites at the mean diurnal scale. The predicted ${r_{\text{s}}}$ and ${r_{\text{a}}}$ show strong physical consistency across the two regularized hybrid models, but are physically implausible in the under-constrained hybrid model. The hybrid models are robust in reproducing consistent results for energy fluxes and resistances across different scales (diurnal, seasonal, and interannual), reflecting their ability to learn the physical dependence of the target variables on the meteorological inputs. As a next step, we propose to test these heavily observation-informed parameterizations derived through hybrid modeling as a substitute for ad hoc formulations in Earth system models.https://doi.org/10.1088/1748-9326/acbbe0hybrid modelingphysics-constrainedmachine learningmulti-task learningevapotranspirationsurface resistance |
spellingShingle | Reda ElGhawi Basil Kraft Christian Reimers Markus Reichstein Marco Körner Pierre Gentine Alexander J Winkler Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning Environmental Research Letters hybrid modeling physics-constrained machine learning multi-task learning evapotranspiration surface resistance |
title | Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning |
title_full | Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning |
title_fullStr | Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning |
title_full_unstemmed | Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning |
title_short | Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning |
title_sort | hybrid modeling of evapotranspiration inferring stomatal and aerodynamic resistances using combined physics based and machine learning |
topic | hybrid modeling physics-constrained machine learning multi-task learning evapotranspiration surface resistance |
url | https://doi.org/10.1088/1748-9326/acbbe0 |
work_keys_str_mv | AT redaelghawi hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT basilkraft hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT christianreimers hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT markusreichstein hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT marcokorner hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT pierregentine hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning AT alexanderjwinkler hybridmodelingofevapotranspirationinferringstomatalandaerodynamicresistancesusingcombinedphysicsbasedandmachinelearning |