Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Improving evapotranspiration estimation by integrating process-based biophysical variables into a deep learning approach

Xie, Mingming LU ; Zhang, Jianyun ; Bao, Zhenxin ; Zhang, Linus LU orcid ; Duan, Zheng LU ; Wang, Guoqing LU ; Liu, Cuishan ; Yuan, Feifei LU and Guan, Xiaoxiang (2026) In Journal of Hydrology: Regional Studies 63. p.103114-103114
Abstract
Study Region 103 FLUXNET2015 flux towers distributed across diverse climatic and ecological regions. Study Focus Accurate estimation of evapotranspiration (ET) is critical for understanding regional ecohydrological processes. Physically based models such as the Penman-Monteith-Leuning (PML) model are robust but often constrained by fixed parameterization schemes, while data-driven approaches such as Long Short-Term Memory (LSTM) networks can capture nonlinearities but depend heavily on training data. To address these limitations, this study developed a hybrid model (PML-LSTM) by integrating biophysical variables from PML simulation into LSTM network. Model performance was systematically evaluated against standalone PML and LSTM across... (More)
Study Region 103 FLUXNET2015 flux towers distributed across diverse climatic and ecological regions. Study Focus Accurate estimation of evapotranspiration (ET) is critical for understanding regional ecohydrological processes. Physically based models such as the Penman-Monteith-Leuning (PML) model are robust but often constrained by fixed parameterization schemes, while data-driven approaches such as Long Short-Term Memory (LSTM) networks can capture nonlinearities but depend heavily on training data. To address these limitations, this study developed a hybrid model (PML-LSTM) by integrating biophysical variables from PML simulation into LSTM network. Model performance was systematically evaluated against standalone PML and LSTM across three modelling levels: local (site), type (vegetation type), and group (forest and non-forest). New Hydrological Insights for the Region The PML-LSTM model achieved superior performance, with median NSE values of 0.851, 0.913, and 0.933 during validation, surpassing both the PML model (0.843, 0.788, 0.766) and the LSTM model (0.818, 0.879, 0.873). Integrating biophysical information in the PML-LSTM model improved ET estimation accuracy and model generalization, leading to more robust spatiotemporal performance under leave-one-out cross-validation and extreme weather extrapolation experiments. Distinct model behaviors emerged under varying sample conditions: the PML model exhibited greater robustness under data-scarce conditions at the local level, while the LSTM and PML-LSTM models benefited from larger training datasets. This study highlights the potential of combining process-based and data-driven approaches to improve ET estimation and provides insights for regional ecohydrological modelling. (Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Evapotranspiration estimation, Deep learning, Long short-term memory, Penman-monteith-leuning, Hybrid model
in
Journal of Hydrology: Regional Studies
volume
63
pages
1 pages
publisher
Elsevier
ISSN
2214-5818
DOI
10.1016/j.ejrh.2026.103114
language
English
LU publication?
yes
id
503966dc-e3c6-4b8c-9064-ad79375adf37
date added to LUP
2026-01-09 11:46:06
date last changed
2026-01-15 12:16:30
@article{503966dc-e3c6-4b8c-9064-ad79375adf37,
  abstract     = {{Study Region 103 FLUXNET2015 flux towers distributed across diverse climatic and ecological regions. Study Focus Accurate estimation of evapotranspiration (ET) is critical for understanding regional ecohydrological processes. Physically based models such as the Penman-Monteith-Leuning (PML) model are robust but often constrained by fixed parameterization schemes, while data-driven approaches such as Long Short-Term Memory (LSTM) networks can capture nonlinearities but depend heavily on training data. To address these limitations, this study developed a hybrid model (PML-LSTM) by integrating biophysical variables from PML simulation into LSTM network. Model performance was systematically evaluated against standalone PML and LSTM across three modelling levels: local (site), type (vegetation type), and group (forest and non-forest). New Hydrological Insights for the Region The PML-LSTM model achieved superior performance, with median NSE values of 0.851, 0.913, and 0.933 during validation, surpassing both the PML model (0.843, 0.788, 0.766) and the LSTM model (0.818, 0.879, 0.873). Integrating biophysical information in the PML-LSTM model improved ET estimation accuracy and model generalization, leading to more robust spatiotemporal performance under leave-one-out cross-validation and extreme weather extrapolation experiments. Distinct model behaviors emerged under varying sample conditions: the PML model exhibited greater robustness under data-scarce conditions at the local level, while the LSTM and PML-LSTM models benefited from larger training datasets. This study highlights the potential of combining process-based and data-driven approaches to improve ET estimation and provides insights for regional ecohydrological modelling.}},
  author       = {{Xie, Mingming and Zhang, Jianyun and Bao, Zhenxin and Zhang, Linus and Duan, Zheng and Wang, Guoqing and Liu, Cuishan and Yuan, Feifei and Guan, Xiaoxiang}},
  issn         = {{2214-5818}},
  keywords     = {{Evapotranspiration estimation; Deep learning; Long short-term memory; Penman-monteith-leuning; Hybrid model}},
  language     = {{eng}},
  pages        = {{103114--103114}},
  publisher    = {{Elsevier}},
  series       = {{Journal of Hydrology: Regional Studies}},
  title        = {{Improving evapotranspiration estimation by integrating process-based biophysical variables into a deep learning approach}},
  url          = {{http://dx.doi.org/10.1016/j.ejrh.2026.103114}},
  doi          = {{10.1016/j.ejrh.2026.103114}},
  volume       = {{63}},
  year         = {{2026}},
}