Automated actual evapotranspiration estimation: Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms

Alireza Ghaderi Bafti; Arman Ahmadi; Ali Abbasi; Hamid Kamangir; Sadegh Jamali; Hossein Hashemi

doi:10.1016/j.atmosres.2023.107107

Automated actual evapotranspiration estimation: Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms

Alireza Ghaderi Bafti^*, Arman Ahmadi, Ali Abbasi, Hamid Kamangir, Sadegh Jamali, Hossein Hashemi

^*Corresponding author for this work

Water Resources

Research output: Contribution to journal › Article › Scientific › peer-review

Abstract

Actual evapotranspiration (ET_a) plays a crucial role in the water and energy cycles of the earth. An accurate estimate of the ET_a is essential for management of the water resources, agriculture, and irrigation, as well as research on atmospheric variations. Despite the importance of accurate ET_a values, estimating and mapping them remains challenging due to the physical and biological complexity of the ET process. As a novel approach for rapid and reliable estimation of ET_a, the present study develops automated deep learning (AutoDL) models that incorporate a metaheuristic optimization algorithm for image processing, architectural design, and hyperparameter tuning. The proposed AutoDL models integrate three different spatial and channel attention mechanisms, including a novel activated spatial attention mechanism (ASPAM), with the U-Net architecture. Bypassing the need for meteorological inputs, the proposed framework uses Moderate Resolution Imaging Spectrometer (MODIS) products and Digital Elevation Model (DEM) data as inputs. To evaluate the performance of the models, they are applied to three study areas in the United States with various climatic characteristics. According to the results, during the spring and summer, when the target values have higher certainty, the estimations are highly promising, with R² as high as 0.91 and MAPE as low as 6.40%. Furthermore, the proposed ASPAM module improves the accuracy of ET_a estimations compared to attention gate (AG) and squeeze and excitation (SE) attention modules. The results also indicate that the MODIS raw products and derived vegetation and water indices can predict ET_a within a reliable range of accuracy, with the addition of DEM data marginally enhancing the models' performance. The automatic workflow of this model makes it significantly easy to use, contributing to its applicability and generalizability for enhancing atmospheric research.

Original language	English
Article number	107107
Number of pages	16
Journal	Atmospheric Research
Volume	297
DOIs	https://doi.org/10.1016/j.atmosres.2023.107107
Publication status	Published - 2023

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Actual evapotranspiration mapping
Attention mechanism
Deep learning
Harris Hawks Optimization
Simplified surface energy balance (SSEBop)
U-Net architecture

Access to Document

10.1016/j.atmosres.2023.107107

Cite this

@article{4783b4180e754cd2930033782bef8cfd,

title = "Automated actual evapotranspiration estimation: Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms",

abstract = "Actual evapotranspiration (ETa) plays a crucial role in the water and energy cycles of the earth. An accurate estimate of the ETa is essential for management of the water resources, agriculture, and irrigation, as well as research on atmospheric variations. Despite the importance of accurate ETa values, estimating and mapping them remains challenging due to the physical and biological complexity of the ET process. As a novel approach for rapid and reliable estimation of ETa, the present study develops automated deep learning (AutoDL) models that incorporate a metaheuristic optimization algorithm for image processing, architectural design, and hyperparameter tuning. The proposed AutoDL models integrate three different spatial and channel attention mechanisms, including a novel activated spatial attention mechanism (ASPAM), with the U-Net architecture. Bypassing the need for meteorological inputs, the proposed framework uses Moderate Resolution Imaging Spectrometer (MODIS) products and Digital Elevation Model (DEM) data as inputs. To evaluate the performance of the models, they are applied to three study areas in the United States with various climatic characteristics. According to the results, during the spring and summer, when the target values have higher certainty, the estimations are highly promising, with R2 as high as 0.91 and MAPE as low as 6.40%. Furthermore, the proposed ASPAM module improves the accuracy of ETa estimations compared to attention gate (AG) and squeeze and excitation (SE) attention modules. The results also indicate that the MODIS raw products and derived vegetation and water indices can predict ETa within a reliable range of accuracy, with the addition of DEM data marginally enhancing the models' performance. The automatic workflow of this model makes it significantly easy to use, contributing to its applicability and generalizability for enhancing atmospheric research.",

keywords = "Actual evapotranspiration mapping, Attention mechanism, Deep learning, Harris Hawks Optimization, Simplified surface energy balance (SSEBop), U-Net architecture",

author = "Bafti, {Alireza Ghaderi} and Arman Ahmadi and Ali Abbasi and Hamid Kamangir and Sadegh Jamali and Hossein Hashemi",

note = "Green Open Access added to TU Delft Institutional Repository {\textquoteleft}You share, we take care!{\textquoteright} – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. ",

year = "2023",

doi = "10.1016/j.atmosres.2023.107107",

language = "English",

volume = "297",

journal = "Atmospheric Research",

issn = "0169-8095",

publisher = "Elsevier",

}

TY - JOUR

T1 - Automated actual evapotranspiration estimation

T2 - Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms

AU - Bafti, Alireza Ghaderi

AU - Ahmadi, Arman

AU - Abbasi, Ali

AU - Kamangir, Hamid

AU - Jamali, Sadegh

AU - Hashemi, Hossein

N1 - Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2023

Y1 - 2023

N2 - Actual evapotranspiration (ETa) plays a crucial role in the water and energy cycles of the earth. An accurate estimate of the ETa is essential for management of the water resources, agriculture, and irrigation, as well as research on atmospheric variations. Despite the importance of accurate ETa values, estimating and mapping them remains challenging due to the physical and biological complexity of the ET process. As a novel approach for rapid and reliable estimation of ETa, the present study develops automated deep learning (AutoDL) models that incorporate a metaheuristic optimization algorithm for image processing, architectural design, and hyperparameter tuning. The proposed AutoDL models integrate three different spatial and channel attention mechanisms, including a novel activated spatial attention mechanism (ASPAM), with the U-Net architecture. Bypassing the need for meteorological inputs, the proposed framework uses Moderate Resolution Imaging Spectrometer (MODIS) products and Digital Elevation Model (DEM) data as inputs. To evaluate the performance of the models, they are applied to three study areas in the United States with various climatic characteristics. According to the results, during the spring and summer, when the target values have higher certainty, the estimations are highly promising, with R2 as high as 0.91 and MAPE as low as 6.40%. Furthermore, the proposed ASPAM module improves the accuracy of ETa estimations compared to attention gate (AG) and squeeze and excitation (SE) attention modules. The results also indicate that the MODIS raw products and derived vegetation and water indices can predict ETa within a reliable range of accuracy, with the addition of DEM data marginally enhancing the models' performance. The automatic workflow of this model makes it significantly easy to use, contributing to its applicability and generalizability for enhancing atmospheric research.

AB - Actual evapotranspiration (ETa) plays a crucial role in the water and energy cycles of the earth. An accurate estimate of the ETa is essential for management of the water resources, agriculture, and irrigation, as well as research on atmospheric variations. Despite the importance of accurate ETa values, estimating and mapping them remains challenging due to the physical and biological complexity of the ET process. As a novel approach for rapid and reliable estimation of ETa, the present study develops automated deep learning (AutoDL) models that incorporate a metaheuristic optimization algorithm for image processing, architectural design, and hyperparameter tuning. The proposed AutoDL models integrate three different spatial and channel attention mechanisms, including a novel activated spatial attention mechanism (ASPAM), with the U-Net architecture. Bypassing the need for meteorological inputs, the proposed framework uses Moderate Resolution Imaging Spectrometer (MODIS) products and Digital Elevation Model (DEM) data as inputs. To evaluate the performance of the models, they are applied to three study areas in the United States with various climatic characteristics. According to the results, during the spring and summer, when the target values have higher certainty, the estimations are highly promising, with R2 as high as 0.91 and MAPE as low as 6.40%. Furthermore, the proposed ASPAM module improves the accuracy of ETa estimations compared to attention gate (AG) and squeeze and excitation (SE) attention modules. The results also indicate that the MODIS raw products and derived vegetation and water indices can predict ETa within a reliable range of accuracy, with the addition of DEM data marginally enhancing the models' performance. The automatic workflow of this model makes it significantly easy to use, contributing to its applicability and generalizability for enhancing atmospheric research.

KW - Actual evapotranspiration mapping

KW - Attention mechanism

KW - Deep learning

KW - Harris Hawks Optimization

KW - Simplified surface energy balance (SSEBop)

KW - U-Net architecture

UR - http://www.scopus.com/inward/record.url?scp=85177058090&partnerID=8YFLogxK

U2 - 10.1016/j.atmosres.2023.107107

DO - 10.1016/j.atmosres.2023.107107

M3 - Article

AN - SCOPUS:85177058090

SN - 0169-8095

VL - 297

JO - Atmospheric Research

JF - Atmospheric Research

M1 - 107107

ER -

Automated actual evapotranspiration estimation: Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Embargoed Document

Fingerprint

Cite this