Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion

Estimating wheat traits based on spectral reflectance measurements and machine learning remains challenging due to the large datasets required for model training and testing. To overcome this limitation, a simulated dataset was generated using the radiative transfer model (RTM) PROSAIL and inverted...

Full description

Saved in:
Bibliographic Details
Main Authors: Lukas J. Koppensteiner, Hans-Peter Kaul, Sebastian Raubitzek, Philipp Weihs, Pia Euteneuer, Jaroslav Bernas, Gerhard Moitzi, Thomas Neubauer, Agnieszka Klimek-Kopyra, Norbert Barta, Reinhard W. Neugschwandtner
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/11/1904
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850160963977740288
author Lukas J. Koppensteiner
Hans-Peter Kaul
Sebastian Raubitzek
Philipp Weihs
Pia Euteneuer
Jaroslav Bernas
Gerhard Moitzi
Thomas Neubauer
Agnieszka Klimek-Kopyra
Norbert Barta
Reinhard W. Neugschwandtner
author_facet Lukas J. Koppensteiner
Hans-Peter Kaul
Sebastian Raubitzek
Philipp Weihs
Pia Euteneuer
Jaroslav Bernas
Gerhard Moitzi
Thomas Neubauer
Agnieszka Klimek-Kopyra
Norbert Barta
Reinhard W. Neugschwandtner
author_sort Lukas J. Koppensteiner
collection DOAJ
description Estimating wheat traits based on spectral reflectance measurements and machine learning remains challenging due to the large datasets required for model training and testing. To overcome this limitation, a simulated dataset was generated using the radiative transfer model (RTM) PROSAIL and inverted based on an artificial neural network (ANN). Field experiments were conducted in Eastern Austria to measure spectral reflectance and destructively sample plants to measure the wheat traits plant area index (PAI), nitrogen yield (NY), canopy water content (CWC), and above-ground dry matter (AGDM). Four ANN-based RTM inversion models were setup, which varied in their spectral resolution, hyperspectral or multispectral, and the inclusion or exclusion of background soil spectra correction. The models were also compared to a simple vegetation index approach using Normalized Difference Vegetation Index (NDVI) and Normalized Difference Red-Edge (NDRE). The RTM inversion model with hyperspectral input data and background soil spectra correction was the best among all tested models for estimating wheat traits during the vegetative developmental stages (PAI: R<sup>2</sup> = 0.930, RRMSE = 17.9%; NY: R<sup>2</sup> = 0.908, RRMSE = 14.4%; CWC: R<sup>2</sup> = 0.967, RRMSE = 17.0%) as well as throughout the whole growing season (PAI: R<sup>2</sup> = 0.845, RRMSE = 27.7%; CWC: R<sup>2</sup> = 0.884, RRMSE = 20.0%; AGDM: R<sup>2</sup> = 0.960, RRMSE = 13.7%). Many models presented in this study provided suitable estimations of the relevant wheat traits PAI, NY, CWC, and AGDM for application in agronomy, breeding, and crop sciences in general.
format Article
id doaj-art-53e471060c834e8b80f549c77805a4f0
institution OA Journals
issn 2072-4292
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-53e471060c834e8b80f549c77805a4f02025-08-20T02:23:00ZengMDPI AGRemote Sensing2072-42922025-05-011711190410.3390/rs17111904Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model InversionLukas J. Koppensteiner0Hans-Peter Kaul1Sebastian Raubitzek2Philipp Weihs3Pia Euteneuer4Jaroslav Bernas5Gerhard Moitzi6Thomas Neubauer7Agnieszka Klimek-Kopyra8Norbert Barta9Reinhard W. Neugschwandtner10Institute of Agronomy, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Konrad-Lorenz-Straße 24, 3430 Tulln, AustriaInstitute of Agronomy, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Konrad-Lorenz-Straße 24, 3430 Tulln, AustriaInstitute of Information Systems Engineering—Data Science, University of Technology Vienna, Favoritenstraße 9-11, 1040 Vienna, AustriaInstitute of Meteorology and Climatology, Department of Ecosystem Management, Climate and Biodiversity, University of Natural Resources and Life Sciences, Vienna, Gregor-Mendel-Straße 33, 1180 Vienna, AustriaExperimental Farm Groß-Enzersdorf, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Schloßhofer Straße 31, 2301 Groß-Enzersdorf, AustriaDepartment of Agroecosystems, Faculty of Agriculture and Technology, University of South Bohemia in Ceske Budejovice, Studentska 1668, 370 05 Ceske Budejovice, Czech RepublicExperimental Farm Groß-Enzersdorf, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Schloßhofer Straße 31, 2301 Groß-Enzersdorf, AustriaInstitute of Information Systems Engineering—Data Science, University of Technology Vienna, Favoritenstraße 9-11, 1040 Vienna, AustriaDepartment of Agroecology and Plant Production, University of Agriculture in Kraków, Al. Mickiewicza 21, 31-120 Kraków, PolandInstitute of Agricultural Engineering, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Peter-Jordan-Straße 82, 1190 Vienna, AustriaInstitute of Agronomy, Department of Agricultural Sciences, University of Natural Resources and Life Sciences, Vienna, Konrad-Lorenz-Straße 24, 3430 Tulln, AustriaEstimating wheat traits based on spectral reflectance measurements and machine learning remains challenging due to the large datasets required for model training and testing. To overcome this limitation, a simulated dataset was generated using the radiative transfer model (RTM) PROSAIL and inverted based on an artificial neural network (ANN). Field experiments were conducted in Eastern Austria to measure spectral reflectance and destructively sample plants to measure the wheat traits plant area index (PAI), nitrogen yield (NY), canopy water content (CWC), and above-ground dry matter (AGDM). Four ANN-based RTM inversion models were setup, which varied in their spectral resolution, hyperspectral or multispectral, and the inclusion or exclusion of background soil spectra correction. The models were also compared to a simple vegetation index approach using Normalized Difference Vegetation Index (NDVI) and Normalized Difference Red-Edge (NDRE). The RTM inversion model with hyperspectral input data and background soil spectra correction was the best among all tested models for estimating wheat traits during the vegetative developmental stages (PAI: R<sup>2</sup> = 0.930, RRMSE = 17.9%; NY: R<sup>2</sup> = 0.908, RRMSE = 14.4%; CWC: R<sup>2</sup> = 0.967, RRMSE = 17.0%) as well as throughout the whole growing season (PAI: R<sup>2</sup> = 0.845, RRMSE = 27.7%; CWC: R<sup>2</sup> = 0.884, RRMSE = 20.0%; AGDM: R<sup>2</sup> = 0.960, RRMSE = 13.7%). Many models presented in this study provided suitable estimations of the relevant wheat traits PAI, NY, CWC, and AGDM for application in agronomy, breeding, and crop sciences in general.https://www.mdpi.com/2072-4292/17/11/1904phenotypingmachine learningsimulation learningvegetation indicesbackground soil spectra
spellingShingle Lukas J. Koppensteiner
Hans-Peter Kaul
Sebastian Raubitzek
Philipp Weihs
Pia Euteneuer
Jaroslav Bernas
Gerhard Moitzi
Thomas Neubauer
Agnieszka Klimek-Kopyra
Norbert Barta
Reinhard W. Neugschwandtner
Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
Remote Sensing
phenotyping
machine learning
simulation learning
vegetation indices
background soil spectra
title Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
title_full Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
title_fullStr Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
title_full_unstemmed Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
title_short Estimating Wheat Traits Using Artificial Neural Network-Based Radiative Transfer Model Inversion
title_sort estimating wheat traits using artificial neural network based radiative transfer model inversion
topic phenotyping
machine learning
simulation learning
vegetation indices
background soil spectra
url https://www.mdpi.com/2072-4292/17/11/1904
work_keys_str_mv AT lukasjkoppensteiner estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT hanspeterkaul estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT sebastianraubitzek estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT philippweihs estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT piaeuteneuer estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT jaroslavbernas estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT gerhardmoitzi estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT thomasneubauer estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT agnieszkaklimekkopyra estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT norbertbarta estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion
AT reinhardwneugschwandtner estimatingwheattraitsusingartificialneuralnetworkbasedradiativetransfermodelinversion