Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data

Creating accurate digital maps of the agrochemical properties of soils on a field scale with a limited data set is a problem that slows down the introduction of precision farming. The use of machine learning methods based on the use of direct and indirect predictors of spatial changes in the agroche...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ilnas Sahabiev, Elena Smirnova, Kamil Giniyatullin
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	precision agriculture digital maps machine learning methods remote sensing Agriculture S
Acceso en línea:	https://doaj.org/article/d24790a9e76a4a05b5f6f30e817209e3
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:d24790a9e76a4a05b5f6f30e817209e3
record_format	dspace
spelling	oai:doaj.org-article:d24790a9e76a4a05b5f6f30e817209e32021-11-25T16:09:26ZSpatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data10.3390/agronomy111122662073-4395https://doaj.org/article/d24790a9e76a4a05b5f6f30e817209e32021-11-01T00:00:00Zhttps://www.mdpi.com/2073-4395/11/11/2266https://doaj.org/toc/2073-4395Creating accurate digital maps of the agrochemical properties of soils on a field scale with a limited data set is a problem that slows down the introduction of precision farming. The use of machine learning methods based on the use of direct and indirect predictors of spatial changes in the agrochemical properties of soils is promising. Spectral indicators of open soil based on remote sensing data, as well as soil properties, were used to create digital maps of available forms of nitrogen, phosphorus, and potassium. It was shown that machine learning methods based on support vectors (SVMr) and random forest (RF) using spectral reflectance data are similarly accurate at spatial prediction. An acceptable prediction was obtained for available nitrogen and available potassium; the variability of available phosphorus was modeled less accurately. The coefficient of determination (R<sup>2</sup>) of the best model for nitrogen is R<sup>2</sup><sub>SVMr</sub> = 0.90 (Landsat 8 OLI) and R<sup>2</sup><sub>SVMr</sub> = 0.79 (Sentinel 2), for potassium—R<sup>2</sup><sub>SVMr</sub> = 0.82 (Landsat 8 OLI) and R<sup>2</sup><sub>SVMr</sub> = 0.77 (Sentinel 2), for phosphorus—R<sup>2</sup><sub>SVMr</sub> = 0.68 (Landsat 8 OLI), R<sup>2</sup><sub>SVMr</sub> = 0.64 (Sentinel 2). The models based on remote sensing data were refined when soil organic matter (SOC) and fractions of texture (Silt, Clay) were included as predictors. The SVMr models were the most accurate. For Landsat 8 OLI, the SVMr model has a R<sup>2</sup> value: nitrogen—R<sup>2</sup> = 0.95, potassium—R<sup>2</sup> = 0.89 and phosphorus—R<sup>2</sup> = 0.65. Based on Sentinel 2, nitrogen—R<sup>2</sup> = 0.92, potassium—R<sup>2</sup> = 0.88, phosphorus—R<sup>2</sup> = 0.72. The spatial prediction of nitrogen content is influenced by SOC, potassium—by SOC and texture, phosphorus—by texture. The validation of the final models was carried out on an independent sample on soils from a chernozem zone. For nitrogen based on Landsat 8 OLI R<sup>2</sup> = 0.88, for potassium R<sup>2</sup> = 0.65, and for phosphorus R<sup>2</sup> = 0.31. Based on Sentinel 2, for nitrogen R<sup>2</sup> = 0.85, for potassium R<sup>2</sup> = 0.62, and for phosphorus R<sup>2</sup> = 0.71. The inclusion of SOC and texture in remote sensing-based machine learning models makes it possible to improve the spatial prediction of nitrogen, phosphorus and potassium availability of soils in chernozem zones and can potentially be widely used to create digital agrochemical maps on the scale of a single field.Ilnas SahabievElena SmirnovaKamil GiniyatullinMDPI AGarticleprecision agriculturedigital mapsmachine learning methodsremote sensingAgricultureSENAgronomy, Vol 11, Iss 2266, p 2266 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	precision agriculture digital maps machine learning methods remote sensing Agriculture S
spellingShingle	precision agriculture digital maps machine learning methods remote sensing Agriculture S Ilnas Sahabiev Elena Smirnova Kamil Giniyatullin Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
description	Creating accurate digital maps of the agrochemical properties of soils on a field scale with a limited data set is a problem that slows down the introduction of precision farming. The use of machine learning methods based on the use of direct and indirect predictors of spatial changes in the agrochemical properties of soils is promising. Spectral indicators of open soil based on remote sensing data, as well as soil properties, were used to create digital maps of available forms of nitrogen, phosphorus, and potassium. It was shown that machine learning methods based on support vectors (SVMr) and random forest (RF) using spectral reflectance data are similarly accurate at spatial prediction. An acceptable prediction was obtained for available nitrogen and available potassium; the variability of available phosphorus was modeled less accurately. The coefficient of determination (R<sup>2</sup>) of the best model for nitrogen is R<sup>2</sup><sub>SVMr</sub> = 0.90 (Landsat 8 OLI) and R<sup>2</sup><sub>SVMr</sub> = 0.79 (Sentinel 2), for potassium—R<sup>2</sup><sub>SVMr</sub> = 0.82 (Landsat 8 OLI) and R<sup>2</sup><sub>SVMr</sub> = 0.77 (Sentinel 2), for phosphorus—R<sup>2</sup><sub>SVMr</sub> = 0.68 (Landsat 8 OLI), R<sup>2</sup><sub>SVMr</sub> = 0.64 (Sentinel 2). The models based on remote sensing data were refined when soil organic matter (SOC) and fractions of texture (Silt, Clay) were included as predictors. The SVMr models were the most accurate. For Landsat 8 OLI, the SVMr model has a R<sup>2</sup> value: nitrogen—R<sup>2</sup> = 0.95, potassium—R<sup>2</sup> = 0.89 and phosphorus—R<sup>2</sup> = 0.65. Based on Sentinel 2, nitrogen—R<sup>2</sup> = 0.92, potassium—R<sup>2</sup> = 0.88, phosphorus—R<sup>2</sup> = 0.72. The spatial prediction of nitrogen content is influenced by SOC, potassium—by SOC and texture, phosphorus—by texture. The validation of the final models was carried out on an independent sample on soils from a chernozem zone. For nitrogen based on Landsat 8 OLI R<sup>2</sup> = 0.88, for potassium R<sup>2</sup> = 0.65, and for phosphorus R<sup>2</sup> = 0.31. Based on Sentinel 2, for nitrogen R<sup>2</sup> = 0.85, for potassium R<sup>2</sup> = 0.62, and for phosphorus R<sup>2</sup> = 0.71. The inclusion of SOC and texture in remote sensing-based machine learning models makes it possible to improve the spatial prediction of nitrogen, phosphorus and potassium availability of soils in chernozem zones and can potentially be widely used to create digital agrochemical maps on the scale of a single field.
format	article
author	Ilnas Sahabiev Elena Smirnova Kamil Giniyatullin
author_facet	Ilnas Sahabiev Elena Smirnova Kamil Giniyatullin
author_sort	Ilnas Sahabiev
title	Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
title_short	Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
title_full	Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
title_fullStr	Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
title_full_unstemmed	Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data
title_sort	spatial prediction of agrochemical properties on the scale of a single field using machine learning methods based on remote sensing data
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/d24790a9e76a4a05b5f6f30e817209e3
work_keys_str_mv	AT ilnassahabiev spatialpredictionofagrochemicalpropertiesonthescaleofasinglefieldusingmachinelearningmethodsbasedonremotesensingdata AT elenasmirnova spatialpredictionofagrochemicalpropertiesonthescaleofasinglefieldusingmachinelearningmethodsbasedonremotesensingdata AT kamilginiyatullin spatialpredictionofagrochemicalpropertiesonthescaleofasinglefieldusingmachinelearningmethodsbasedonremotesensingdata
_version_	1718413343795970048

Spatial Prediction of Agrochemical Properties on the Scale of a Single Field Using Machine Learning Methods Based on Remote Sensing Data

Ejemplares similares