東華大學圖書館 |

Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data./
作者:	Saifuzzaman, Md.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2020,
面頁冊數:	171 p.
附註:	Source: Dissertations Abstracts International, Volume: 82-10, Section: B.
Contained By:	Dissertations Abstracts International82-10B.
標題:	Maps. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28384049
ISBN:	9798708710505

Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data.
Saifuzzaman, Md.

Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data. - Ann Arbor : ProQuest Dissertations & Theses, 2020 - 171 p.

Source: Dissertations Abstracts International, Volume: 82-10, Section: B.

Thesis (Ph.D.)--McGill University (Canada), 2020.

This item must not be sold to any third party vendors.

Emerging technologies in precision agriculture (PA) offer a wide array of advanced methods to assess soil properties and to determine soil variability. Remote sensing (RS) and proximal soil sensing (PSS) technologies, widely used in quantifying surface and subsurface soil parameters, can be combined to infer spatial patterns of soil heterogeneity and to develop thematic maps for site-specific management. However, the use of these soil sensors must be reviewed constantly to maintain their efficiency and precision in delineating the soil-crop relationship and to inform PA approaches. Data mining and model optimization are key to evaluating high-density geospatial data in a dynamic production system. High-density PSS and RS-based soil characterization was explored and optimization techniques for digital soil mapping in PA were evaluated.In a first study, sensor measurements were subjected to multivariate statistical analysis, followed by an evaluation of a new Neighborhood Search Analyst (NSA) and the capacity of other data clustering algorithms to delineate spatially contiguous zones in agricultural fields and to optimize soil sampling locations to inform best management practices. PSS-based topography, apparent electrical conductivity (ECa), and RS-based indices data from 3 sites in Ontario, Canada, were employed to assess the novel technique's performance in accurate zone delineation. In creating homogeneous zones, a maximum of 70% field variance (R2 = 0.70) was achieved. The R2 of the k-means cluster compared to that of the NSA was relatively higher (R2 = 0.80) where, the k-means cluster map consisted of groups or pixels with isolated boundaries in various parts of the field. The NSA's unique capacity, across various locations, to produce an optimum (or userdefined) number of zones highlighted its superiority to k-means' partitioning with isolated boundaries.A second study assessed the utility of PSS-based soil characterization in developing an optimum prediction method for multiple soil properties at 12 sites across Ontario, Canada. Targeted soil sampling locations were determined and optimized using NSA clustering tools. Measured ECa, topographic parameters and six lab-quantified soil properties [pH, buffer pH, soil organic matter (SOM), Phosphorus (P), Potassium (K) and Cation Exchange Capacity (CEC)] were used in evaluating the method's predictive capacity and to compare different fields' propagated soil measurement errors by drawing on the results of the North American Proficiency Testing program. Pearson's correlation coefficients exceeding 0.60 indicated strong relationships between sensor variables and field-measured soil properties, topographic parameters and shallow ECa sensor variables, allowing effective predictions of several soil chemical properties (i.e., SOM, P, and CEC).Lastly, supervised machine learning models drawing on high-density information from multiple sensors (PSS and RS) operating at different geospatial scales, were used to generate thematic soil maps for an agricultural field in Ontario, Canada. A random forest (RF) regression model delineated the complex hierarchical relationships existing among the sensor variables and evaluated prediction efficiencies for multiple soil nutrients. The reduction of variables based on their relative importance and parameter optimization (i.e., by defining the number of trees) of the regression forest improved the predictive accuracy for nine soil properties at the cross-validation stage. The best prediction capacity has been achieved for soil pH, K, and Zn (R2 ≥ 0.80).Sophisticated technologies are critical to generating finer resolution thematic maps for PA and to address soil management at various geospatial scales. Multilayer data optimization techniques used in multiple sensor-based mapping provide information of field-scale variability and soil prediction at the local-scale. This research indicated that soil variability which was determined using sensor-fused data and optimization techniques could assist in constructing precise soil property prediction models and in developing reliable thematic maps for site-specific crop management.

ISBN: 9798708710505Subjects--Topical Terms:

544078
Maps.
Subjects--Index Terms:

Geospatial data

Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data.
LDR:10260nmm a2200385 4500 001 2280673
005 20210907071128.5
008 220723s2020 ||||||||||||||||| ||eng d
020 $a 9798708710505
035 $a (MiAaPQ)AAI28384049
035 $a (MiAaPQ)McGill_05741x41m
035 $a AAI28384049
040 $a MiAaPQ $c MiAaPQ
100 1 $a Saifuzzaman, Md. $3 3559213
245 1 0 $a Optimization of Geospatial Data Modelling for Crop Production by Integrating Proximal Soil Sensing and Remote Sensing Data.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2020
300 $a 171 p.
500 $a Source: Dissertations Abstracts International, Volume: 82-10, Section: B.
500 $a Advisor: Adamchuk, Viacheslav.
502 $a Thesis (Ph.D.)--McGill University (Canada), 2020.
506 $a This item must not be sold to any third party vendors.
520 $a Emerging technologies in precision agriculture (PA) offer a wide array of advanced methods to assess soil properties and to determine soil variability. Remote sensing (RS) and proximal soil sensing (PSS) technologies, widely used in quantifying surface and subsurface soil parameters, can be combined to infer spatial patterns of soil heterogeneity and to develop thematic maps for site-specific management. However, the use of these soil sensors must be reviewed constantly to maintain their efficiency and precision in delineating the soil-crop relationship and to inform PA approaches. Data mining and model optimization are key to evaluating high-density geospatial data in a dynamic production system. High-density PSS and RS-based soil characterization was explored and optimization techniques for digital soil mapping in PA were evaluated.In a first study, sensor measurements were subjected to multivariate statistical analysis, followed by an evaluation of a new Neighborhood Search Analyst (NSA) and the capacity of other data clustering algorithms to delineate spatially contiguous zones in agricultural fields and to optimize soil sampling locations to inform best management practices. PSS-based topography, apparent electrical conductivity (ECa), and RS-based indices data from 3 sites in Ontario, Canada, were employed to assess the novel technique's performance in accurate zone delineation. In creating homogeneous zones, a maximum of 70% field variance (R2 = 0.70) was achieved. The R2 of the k-means cluster compared to that of the NSA was relatively higher (R2 = 0.80) where, the k-means cluster map consisted of groups or pixels with isolated boundaries in various parts of the field. The NSA's unique capacity, across various locations, to produce an optimum (or userdefined) number of zones highlighted its superiority to k-means' partitioning with isolated boundaries.A second study assessed the utility of PSS-based soil characterization in developing an optimum prediction method for multiple soil properties at 12 sites across Ontario, Canada. Targeted soil sampling locations were determined and optimized using NSA clustering tools. Measured ECa, topographic parameters and six lab-quantified soil properties [pH, buffer pH, soil organic matter (SOM), Phosphorus (P), Potassium (K) and Cation Exchange Capacity (CEC)] were used in evaluating the method's predictive capacity and to compare different fields' propagated soil measurement errors by drawing on the results of the North American Proficiency Testing program. Pearson's correlation coefficients exceeding 0.60 indicated strong relationships between sensor variables and field-measured soil properties, topographic parameters and shallow ECa sensor variables, allowing effective predictions of several soil chemical properties (i.e., SOM, P, and CEC).Lastly, supervised machine learning models drawing on high-density information from multiple sensors (PSS and RS) operating at different geospatial scales, were used to generate thematic soil maps for an agricultural field in Ontario, Canada. A random forest (RF) regression model delineated the complex hierarchical relationships existing among the sensor variables and evaluated prediction efficiencies for multiple soil nutrients. The reduction of variables based on their relative importance and parameter optimization (i.e., by defining the number of trees) of the regression forest improved the predictive accuracy for nine soil properties at the cross-validation stage. The best prediction capacity has been achieved for soil pH, K, and Zn (R2 ≥ 0.80).Sophisticated technologies are critical to generating finer resolution thematic maps for PA and to address soil management at various geospatial scales. Multilayer data optimization techniques used in multiple sensor-based mapping provide information of field-scale variability and soil prediction at the local-scale. This research indicated that soil variability which was determined using sensor-fused data and optimization techniques could assist in constructing precise soil property prediction models and in developing reliable thematic maps for site-specific crop management.
520 $a Les technologies emergentes en agriculture de precision (AP) offrent un large eventail de methodes avancees pour evaluer les proprietes du sol et determiner leur variabilite. Les technologies de teledetection (RS) et de detection de sol proximale (PSS), largement utilisees pour quantifier les parametres pedologiques de surface et souterrains, peuvent etre combinees de maniere a deduire des modeles spatiaux d'heterogeneite des sols et pour developper des cartes thematiques pour une gestion specifique au site. Cependant, ces explorations avec capteurs de sol doivent etre revues en permanence pour maintenir leur efficacite et leur precision dans l'encadrement des relations sol-culture et des approches PA. L'exploration de donnees et l'optimisation des modeles sont essentielles a evaluation des donnees geospatiales a haute densite dans un systeme de production dynamique. La caracterisation des sols a base de PSS et RS a haute densite fut exploree et les techniques d'optimisation pour la cartographie numerique des sols en PA furent evaluees.Dans une premiere etude, des mesures des capteurs furent soumises a une analyse statistique multivariee, suivie d'une evaluation de la capacite de Neighbourhood Analyst (NSA) et d'autres algorithmes de regroupement de donnees a delimiter des zones spatialement contigues dans les champs agricoles et d'optimiser les emplacements d'echantillonnage du sol pour eclairer les meilleures pratiques de gestion. La topographie basee sur PSS, la conductivite electrique apparente (ECa) et les donnees d'indices bases sur RS de 3 sites en Ontario, au Canada, ont permis une evaluation des performances de la nouvelle technique dans la delimitation precise des zones. Le R2 du groupe de k-moyennes par rapport a celui de la NSA etait relativement plus eleve (R2 = 0,80) ou, la carte du groupe de k-moyennes consistait en groupes ou pixels avec des limites isolees dans diverses parties du champ. La capacite unique des NSA, sur divers sites, a produire un nombre optimal (ou defini par l'utilisateur) de zones, a mis en evidence sa superiorite sur le partitionnement par k-means avec des limites isolees.Une seconde etude evalua l'utilite de la caracterisation des sols basee sur PSS dans le developpement d'une methode de prediction optimale pour plusieurs proprietes des sols, pour 12 sites a travers l'Ontario, Canada. Les emplacements d'echantillonnage des sols cibles furent determines et optimises a l'aide d'outils de regroupement NSA. L'ECa mesuree, les parametres topographiques et six proprietes du sol quantifiees en laboratoire (pH, pH tampon, SOM, P, K et CEC) servirent a evaluer la capacite predictive de la methode et a comparer l'erreur de mesure du sol propagee de differents champs en s'appuyant sur les resultats du Programme North American Proficiency Testing. Les coefficients de correlation de Pearson superieurs a 0,60 indiquaient de fortes relations entre les variables du capteur et les proprietes du sol mesurees sur le terrain, les parametres topographiques et les variables du capteur ECa peu profondes, permettant des predictions efficaces de plusieurs proprietes chimiques du sol (c.-a-d. SOM, P et CEC).Enfin, des modeles d'apprentissage automatique supervise s'appuyant sur des informations a haute densite provenant de plusieurs capteurs (PSS et RS) fonctionnant a differentes echelles geospatiales ont servi a generer des cartes thematiques des sols pour un champ agricole en Ontario, au Canada. Un modele de regression aleatoire en foret (RF) a delimite les relations hierarchiques complexes existant entre les variables du capteur et evalue l'efficacite de la prediction pour plusieurs nutriments du sol. L'importance reduction en fonction de leur relative variable et l'optimisation des parametres (c'est-a-dire en definissant le nombre d'arbres) de regression ont ameliore la precision predictive pour neuf proprietes du sol au stade de la validation croisee. Le coefficient de determination (R2) a montre que la plus grande precision (ajustement du modele) a ete atteinte pour la prediction du pH, du K et du Zn (R2 ≥ 0.80).Des technologies sophistiquees sont essentielles a la generation de cartes thematiques a resolution plus fine pour l'AP et a la gestion des sols a differentes echelles geospatiales. Les techniques d'optimisation des donnees multicouches utilisees dans la cartographie basee sur plusieurs capteurs permettent de comprendre la variabilite a l'echelle du terrain et la prevision du sol a l'echelle locale. Cette recherche a indique que la variabilite du sol determinee a l'aide de donnees fusionnees par capteur et de techniques d'optimisation pourrait aider a construire des modeles precis de prevision des proprietes du sol et a developper des cartes thematiques fiables pour la gestion des cultures specifiques au site.
590 $a School code: 0781.
650 4 $a Maps. $3 544078
650 4 $a Mapping. $3 3355992
650 4 $a Aircraft accidents & safety. $3 3559214
650 4 $a Software. $2 gtt. $3 619355
650 4 $a Deep learning. $3 3554982
650 4 $a Datasets. $3 3541416
650 4 $a Remote sensing. $3 535394
650 4 $a Clustering. $3 3559215
653 $a Geospatial data
653 $a Crop production
653 $a Proximal soil sensing
653 $a Remote sensing data
690 $a 0202
690 $a 0799
690 $a 0473
710 2 $a McGill University (Canada). $3 1018122
773 0 $t Dissertations Abstracts International $g 82-10B.
790 $a 0781
791 $a Ph.D.
792 $a 2020
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28384049