Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 40 A review of methodol

Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 40 A review of methodological contributions of geostatistics to precision agriculture Douaik Ahmed ahmed.douaik@inra.ma Research Unit on Environment and Conservation of Natural Resources, Regional Center of Agricultural Research of Rabat, INRA, Rabat, Morocco. Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 41 Abstract Stronger competitiveness and better productivity of Moroccan agriculture could only be achieved through the improvement of current cultivation techniques by the adoption of more efficient and rational techniques. In this way, precision agriculture (PA), a site- specific crop management, is a promising way. It seeks to apply the right amount, when and where it is needed using variable rate technology. This approach adjusts each input based on the specific condition of each part of the study area. To achieve this goal, the spatial variability of the soil must be known everywhere using both soil samples and ancillary data or environmental covariates like proximal and remote sensing imagery, digital elevation model, land use types, yield monitor data, etc. Geostatistics, the use of statistical methods for studying spatial data, offers excellent tools for describing and modelling spatial variability using variogram, interpolating at unsampled locations using kriging, mapping, simulating different scenarios by incorporating uncertainty, optimizing experimental designs and sampling schemes, etc. These different contributions of geostatistics to precision agriculture will be illustrated from examples taken from worldwide published research works regarding different crop production factors at various spatial scales. Keywords: Digital mapping, Kriging, Site-specific crop management, Spatial variability, Stochastic simulation, Variogram. Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 42 Une revue des contributions méthodologiques de la géostatistique à l'agriculture de précision Résumé Une compétitivité plus forte et une meilleure productivité de l'agriculture marocaine ne peuvent être atteintes que par l'amélioration des techniques culturales actuelles grâce à l'adoption de techniques plus efficaces et plus rationnelles. De cette manière, l'agriculture de précision (AP), une gestion des cultures spécifique au site, est une voie prometteuse. Elle cherche à appliquer la bonne quantité, quand et où elle est nécessaire en utilisant la technologie à débit variable. Cette approche ajuste chaque entrée en fonction de la condition spécifique de chaque partie de la zone d'étude. Pour atteindre cet objectif, la variabilité spatiale du sol doit être connue partout en utilisant à la fois des échantillons de sol et des données auxiliaires ou des co-variables environnementales comme l'imagerie de détection proximale et de télédétection, le modèle numérique d'élévation, les types d'utilisation des terres, les données du moniteur du rendement, etc. La géostatistique, l’utilisation des méthodes statistiques pour l'étude des données spatiales, offre d'excellents outils pour décrire et modéliser la variabilité spatiale en utilisant le variogramme, interpoler à des endroits non échantillonnés en utilisant le krigeage, cartographier, simulater différents scénarios en incorporant l'incertitude, optimiser des plans expérimentaux et des plans d'échantillonnage, etc. Les différentes contributions de la géostatistique à l'agriculture de précision seront illustrées à partir d'exemples tirés de travaux de recherche publiés dans le monde entier concernant différents facteurs de production végétale à diverses échelles spatiales. Mots clés: Cartographie numérique, Krigeage, Gestion des cultures spécifiques au site, Variabilité spatiale, Simulation stochastique, Variogramme. Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 43 مراجعة للمساهمات المنهجية ل لجيوإحصاء في الزراعة الدقيقة أحمد الدويك خص مل ال يمكن تحقيق قدرة تنافسية أقوى وإنتاجية أفضل للزراعة المغربية إال من خالل تحسين تقنيات الزراعة الحالية من خالل اعتماد تقنيات أكثر كفاءة وعقالنية. بهذه،الطريقة تعتبر الزراعة الدقيقة إدارة محاصيل خاصة،بالموقع .طريقة واعدة إنها ت سعى إلى تطبيق الكمية ،المناسبة متى وأين تكون هناك حاجة إليه ا باستخدام تقنية معدل متغير. يقوم هذا األسلوب بضبط كل إدخال بناءً على الحالة المحددة لكل جزء من منطقة الدراسة. لتحقيق هذا،الهدف يجب أن يكون التباين المكاني للتربة معروفًا في كل مكان باستخدام عينات التربة والبيانات الم ساعدة أو المتغيرات البيئية مثل صور االستشعار عن بعد،والداني ونموذج االرتفاع،الرقمي وأنواع استخدام،األراضي وبيانات غلة المحاصيل .، وما إلى ذلك الجيوإحصاء ، استخدام األساليب اإلحصائية لدراسة البيانات المكانية ، يوفر أدوات ممتازة لوصف ونمذجة التباين المكاني باستخدام الف ا ريوغرام ، و االستنباط في المواقع غير المعينة ب استخدام الكريجاج ورسم الخرائط ، ومحاكاة .السيناريوهات المختلفة من خالل دمج عدم اليقين ، وتحسين التصاميم التجريبية وخطط أخذ العينات ، إلخ سيتم توضيح المساهمات المختلفة ال لجيوإحصاء في الزراعة الدقيقة من األمثلة المأخوذة من األعمال البحثية المنشورة في جميع أنحاء العالم بشأن عوامل إنتاج المحاصيل المختلفة على مقاييس مكانية مختلفة. :الكلمات المفتاحية ، رسم الخرائط الرقمية الكريجاج ، إدارة المحاصيل الخاصة بالموقع ، التباين ، المكاني،المحاكاة العشوائية الف ا ريوغرام. Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 44 Introduction Since the appearance of the modern human being, people were nomadic and lived by collecting wild plants, fishing, and hunting. It is circa 12.000 BC that people became sedentary and experienced farming by domesticating plants and animals (Herrera and Garcia-Bertrand, 2018). This constitutes the first, also called Neolithic, Agricultural Revolution. The Industrial Revolution contributed to the second, or British, Agricultural Revolution that occurred between the mid-17th and late 19th centuries (Thompson, 1968) and was characterized by a huge increase in agricultural production due to its mechanization. The third, or Green, Agricultural Revolution took place in 1960-1970 (Pingali, 2012). It allowed combating hunger by using agrochemicals, biotechnologies, expansion of irrigation infrastructure, hybridized seeds, etc. At present, we are living the fourth, or Digital, Agricultural Revolution that is based on data and information technologies (Klerkx et al, 2019; Saiz-Rubio and Rovira-Mas, 2020). Although there are subtle differences between them, digital agriculture is interchangeably called smart farming or precision agriculture (Klerkx et al, 2019). Smart farming is “the application of information and data technologies for optimizing complex farming systems, the focus is on access to data and the application of these data (how the collected information can be used in a smart way)” (AgroCares website). In fact, “Digital agriculture means to go beyond the mere presence and availability of data and create actionable intelligence and meaningful added value from such data; it integrates both precision agriculture and smart farming” (AgroCares website). Precision agriculture (PA) was defined as a site-specific management that recognizes the variability within a field and involves doing the right thing, in the right way, at the right place, and at the right time (Berry, 1998); this is known as the four Rs approach. Very recently, the International Society of Precision Agriculture (ISPA) defined PA, in 2019, in two formats. The short definition is "PA is a management strategy that takes account of temporal and spatial variability to improve sustainability of agricultural production". PA appeared in the early 1990s (McBratney et al, 2005; Mulla and Khosla, 2016). PA is a very important issue in modern agriculture with the aim of improving crop and land productivity at reasonable cost and protecting the environment (Auernhammer, 2001) and this is confirmed by the devoted structure like the International Society of PA (ISPA) with its continental representations and a series of conferences as well as a specifically dedicated journal (Precision Agriculture published by Springer). All precedent definitions of PA highlight the great importance of spatial and temporal variability or heterogeneity for PA. The temporal component is very important and spatio-temporal data can be evaluated for limited temporal measurement situations using statistical methods for temporal stability or persistence like Spearman rank correlation and relative differences (Vachaud et al, 1985; Douaik, 2005; Douaik et al, 2006, 2007, 2011; Iraqui et al, 2021: this issue) and for large temporal measurement scenarios using space-time statistical methods like geostatistics and Bayesian maximum entropy (Kyriakidis and Journel, 1999; Douaik, 2005; Douaik et al, 2004, 2005, 2008, 2011). However, in the present review, the discussion will be limited only to the spatial component. Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 45 There exists a large number of spatial interpolation methods (Robinson and Metternicht, 2006; Li and Heap, 2014); however, the most frequently used methods are inverse distance weighted, spline, and kriging. Although, all spatial interpolation methods predict a value at any unsampled location as a linear combination of weighted values of neighboring locations, they differ in the way the weights are computed. However, kriging, the geostatistical method of spatial interpolation, offers much more possibilities than the remaining methods since it is based on a specific tool, the variogram. These two geostatistical tools allow to consider spatial variability including configuration geometry and redundancy, take into account eventually the differences in spatial variability for different directions (anisotropy) (Gotway and Hergert, 1997), fuse data with different information supports by using block kriging and area-to-point and area-to-area methods (Sciarretta and Trematerra, 2014), compute a measure of the reliability of interpolated values (Mueller et al, 2004), and include auxiliary information (Gotway and Hartford, 1996). For all the above arguments, the focus in this research work will be exclusively on geostatistics. Geostatistics was applied, in general, to different scientific areas like remote sensing (Van Der Meer, 2012), earth sciences (Sarma, 2009), GIS (Burrough, 2001), etc. For the particular case of agricultural sciences, geostatistics was applied to soil science in the beginning of the 1980s in a series uploads/Societe et culture/ 1-pb 4 .pdf

Documents similaires

Douaik A. (2021). AFRIMED AJ –Al Awamia (132). p. 40-64 40 A review of methodol 0 0

• Détails
• Publié le Dec 24, 2021
• Catégorie Society and Cultur...
• Langue French
• Taille du fichier 0.8225MB