Abstract Increasing the organic matter content of the soil has been presented in the:”4per1000″ proposal as a significant climate mitigation measure able to support the achievement of Sustainable Development Goal 13 - Climate Action of United Nations. At the same time, the report of the Mission Board for Soil health and Food, "Caring for soil is caring for life," indicates that one of the targets that must be reached by 2030 is the conservation and increase of soil organic carbon stock. De facto, the panel clearly indicates the soil organic carbon as one of the indicators that can be used to monitor soil health, and at the same time, if the current soil use is sustainable or not. Thus it is to be expected that the monitoring of SOC will become requested to check and monitor the sustainability of agricultural practices realized in the agricultural areas. For all the above reasons, the development of a reliable and fast indirect methods to evaluate the SOC is necessary to support different stakeholders (government, municipality, farmer) to monitor SOC at different spatial scales (national, regional, local). Over the past two decades, data mining approaches in spatial modeling of soil organic carbon using machine learning techniques and artificial neural network (ANN) to investigate the amount of carbon in the soil using remote sensing data has been widely considered. Accordingly, this study aims to design an accurate and robust neural network model to estimate the soil organic carbon using the data-based field-portable spectrometer and laboratory-based visible and near-infrared (VIS/NIR, 350−2500 nm) spectroscopy of soils. The measurements will be on two sets of the same soil samples, the first by the standard protocol of requested laboratories for soil scanning, The second set of the soil samples without any cultivation to simulate the soil condition in the sampling field emphasizes the predictive capabilities to achieve fast, cheap and accurate soil status. Carbon soil parameter will determine using, multivariate regression method used for prediction with Least absolute shrinkage and selection operator regression (Lasso) in interval way (high, medium, and low). The results will increase accuracy, precision, and cost-effectiveness over traditional ex-situ methods.

In a vineyard, plant water status variability is strongly dependent on soil spatial variability, whose physical characteristics drive the processes involved in the soil water balance. The more the soil and its characteristics vary in space (horizontally and vertically), the less homogeneous the productive and qualitative response within the vineyard will be. In this context, the proximal sensing of apparent soil Electrical Conductivity (ECa) and its monitoring during the growing season can help understanding the nature of spatial variability of vineyard, supporting both viticultural microzoning (identifying Homogeneous and functional Homogeneous Zones, HZs and fHZs) and field experiments. In this contribution, the importance of use of ECa spatial monitoring during the grapevine growing season will be showed, highlighting its importance in the evaluation of the vineyard response and identification of FHZs. In this direction, a predictive model of soil ECa is proposed. It consists of the spatial prediction of ECa starting from punctual measurements obtained by a network of TDR probes able to measure ECA. In order to realise this model, a machine learning method based on Random Forests was used. It was trained to derive the spatial relationships between the apparent value of ECa measured with geophysical instrument and the ECa measured with the ACCLIMA TDR probes. In this way, a spatial prediction of the ECa values of the surveyed area is possible. The study was realized in a vineyard of southern Italy on Greco (white) grapevine, where detailed and precise records on soil and atmosphere systems, in-vivo plant monitoring of eco-physiological parameters have been conducted in 2020 and 2021, and spatial variability of plant status in vineyard monitored by means of UAV multispectral images. Apparent soil ECa was measured five timesduring the growing season 2021 by using the PROFILER EMP 400 electromagnetometer both in vertical and horizontal dipole mode. This instrument allows to simultaneously work with three frequencies (5000, 10000 and 15000 Hz) and explore different depth volumes of sub-soil. The predictive model results were processed in MATLAB and Python software and validated on plant responses obtained from vegetational indexes calculated from UAV multispectral images. The obtained results have shown how the ECa can be estimated by the predictive model carrying out important information to support vineyard monitoring.

Abstract Currently, the main goal of agriculture is to promote the resilience of agricultural systems in a sustainable way through the improvement of use efficiency of farm resources, increasing crop yield and quality, under climate change conditions. This last is expected to drastically modify plant growth, with possible negative effects especially in arid and semi-arid regions of Europe on the viticultural sector. In this context, the monitoring of spatial behavior of grapevine during the growing season represents an opportunity to improve the plant management, winegrowers’ incomes, and to preserve the environmental health, but it has additional costs for the farmer. Nowadays UAS equipped with a VIS-NIR multispectral camera (blue, green, red, red-edge, and NIR) represents a good and relatively cheap solution to assess plant status spatial information (by means of a limited set of spectral vegetation indices), representing important support in precision agriculture management during the growing season. While differences between UAS-based multispectral imagery and point-based spectroscopy are well discussed in the literature, their impact on plant status estimation by vegetation indices is not completely investigated in depth. The aim of this study was to assess the performance level of UAS-based multispectral (5 bands across 450-800nm spectral region with a spatial resolution of 5cm) imagery, reconstructed high-resolution satellite (Sentinel-2A) multispectral imagery (13 bands across 400-2500 nm with a spatial resolution of <2 m) through Convolutional Neural Network (CNN) approach, and point-based field spectroscopy (collecting 600 wavelengths across 400-1000 nm spectral region with a surface footprint of 1-2 cm) in a plant status estimation application, and then, using Bayesian regularization artificial neural network for leaf chlorophyll content (LCC) and plant water status (LWP) prediction. The approach was realized within the Italian regional project GREASE, in an experimental vineyard of Greco of Feudi di San Gregorio winery (southern Italy), where detailed and precise records on soil and atmosphere systems, in-vivo plant monitoring of eco-physiological parameters have been conducted. Keywords: precision agriculture, vineyard monitoring, spectral measurements, CNN applied to viticulture, UAS.

Small mammals, and particularly rodents, are common inhabitants of farmlands, where they play key roles in the ecosystem, but when overabundant, they can be major pests, able to reduce crop production and farmers’ incomes, with tangible effects on the achievement of Sustainable Development Goals no 2 (SDG2, Zero Hunger) of the United Nations. Farmers do not currently have a standardized, accurate method of detecting the presence, abundance, and locations of rodents in their fields, and hence do not have environmentally efficient methods of rodent control able to promote sustainable agriculture oriented to reduce the environmental impacts of cultivation. New developments in unmanned aerial system (UAS) platforms and sensor technology facilitate cost-effective data collection through simultaneous multimodal data collection approaches at very high spatial resolutions in environmental and agricultural contexts. Object detection from remote-sensing images has been an active research topic over the last decade. With recent increases in computational resources and data availability, deep learning-based object detection methods are beginning to play an important role in advancing remote-sensing commercial and scientific applications. However, the performance of current detectors on various UAS-based datasets, including multimodal spatial and physical datasets, remains limited in terms of small object detection. In particular, the ability to quickly detect small objects from a large observed scene (at field scale) is still an open question. In this paper, we compare the efficiencies of applying one- and two-stage detector models to a single UAS-based image and a processed (via Pix4D mapper photogrammetric program) UAS-based orthophoto product to detect rodent burrows, for agriculture/environmental applications as to support farmer activities in the achievements of SDG2. Our results indicate that the use of multimodal data from low-cost UASs within a self-training YOLOv3 model can provide relatively accurate and robust detection for small objects (mAP of 0.86 and an F1-score of 93.39%), and can deliver valuable insights for field management with high spatial precision able to reduce the environmental costs of crop production in the direction of precision agriculture management.