RECREO: Resource-Efficient and Climate-change REsilient hazelnut Orchard
23.04.2024Grant
Call for proposalRegione Lazio, POR-FESR Lazio 2021-2027, Public notice "Riposizionamento Competitivo RSI", Ambito 2 "Economia del Mare, Green Economy e Agrifood", approved by executive determination N. G18823 del 28/12/2022, published in BUR Lazio N.108 del 29/12/2022.
Grant
403.278,88 eu.
Project accepted for funding with Det. n.G14867 del 09/11/2023, published in BUR Lazio n.93 - Supplemento n.1 del 21/11/2023.
18 months
Partners
- Sigma Consulting Srl (capofila)
- Terrasystem Srl
- Università degli Studi della TUSCIA - Dipartimento di Scienze Agrarie e Forestali (DAFNE)
- Università degli Studi ROMA TRE - Dipartimento di Ingegneria Civile, Informatica e delle Tecnologie Aeronautiche (ICITA)
Description
The RECREO project pursues the goal of developing a decision support system (DSS) for some agronomic aspects of the hazel grove, such as:
1. Rational management of the irrigation supply in the hazel grove, based on the actual needs of the crop in relation to the phenological stage and expected seasonal production
2. Development of "early detection" systems of the main nutritional deficiencies, for a sustainable management of hazel grove nutrition functional to the definition of site-cultivar specific hazelnut fertilization protocols and to the significant reduction of nutrient drift from the agroecosystem hazel grove
These objectives will be pursued by integrating the information collected by IoT instrumentation installed in situ (weather stations, soil humidity sensors, AgroCam for the observation of phenology, Sap Flow sensors for the measurement of sap flows, integrated with data observed remotely via multispectral optical satellite sensors (Sentinel 2 and Landsat 8), radiances emitted in the thermal infrared (Landsat 8 and MODIS) and SAR data (Synthetic Aperture Radar from Sentinel1). Furthermore, data will also be acquired at key moments in the hazelnut phenology remotely sensed, remote and proximal, for example through sensors carried by drones (Unmanned Aerial Vehicle - UAV) with very high spatial and radiometric resolution.
The data will be processed in order to obtain an adequate representation of the state variables of the Soil Vegetation ATmosphere (SVAT) system such as water content, vigor, nutritional status of the crop and real evapotranspiration, as well as the tracking of irrigation interventions, even remotely. In particular, satellite data will be processed taking into account radiative transfer models, energy balance simulation and numerical inversion methods, including those based on artificial intelligence, in order to obtain estimates of relevant crop state variables, such as the water content and chlorophyll content at leaf and canopy level, the leaf area index LAI, the actual evapotranspiration (ETa).
All this information will be collected, archived and organized in a database returning an exhaustive description of the state of the system, which will be evaluated by the DSS models in order to provide prescriptions on the appropriate irrigation and fertilization interventions to deal with any nutritional deficiencies . These prescription models provide, in turn, further and new methods of machine learning, statistical data processing and signal processing on images, videos and field data. In particular:
- for the definition of the forecast model for the irrigation suggestion, the use of methodologies ranging from systems theory for the creation of a "model-based" representation to machine learning for the creation of a "model-based" representation will be evaluated. free;
- for the definition of the "early detection" system of the main nutritional deficiencies of the hazelnut, solutions based on Machine Learning techniques will be analysed.