The project aims to address key challenges in rural areas by creating a platform and a database for monitoring and analyzing agro-forestry systems, leveraging digital and remote sensing technologies. Key points include: rural transformation and regeneration, EU policies adaptation, digital agriculture, Earth Observation programs exploitation, technological integration, Socio-economic benefit for farmers, environmental awareness.
This kind of system must be operable in different areas and with different technologies. For this reason, experiments in different fields of application will be carried out and will be managed by five Operative Unites (OUs).
The project is organized in different Work Packages (WPs):
- WP1 – Coordination of the project
- WP2 – Geospatial database designing and management
- WP3 – Local application
- WP4 – Validation and quality control
After creating the database and defining usage policies, within the WP3 – local application, each OU will focus on specific research topics:
OU1 – VT
University of Tuscia
Remote sensing application to hyperspectral and SAR imagery
WP1 – Coordination of the project
Research
OU1 will carry out the classification of remote sensing imagery from the Prisma hyperspectral satellite and from Sentinel 1 SAR in C band and Sentinel 2 satellites of the Copernicus program together with LiDAR data.The research areas are: the Presidential Reserve of Castel Porziano, the reserve of Lake Vico and the area of Campania improperly called “the land of fires”. For the experimentation in the area of Castel Porziano, the application in the forestry field will be tested in cooperation with the staff on duty. In this case, the use of LiDAR app will also be tested to measure the forestry basal area; a quick technique will also be used to quantify the percentage of the canopy coverage (15). Five contributors and 500 observations are expected. For the experimentation on Lake Vico both reserve rangers and trained volunteers will be employed, 6 in total. It is planned to acquire 200 GT per year, to be compared with the simultaneous images acquired from the Prisma hyperspectral satellite. For the experimental area in Campania, it is planned to use volunteer personnel (n 5) and Federico II researchers and scholars (n 4), with the aim of acquiring 200 Ground Truth points per year, to be compared with the simultaneous images acquired from the Prisma hyperspectral satellite.
Expected results
- Assessment of the quality of the data collected; this assessment will be applied both to on field samplings and to the results of images classification.
- Development of a Neural Network approach to classify hyperspectral images and to assess the stress of tree vegetation in the forests.
OU2 – UD
University of Udine
Small landscape features mapping
Research
OU2-UD research unit will develop an experiment to support survey and monitoring of the small-scale features of landscape like hedgerows, brooks, lines of trees, terraces, dry stone walls, ponds. Non-cropped habitats are primarily made up of trees and grassy margins, sometime water, their amount, quality and spatial configuration can have strong implications for the delivery of various ecosystem services (7), but are also central for farm design (9) and management, as for example the application of “green direct payment” (or “greening”) required by the EU Agricultural policy. The objective of the case study is to design and carry out the survey and classification of these landscape elements in the surrounding areas. New models for collecting reliable spatial information related to ground truths will be tested using a distributed approach based on participatory geospatial systems powered by a combination of remotely sensed and field-crowd sensed data. In particular the activities will be included in Living Lab experiments and will investigate selected rural and peri-urban study areas within the plane of the Friuli Venezia Giulia region (6 representative sites along the rural-urban gradient). The relevant stakeholders (local administrations, public and private actors involved in management of rural areas, farmers and professional associations) will be involved since the early stages of the study. This will promote the active participation of the surveyors and their awareness on the potential advantages in the use of the data.
Expected results
- A detailed map of small landscape features useful for landscape characterization.
- The assessment of the impact of these small elements on deliverable ecosystem services and on farm management.
OU3 – RC
Mediterranea University of Reggio Calabria
Remote sensing application for dry stone artefacts and LULC maps
Research
The case study here analysed is located in the north side of Etna volcano, an area of relevant naturalistic value. Since 1987 it has belonged to the territory of the Etna Park (Etna Park) and since 2013 it has been part of the UNESCO World Heritage List Mount Etna. This landscape shows relevant forest, agricultural and rural architectural features. The dry-stone rural architecture has been inscribed in the list of Intangible Cultural Heritage of Humanity (Art of dry stone walling, knowledge and techniques of UNESCO). In the Etna area there are numerous dry stone buildings that are landmarks to that landscape: the walls of the terraces and locks; the “turrets” stepped pyramids; rural houses and pastoral shelters so called “a thòlos” (hundreds in this area).
In this area, just like in other Italian areas, there is a lack of information about the amount and location of this cultural heritage, about its specific characteristics and its state of conservation. This depends on many factors: it is generally scattered and located in inaccessible areas, the remote surveying poses problems in identifying the buildings (they are small, the covering stone is mistaken for the piles of stones or with the stony surface of the site). The objective of the case study is to design, detect and classify these artefacts and their surrounding area by using integrated and advanced techniques for remote sensing and field surveys. The former will be based on images obtained both from UAVs (using multispectral and thermal sensors) and from satellites (using multispectral, hyperspectral and RADAR sensors) with high and very high temporal, radiometric and geometric resolution. The classification techniques are object-oriented and / or pixel-based, using Machine Learning (ML) or Deep Learning (DL) (14, 22). The latter will be carried out using the spatial infrastructure developed in the project. They will be conducted with a participatory approach, the direct interlocutors will be involved, and so excursion groups, Etna Park Authority, and tourist-cultural associations will. The surveying can be implemented with simple devices (Smartphone and Pad) and in compliance with a protocol predefined by the OU; the involvement of different participants is important both to collect GT related to satellite images and to identify an adequate number of samples. They will be arranged by type, thus allowing to perform postdictive analyses to classify the objects identified by the remote sensing. Furthermore, for the geometric irregularities and curvilinear shapes typical of the pastoral shelters in tholos, the use of advanced terrestrial and aerial architectural survey techniques based on 3D laser scanners and UAVs will be required.
Expected results
- Survey campaign in a large sampling area to identify, geolocate, characterize, classify the rural dry-stone cultural heritage and the territory which hosts it.
- LULC classification for two historical periods in order to analyse the transformations of the landscape.
- Implementation of procedures based on Machine Learning and Deep Learning to classify the cultural heritage and landscape.
- In addition to the LULC classification, the metrics of landscape composition and configuration will be calculated in order to understand the relevant dynamics.
OU4 – BO
University of Bologna
Rural structures and heritage
Research
OU4-BO will test a system aimed at supporting an updated and cost-effective monitoring of rural structures and heritage, as a crucial building block to enable efficient planning and management of rural areas. Lack of resources and cross-sector and cross-governmental issues make it difficult to achieve frequently updated, detailed and reliable geodatabases with extensive coverage only by means of top-down or volunteered geodatabase creation. The activities will test new models of systematic widespread generation of reliable spatial information, harvesting data about rural structures and heritage (features, conservation status, technical and energetic efficiency, current use) using a distributed approach powered by participatory geospatial systems fed through a combination of remotely sensed and on-field crowd-sensed data (1, 20, 22). Trials will be carried out in Living Lab experiments over rural and periurban study-areas selected within the Metropolitan Territory of Bologna (2 pilots per year, 5 rural sites representative of the size, type and location of farms and buildings in each pilot). The relevant stakeholders (local administrations, public and private actors involved in rural areas management, farmers and professional associations) will be involved and will collaborate since the early stage of the project when the specific data to be collected and the contributors’ engagement plan will be defined. A win-win perspective assuring sustainability of the proposed approach also beyond the project duration will be pursued. In each pilot an actor from the academic-public-private triangle will participate in the contributors’ team.
Expected results
- The expected outcome is a tested platform providing a boost in GTs acquisition in rural structures and heritage, allowing public and private institutions and farmers to plan effective maintenance, restoration, and implementation actions. The platform will support public institutions I) in assessing new construction needs, based on the existing building stock; ii) detecting and addressing the evolving needs of farmers and newcomers in terms of structures and infrastructure; iii) monitoring the impact of rural planning; iv) deriving quantitative information aimed at defining design criteria. At the same time, farmers will access more easily targeted and contextual information about planning, building and environmental constraints and opportunities, and the potential to reuse and regenerate damaged or abandoned buildings.
OU5 – AN
Polytechnic University of Marche
Landscape quality
Research
The OU5-AN (as well as the OU2-UD) will focus on the survey and monitoring of the small scale features of landscape (uncultivated, vegetation of permanent ditches and farm roads, permanent crops -olive groves, vineyards, orchards-, plants and rows of trees); these important elements are able to affect the delivery of ecosystem services, and their presence can be used to describe the environmental characteristics of the landscape. Mainly referring to olive tree growing landscape, the possibility of assigning a quality label based on landscape and farms characters will be explored; outcomes on agritourist sector will be explored too (6). The effectiveness and compliance with the commitments of EU funds will be also analysed. Local stakeholders (regional administration, municipality, etc.) committed in landscape and environmental planning will be engaged by means of workshop and Focus groups. Furthermore, in collaboration with Cluster Agri-Food (CLAM) of Marche Region and Smart Farming Research Centre of UNIVPM, a permanent “Contamination Lab” involving farmers, researchers and expert of agricultural sector will be set up. The dynamics of olive tree landscape will be analysed integrating GIS and landscape ecology models with the aim to define specific spatial indices.
Expected results
- Expected outcome is the triggering of a participatory process by means of the setting up of a permanent virtual and/or physic space (“Contamination Lab”) aimed at the sharing of experiences and ideas between farmers, experts and operators in the sector of rural areas valorisation (rural tourism, quality labelling, etc.).