#	Pagina
attuale pagina	/open-h2020/projects/212427/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - FOCUS (Forest Operational monitoring using Copernicus and UAV hyperSpectral data)

Teaser

Summary

FOCUS is a H2020 project aimed at demonstrating an innovative extension of an existing forest monitoring service (www.silvisense.com) using a combination of Copernicus Sentinel-2 imagery with hyperspectral data captured by UAV platforms. The new product line will focus on providing its customers with earlier detection of forest disease outbreaks, threatening Portuguese and European forests.
To this end, new methodologies and tools are being developed to detect trees infected by Pinewood Nematode (Bursaphelenchus xylophilus) (PWN), an important Quarantine Organism.
Widespread decline of Maritime Pine stands in Portugal offers an ideal case for the development, testing, and implementation of integrated information solutions tailored to meet the needs signaled by stakeholders.
The challenges of detecting subtle signatures in heterogeneous landscapes is being addressed through the use of multi-mission data and the collection of important field data for calibration and validation purposes.
FOCUS is employing Copernicus (Sentinel-2) multi-spectral data for wide-area and high-frequency monitoring. Tools running on Sentinel-2 data are being developed using a dense dataset of field and laboratory analysis including for instance spectral characterization or pigment and water content evaluation.
Multi- and Hyperspectral data are being collected from Remotely Piloted Aerial Systems (RPAS) to demonstrate the feasibility of detecting declining trees using these platforms in operational contexts.
The first year of activity resulted in the collection of substantial datasets and the development of early algorithms, which will be demonstrated in the second year of the project.

The economy, stability, and safety of local communities in densely forested territories rely very heavily on the sustainability of the sector. As such, the threat of biotic agents and the degradation of forests causes significant stress in sometimes already vulnerable communities. Using space-based technologies and other solutions to address these challenges is a commitment of FOCUS, towards a precision management of the forest.

Work performed

Activities in the first year of the project were focused on collecting high-quality field and aerial data to calibrate and validate the enhanced forest disturbance and tree health algorithms of the consortium.
In particular, the consortium conducted intensive mapping and sampling campaigns in central Portugal to identify, characterize, and monitor critical variables of interest. This effort allowed the description of baseline conditions and the full decline cycle of Maritime Pines infected with Pinewood Nematode. The spectral analysis performed on samples collected at the test sites also provided important information for the development of successful detection algorithms applied to multi-mission data, centered around Sentinel-2 imagery.
Machine-learning algorithms were developed and tested by the consortium at University of Coimbra, to implement a flexible and robust detection solution capable of serving the forestry community. The algorithms will be demonstrated during the second year of the project.
VITO, deployed a substantial fleet of RPAS to collect multi-, hyperspectral, and thermal data. These valuable and novel datasets will allow the demonstration of the applicability of these platforms and of the data collected by the sensors. Furthermore, the data will be employed in the development of robust algorithms for the detection and characterization of forest disturbance connected to Pine Wilt Disease.
The reinforcement of processing capabilities of the Silvisense (S&T) was also achieved, in preparation of the integration of the datasets developed by the consortium.
The team also maintained a strong end-user engagement and mapped needs and requirements. These activities support the customization of the future data products to meet the needs of the user community. The engagement of the end-user community happened through a User Workshop, participation in trade shows and forestry events, and dedicated bilateral discussions.

Final results

Project FOCUS is expected to support enhanced practices in the management of forest disturbance and tree health. This will be accomplished through the enhancement of monitoring algorithms relying on multi-sensor data. Improved practices ensure a sustainable forestry sector, contributing to a better management of the territory and reinforcing the economy and cohesion of rural areas.
The volume of field, laboratorial and aerial data is unprecedented when integrated and will provide a source of information to be analyzed in the coming years from multiple perspectives.
Data collected so far yielded new perspectives on the interaction of physiological variables throughout the decline cycle upon Pinewood Nematode infection (first â€˜real-worldâ€™ monitoring of infected trees without inoculation).
The data also provides a new vision on the evolution of spectral signatures beyond the current body of knowledge, with a complete characterization of the physiological drivers motivating such changes.
Intensive field campaigns also supported the development of dense sets of calibration data for the development of detection algorithms applied to Sentinel-2. Preliminary results demonstrate a superior performance to previous solutions.
The hyperspectral campaign already conducted also generated an unprecedented dataset, which supports the detailed spectral characterization of infected trees and the development of entirely new integrated Satellite/RPAS solutions for decline detection in heterogeneous landscapes.