Forestry Real-Life Data Collection Use Case

Soil Mission

Forests are changing faster than we can walk through them.

This RAISE Suite use case pairs drone imaging with ground-level soil sensing to spot early signs of forest stress, turning aerial and field data into an openly shared, FAIR-by-design resource for protecting ecosystems.

At a glance

Pilot Period

M6 – M40

(Feb 2026 – Dec 2028)

Partners Involved
Domain

Forestry / Environmental

Location / Study Sites

Mikrokleisoura Forest, Grevena, Greece (next to Aliakmonas river) 

Participants / Scale
  • 1-2 drone flights/month 
  • Ground-based sensors
  • Sentinel-2 for landscape context​
Previous
Next

Main Goal

To investigate how drone-based multispectral imaging, complemented by ground-level sensing, can support systematic near real-time monitoring of forest ecosystem health and deliver, for the first time, scientific data on the condition, health and long-term evolution of the Mikrokleisoura forest to ASDAD that manages it.

A DJI Mavic 3M drone conducts 1-2 flights per month over fixed transects across five priority research zones, capturing canopy reflectance across five spectral bands (Green 560 nm, Red 650 nm, Red Edge 730 nm, NIR 840 nm, plus RGB) to
calculate NDVI, NDRE and GNDVI indices. These indices detect stress, disease or drought in the vegetation before it becomes visible to the naked eye. This will help timely decisions on ecosystem protection, erosion risk, and sustainable forest
management.

The study area spans approximately 1,200 of the forest’s 3,950 stremma, across the rugged highland terrain between the Aliakmonas river to the north and the Vourinos slopes to the south, at elevations of 470-850 m. Ground-based sensors deployed within the forest will add in-situ field data to validate and contextualise the drone-derived observations.

More information can be found here: https://dasosmikrokleisouras.gr/research.html

Domain Challenges

  • Forest accessibility: The forest covers a terrain at 570 m+ elevation with
    limited access points. Identifying suitable take-off locations, transporting
    equipment and planning flight paths around the terrain all require careful
    preparation for each mission.

  • Weather dependency: Multispectral flights require adequate, consistent
    sunlight to produce radiometrically reliable data. Cloud cover, haze or rain can affect data quality.

  • Battery flight time: Each DJI Mavic 3M battery provides approximately 35
    minutes of flight time. Covering five research zones totalling ~1,200 stremma requires careful mission planning across multiple sessions to ensure complete, overlap-consistent coverage.

  • Radiometric consistency over time: Each flight must produce data that is
    directly comparable to previous flights, regardless of the time of day or
    weather conditions at the time of capture. Without this consistency, it
    becomes impossible to distinguish real changes in forest health from
    variations caused by the measurement conditions.

Keywords

environment, forest, drones, ecosystem

Key Figures & Facts

Study Area

Mikrokleisoura, Grevena, Western Macedonia, Greece

Drone Platform

DJI Mavic 3M: RGB + 5-band multispectral (Blue, Green, Red, Red Edge, NIR)

Flight Frequency: 1-2 missions per month over fixed transects

Datasets
  • Multispectral orthomosaics (5 bands) per flight session
  • Vegetation index rasters (NDVI, NDRE, GNDVI)
  • Continuous soil sensor time series (moisture, temperature)
  • Flight logs and radiometric calibration records
  • All datasets deposited to an open repository under CC BY 4.0
Soil Sensoring

Continuous in-situ monitoring of moisture and nutrient levels

Monitoring period

6-8 months of active data collection

Image Resolution

~5 cm ground sampling distance at 80 m altitude

Use Case Implementation Phases

May–December 2026
Data Collection Without RAISE
Between May and August 2026 the team completed the foundational work for the pilot. Study visit was conducted to understand the terrain, canopy structure and natural boundaries of the forest. The full 3,950-stremma estate boundary was mapped, producing a precise digital map (KML) that defines the research area. Test flights with the DJI Mavic 3M produced the first NDVI orthophoto of the forest canopy, confirming that the system is functioning correctly and the data is suitable for analysis. A sample dataset is already publicly available on Zenodo (CC BY 4.0). Five priority research zones covering approximately 1,200 stremma were then identified and mapped, selected by accessibility, vegetation diversity and relevance to ASDAD interests. From 31 August 2026 onwards, systematic data collection flights take place at the end of each month across all five research zones. Each session produces over 12,000 photographs and approximately 120 GB of multispectral raw data, openly available for further processing and analysis by the research community.
May–December 2026
January–March 2027
SDK Integration & RAISE Onboarding
The RAISE Suite SDK is integrated into the pilot's data collection environment. The team onboards to the RAISE Portal, creates a machine-actionable Data Management Plan (maDMP), describes datasets with FAIR-aligned metadata, and connects sensors to the automated data ingestion pipeline.
January–March 2027
From March 2027 onwards
Data Collection with RAISE
The same campaign is repeated, now powered by RAISE Suite: data is captured automatically via the SDK and analysed directly on the RAISE Portal. The cycle ends with the publication of a FAIR-compliant forestry dataset on the RAISE Portal and EOSC, plus a scientific publication on the pilot's results.
From March 2027 onwards

Use Case Outcomes

A Lighter Data Workload

Automated Capture, Less Time and Cost

SDK-based automation captures drone and sensor data directly, cutting the time, storage and manual effort forest data management used to demand.

A Guided FAIR Workflow

One Integrated Toolset, No Expertise Gap

A single RAISE Portal workflow and maDMP replace fragmented tools, making FAIR, interoperable data collection achievable without specialist DMP skills.

An Open, Trusted Dataset

Recognised, Citable and GDPR-Compliant

A published, citable forestry dataset on the RAISE Portal and EOSC turns data sharing into visible, credited and compliant research output.

Relevant News

Into the Forest: Noosware Meets ASDAD in Mikrokleisoura 
July 6, 2026