Identifying Primary and Old-Growth Forests in the Western Balkans Using Remote Sensing
Figure 1. Partners of the Forest Beyond Borders project
Remote sensing technology has revolutionized the way we gather information about our planet's natural resources. With advances in satellite imaging and other remote sensing techniques, we can now identify primary and old-growth forests with remarkable accuracy for several reasons: differences in texture and structure of these forest types, the presence of unique spectral characteristics, and the availability of historical satellite data. These reasons will be further explained in the Section: The Technology behind the Pixels.
Recently, Space4Good and EuroNatur partnered to conduct a remote sensing project in the Western Balkans, more precisely in Albania, Bosnia and Herzegovina, Croatia, Kosovo, Montenegro, North Macedonia, and Serbia. The goal of the project is to identify primary and old-growth forests in the region, which are essential for maintaining biodiversity, mitigating climate change, and providing critical ecosystem services. Having these areas detected via data-driven methods, serves first to have an understanding of the geospatial distribution of these forests, second to quantify the amount of primary/old-forest growth hectares, and finally to have robust geo-based documentation of the information for proper management and monitoring of the areas which helps reinforce their protection.
Figure 2. The Balkans and the 7 countries highlighted
Stakeholder Network & Forest Beyond Borders Workshop
Besides the satellite-driven forest identification, other project objectives include establishing a network of environmental NGOs, scientists, and other stakeholders who will work together to protect these forests. Further, capacity development among national and local NGOs and other stakeholders will also be a key outcome.
The first capacity-building activity already took place last month. The “Forest Beyond Borders'' online workshop was hosted (in hybrid mode) at EuroNatur’s Office in Germany alongside the Space4Good team; it was geared towards expanding knowledge and providing networking opportunities amongst participants. The workshop consisted of GIS and Remote Sensing theory, a hands-on exercise, and round tables of discussion on different topics. The second workshop will take place in October whereby we hope to expand the knowledge of remote sensing in a more advanced way, and again promote the networking opportunity that is so important for the success of this project by exchanging knowledge across borders.
The Technology behind the Pixels
The method developed by the project aims to identify primary and old-growth forests along with their location in the Western Balkans.
These forests have unique characteristics, such as large trees, diverse species, and minimal human disturbance, that are relevant for incorporating into a remote sensing analysis and that make them particularly valuable for conservation. Below are some of the distinctive elements exhibited by the remote sensing data of these forest types that are useful for the successful identification of this land cover type:
Different spectral characteristics: Primary and old-growth forests have distinct spectral characteristics, which can be captured by remote sensing technologies. For example, these forests typically have a higher density of vegetation, which results in a unique spectral signature that can be distinguished from other land cover types.
Texture and structure: The texture and structure of primary and old-growth forests are also different from other land cover types. Remote sensing data can capture the vertical and horizontal structure of these forests, which can be used to differentiate them from other forest types.
Historical data: Remote sensing data can be used to track changes in forest cover over time. This allows us to identify areas that have remained undisturbed for long periods, indicating the presence of primary and old-growth forests.
In this manner, the Space4Good team employed a variety of remote sensing analysis techniques and EO-based indices, utilizing satellite images retrieved from the Landsat and Sentinel families all the way from 1984 to the present. The methods include spectral analysis, which involves looking at the spectral characteristics, or the reflectance of light from the forest canopy, to determine its composition, and areas of inconsistent growth patterns, which might indicate more naturally grown forest ecosystems.
These analyses will be combined with machine learning methods and put forth as a robust geospatial model that enables the identification of forests that fit the parameters of choice. The remote sensing application will be combined with on-the-ground fact-finding to ensure accuracy, which will be actioned with the support of our local partners located in these seven 7 countries.
The Albania Sample Case
Albania has over 580 kha hectares of natural forest (source Global Forest Watch). Of these hectares, it is not possible to determine the areas of primary or old-growth forest in a scalable manner, utilizing traditional methods.
Zooming over an area in Albania it can be noticed that using the methodology outlined above, our models accurately identify old forest growth areas, with confidence (see pixels in dark green), including areas that had previously been unrecognized or understudied. This information will be critical for conservation efforts in the region, as it can inform land-use planning, restoration efforts, and other interventions to protect these valuable ecosystems.
Figure 3. Analysis area in Albania
Widescale impact
By the end of the project, maps will be available that visualize all primary and old-growth forests located in the Western Balkans. In this manner, the NGO partners and other interested stakeholders will have tangible, quantifiable information that can further validate the need to protect these forests as they have the necessary knowledge and capacity to do so.
In addition to identifying the primary and old-growth forests, the project will also be able to provide important insights into the health and status of these ecosystems, as our models are also able to detect forest areas experiencing stress or disturbances, such as logging.
The project will also contribute to the protection of these forests by raising awareness among local stakeholders about the importance of protecting these unique forest types. Many local stakeholders are not aware of the exact locations and size of these forests, nor the relevant legislation to protect them, making it difficult to enforce conservation actions around them. Moreover, with this study in hand, these countries can seek a process to request the protection of these areas according to higher European organisations and initiatives such as Natura 2000 and Unesco World Heritage.
In conclusion, the use of remote sensing technology and its diverse and complex techniques, combined with ground-truthing methods are promising approaches to identify primary and old-growth forests in the Western Balkans. The project will help protect these valuable forests, preserve biodiversity and contribute to the sustainable development of the region.
Would you like more information or are you interested in collaborating with Space4Good? Visit our website or contact us via hello@space4good.com.
Comentarii