LAUSR.org creates dashboard-style pages of related content for over 1.5 million academic articles. Sign Up to like articles & get recommendations!

Assessing the Potential of using Sentinel-1 and 2 or high-resolution aerial imagery data with Machine Learning and Data Science Techniques to Model Peatland Restoration Progress – a Northern Scotland case study

Photo from wikipedia

ABSTRACT Peatland is a globally important store of carbon. Peatland restoration efforts are being increasingly undertaken yet effective monitoring of landscape-scale restoration projects has been limited. A particular gap in… Click to show full abstract

ABSTRACT Peatland is a globally important store of carbon. Peatland restoration efforts are being increasingly undertaken yet effective monitoring of landscape-scale restoration projects has been limited. A particular gap in our understanding is the length of time required before a site reaches the target state. To address this, a classification model based on remote sensing data was developed for a peatland restoration area on blanket bog in northern Scotland, UK, to evaluate whether post-restoration trajectories followed predictable trends over time. The model was trained against a chronosequence of sites within a 20 × 10 km study area that are being restored following drainage and intensive non-native afforestation. Two versions of the model were created to compare the accuracy obtainable from the suite of Sentinel-2 satellite data versus sub-metre resolution aerial imagery from GetMapping (RGB and IR). The Sentinel-2 based model greatly outperformed the aerial imagery-based model. Adding surface slope to the classification did not significantly improve the accuracy of prediction. Prediction of starting and target land covers was very robust, and both the most recent and oldest restoration sites were well predicted spatially. The main uncertainties in the model were within sites of intermediate restoration age, and sites which underwent additional treatments after the initial restoration. Using standard vegetation and wetness indices as indicators, it was possible to track the progression of areas that had been felled and rewetted towards the spectral signal of the control blanket bog locations. A further study examined the use of multiple years of satellite data (2015–2021) and including Sentinel-1 SAR imagery, and confirmed the findings obtained with only a single climatically average year, and furthermore examined the efficacy of different restoration methods. We observed consistent trends of restoration sites beginning to resemble the target hydrologically and ecologically functional blanket bog state after 10–20 years post intervention.

Keywords: restoration; aerial imagery; northern scotland; peatland restoration; model; imagery

Journal Title: International Journal of Remote Sensing
Year Published: 2023

Link to full text (if available)


Share on Social Media:                               Sign Up to like & get
recommendations!

Related content

More Information              News              Social Media              Video              Recommended



                Click one of the above tabs to view related content.