Flark area monitoring in boreal aapa mires using multi-resolution optical remote sensing

Peatlands have globally suffered significant degradation due to human activities which has necessitated monitoring of the condition of and changes in peatland ecosystems. With remote sensing, point-based in-situ observations can be upscaled to larger areas but there is a need to develop scalable monitoring methods. We hypothesize that the upscaling can be conducted by combining multispectral uncrewed aerial vehicle (UAV) and optical satellite imagery observations. We tested the hypothesis in predicting wet flark area extent, a key ecological indicator for patterned aapa mires with flarks, in five sites in central Finland. We asked: 1) How does the spatial and spectral resolution of widely used optical satellite image sensors (Landsat 8-9, Sentinel-2, and PlanetScope) influence flark area coverage prediction? 2) Are there seasonal and site-specific differences in prediction accuracy? 3) Is it feasible to upscale flark area coverage to larger mire areas? We employed UAV-derived flark area classification as a ground reference to compare predictive accuracy of satellite imagery data. We predicted flark area coverage using spectral bands and indices as explanatory variables in random forests regressions. Our findings revealed that all sensors provide accurate results, but there were differences in explanatory capacities between Landsat (pseudo-R² 32−84%, root-mean squared error (RMSE) 10−18%), Sentinel-2 (R² 61−92%, RMSE 6−14%), and PlanetScope (R² 46−92%, RMSE 6−17%) data. The shortwave infrared bands of Landsat and Sentinel-2 did not increase the prediction accuracy. There were notable site-specific variations in prediction accuracy despite all the sites having typical aapa mire wet flark–dry string patterns. With single-site models the prediction accuracies were similar for early and late summer conditions, but when transferring the models to the other sites, performance decreased considerably, especially with the models constructed with the late-summer imagery. Finally, we successfully upscaled the single-site models to detect flark area coverage across entire mire areas. Our results demonstrated that UAV-satellite image combination can be used to track key indicators of peatland conditions and monitoring changes in them.