Extensive use of drainage ditches in European boreal forests and in some parts of North America has resulted in a major change in wetland and soil hydrology and impacted the overall ecosystem functions of these regions. An increasing understanding of the environmental risks associated with forest ditches makes mapping these ditches a priority for sustainable forest and land use management. Here, we present the first rigorous deep learning–based methodology to map forest ditches at regional scale. A deep neural network was trained on airborne laser scanning data (ALS) and 1,607 km of manually digitized ditch channels from 10 regions spread across Sweden. The model correctly mapped 86% of all ditch channels in the test data, with a Matthews correlation coefficient of 0.78. Further, the model proved to be accurate when evaluated on ALS data from other heavily ditched countries in the Baltic Sea Region. This study leads the way in using deep learning and airborne laser scanning for mapping fine-resolution drainage ditches over large areas. This technique requires only one topographical index, which makes it possible to implement on national scales with limited computational resources. It thus provides a significant contribution to the assessment of regional hydrology and ecosystem dynamics in forested landscapes.

To meet the sustainable development goals and enable sustainable management and protection of peatlands, there is a strong need for improving the mapping of peatlands. Here we present a novel approach to identify peat soils based on a high-resolution digital soil moisture map that was produced by combining airborne laser scanning-derived terrain indices and machine learning to model soil moisture at 2 m spatial resolution across the Swedish landscape. As soil moisture is a key factor in peat formation, we fitted an empirical relationship between the thickness of the organic layer (measured at 5479 soil plots across the country) and the continuous SLU (Swedish University of Agricultural Science) soil moisture map (R²= 0.66, p < 0.001). We generated categorical maps of peat occurrence using three different definitions of peat (30, 40, and 50 cm thickness of the organic layer) and a continuous map of organic layer thickness. The predicted peat maps had a higher overall quality (MCC = 0.69–0.73) compared to traditional Quaternary deposits maps (MCC = 0.65) and topographical maps (MCC = 0.61) and captured the peatlands with a recall of ca. 80 % compared to 50 %–70 % on the traditional maps. The predicted peat maps identified more peatland area than previous maps, and the areal coverage estimates fell within the same order as upscaling estimates from national field surveys. Our method was able to identify smaller peatlands resulting in more accurate maps of peat soils, which was not restricted to only large peatlands that can be visually detected from aerial imagery – the historical approach of mapping. We also provided a continuous map of the organic layer, which ranged 6–88 cm organic layer thickness, with an R² of 0.67 and RMSE (root mean square error) of 19 cm. The continuous map exhibits a smooth transition of organic layers from mineral soil to peat soils and likely provides a more natural representation of the distribution of soils. The continuous map also provides an intuitive uncertainty estimate in the delineation of peat soils, critically useful for sustainable spatial planning, e.g., greenhouse gas or biodiversity inventories and landscape ecological research.

Application of Object Oriented Image Classification and Markov Chain Modeling for Land Use and Land Cover Change Analysis

Human modification of hydrological connectivity of landscapes has had significant consequences on ecosystem functioning. Artificial drainage practices have fundamentally altered northern landscapes, yet these man made channels are rarely considered in ecosystem management. To better understand the effects of drainage ditches, we conducted a landscape-scale analysis across eleven selected study regions in Sweden. We implemented a unique approach by backfilling ditches in the current digital elevation model to recreate the prehistoric landscape, thus quantifying and characterizing the channel networks of prehistoric (natural) and current (drained) landscapes. Our analysis detected that 58% of the prehistoric natural channels had been converted to ditches. Even more striking was that the average channel density increased from 1.33 km km⁻² in the prehistoric landscape to 4.66 km km⁻² in the current landscape, indicating the extent of ditching activities in the northern regions. These results highlight that man-made ditches should be accurately mapped across northern landscapes to enable more informed decisions in ecosystem management.

Monitoring the changes in soil organic carbon (SOC) pools is critical for sustainable soil and agricultural management. This case study models total and active organic carbon dynamics (2015/2016 to 2019/2020) using digital soil mapping (DSM) techniques. Model predictors include topographic variables generated from light detection and ranging data; soil and vegetation indices derived from Landsat satellite images; and soil and crop inventory information from Agriculture and Agri-Food Canada to predict total organic carbon (TOC) and permanganate oxidizable carbon (POXc) at the 0–15 cm depth increment for a 37 km² study area in Truro, Nova Scotia. Quantile Regression Forest and stochastic Gradient Boosting Model were utilized for prediction. Although both models performed equally well for predicting TOC and POXc, the accuracy of TOC predictions (e.g., concordance correlation coefficient (CCC) = 0.67) was better than POXc predictions (e.g., CCC = 0.53). The Landsat variables and crop inventory were dominant predictors, while topographic variables across the relatively homogeneous terrain had relatively little influence. During the study period, changes in POXc were predicted across 98% of the study area, with a mean absolute loss of 5.77 (±11.48) mg/kg/year, and in TOC on 27% of the area, with a mean absolute loss of 0.15 (±0.09) g/kg/year. While the annual crop fields observed the highest loss of TOC and POXc, the decline in pasture–grassland–forage fields was relatively low. The study reinforced the effectiveness of DSM for modeling multiple SOC pools at the farm to landscape scales.