Mapping Surface Organic Soil Properties in Arctic Tundra Using C-Band SAR Data

Abstract

Surface soil organic carbon (SOC) content is among the first-order controls on the rate and extent of Arctic permafrost thaw. There is a large discrepancy in current SOC estimates in Arctic tundra, where sparse measurements are unable to capture SOC complexity over the vast tundra region. Synthetic Aperture Radar (SAR) data are sensitive to surface vegetation, roughness and moisture conditions, and may provide useful information on surface SOC properties. Here we investigated the potential of multitemporal Sentinel-1 C-band SAR data for regional SOC mapping in Arctic tundra through principal component analysis (PCA). Multiple in-situ SOC datasets in the Alaska North Slope were assembled to generate a consistent surface (0-10 cm) SOC and bulk density dataset (n=97). The radar VV backscatter shows strong correlation with surface SOC, but the correlation varies greatly with surface snow, moisture and freeze/thaw conditions. However, the first principal component (PC1) of radar backscatter time series from different years shows spatial consistency representing dominant and persistent surface backscatter behavior. The PC1 also shows strong linear correlation with surface SOC concentration (R=0.65, p<0.01), and an exponential relationship with bulk density (R=-0.65, p<0.01). The resulting predicted SOC maps show much lower soil bulk density and higher SOC concentration in the southern shrub tundra area than the northern coastal region, consistent with insitu data. Our analysis shows it is possible to separate the effects of different factors on the radar backscatter response using PCA and multitemporal SAR data, which may lead to more effective satellite-based methods for Arctic SOC mapping.

Hybrid Transformer Network for Soil Moisture Estimation in Precision Irrigation

“…(2) using the transformed auxiliary-RZSM variables in Eq. (4) as, ŷg (t) = fp g ( cs g (t))∀ P g , s ∈ {1, ..., S}, t ∈ {1, ..., T }, (5) where ŷg (t) is the predicted RZSM. The objective of each hybrid TNN predictor neuron, denoted as P g , is to find out the relationship between cs g (t) and ŷg (t).…”

Using Relationships Between Vegetation and Surface Soil Biogeochemical Properties to Assess Regional Soil Carbon Inventories for South Baffin Island, Nunavut, Canada

“…SoilGrids is a global database that uses soil profiles combined with environmental covariates like vegetation and hydrology to predict soil C at 250‐m resolution (Batjes et al., 2020; Poggio et al., 2021). Finally, C‐Band SAR is used to measure and upscale soil C by remotely sensing surface roughness and vegetation structure as a proxy for soil C content (Bartsch et al., 2016; Yonghong et al., 2023).…”