Arctic tundra landscapes are highly complex and are rapidly changing due to the warming climate. Datasets that document the spatial and temporal variability of the landscape are needed to monitor the rapid changes. Synthetic Aperture Radar (SAR) imagery is specifically suitable for monitoring the Arctic, as SAR, unlike optical remote sensing, can provide time series regardless of weather and illumination conditions. This study examines the potential of seasonal backscatter mechanisms in Arctic tundra environments for improving land cover classification purposes by using a time series of HH/HV TerraSAR-X (TSX) imagery. A Random Forest (RF) classification was applied on multi-temporal Sigma Nought intensity and multi-temporal Kennaugh matrix element data. The backscatter analysis revealed clear differences in the polarimetric response of water, soil, and vegetation, while backscatter signal variations within different vegetation classes were more nuanced. The RF models showed that land cover classes could be distinguished with 92.4% accuracy for the Kennaugh element data, compared to 57.7% accuracy for the Sigma Nought intensity data. Texture predictors, while improving the classification accuracy on the one hand, degraded the spatial resolution of the land cover product. The Kennaugh elements derived from TSX winter acquisitions were most important for the RF model, followed by the Kennaugh elements derived from summer and autumn acquisitions. The results of this study demonstrate that multi-temporal Kennaugh elements derived from dual-polarized X-band imagery are a powerful tool for Arctic tundra land cover mapping.

Seasonal and interannual variations in global wetland area are a strong driver of fluctuations in global methane (CH4) emissions. Current maps of global wetland extent vary in their wetland definition, causing substantial disagreement between and large uncertainty in estimates of wetland methane emissions. To reconcile these differences for large-scale wetland CH4 modeling, we developed the global Wetland Area and Dynamics for Methane Modeling (WAD2M) version 1.0 dataset at a similar to 25 km resolution at the Equator (0.25 degrees) at a monthly time step for 2000-2018. WAD2M combines a time series of surface inundation based on active and passive microwave remote sensing at a coarse resolution with six static datasets that discriminate inland waters, agriculture, shoreline, and non-inundated wetlands. We excluded all permanent water bodies (e.g., lakes, ponds, rivers, and reservoirs), coastal wetlands (e.g., mangroves and sea grasses), and rice paddies to only represent spatiotem-poral patterns of inundated and non-inundated vegetated wetlands. Globally, WAD2M estimates the long-term maximum wetland area at 13 :0 x 106 km(2) (13.0Mkm(2)), which can be divided into three categories: mean annual minimum of inundated and non-inundated wetlands at 3.5Mkm(2), seasonally inundated wetlands at 4.0Mkm(2) (mean annual maximum minus mean annual minimum), and intermittently inundated wetlands at 5.5Mkm(2) (long-term maximum minus mean annual maximum). WAD2M shows good spatial agreements with independent wetland inventories for major wetland complexes, i.e., the Amazon Basin lowlands and West Siberian lowlands, with Cohen's kappa coefficient of 0.54 and 0.70 respectively among multiple wetland products. By evaluating the temporal variation in WAD2M against modeled prognostic inundation (i.e., TOPMODEL) and satellite observations of inundation and soil moisture, we show that it adequately represents interannual variation as well as the effect of El Nino-Southern Oscillation on global wetland extent. This wetland extent dataset will improve estimates of wetland CH4 fluxes for global-scale land surface modeling. The dataset can be found at https://doi.org/10.5281/zenodo.3998454 (Zhang et al., 2020).