Sea ice growth and decay are critical processes in the Arctic climate system, but comprehensive observations are very sparse. We analyzed data from 23 sea ice mass balance buoys (IMBs) deployed during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition in 2019–2020 to investigate the seasonality and timing of sea ice thermodynamic mass balance in the Arctic Transpolar Drift. The data reveal four stages of the ice season: (I) onset of ice basal freezing, mid-October to November; (II) rapid ice growth, December–March; (III) slow ice growth, April–May; and (IV) melting, June onward. Ice basal growth ranged from 0.64 to 1.38 m at a rate of 0.004–0.006 m d–1, depending mainly on initial ice thickness. Compared to a buoy deployed close to the MOSAiC setup site in September 2012, total ice growth was about twice as high, due to the relatively thin initial ice thickness at the MOSAiC sites. Ice growth from the top, caused by surface flooding and subsequent snow-ice formation, was observed at two sites and likely linked to dynamic processes. Snow reached a maximum depth of 0.25 ± 0.08 m by May 2, 2020, and had melted completely by June 25, 2020. The relatively early onset of ice basal melt on June 7 (±10 d), 2019, can be partly attributed to the unusually rapid advection of the MOSAiC floes towards Fram Strait. The oceanic heat flux, calculated based on the heat balance at the ice bottom, was 2.8 ± 1.1 W m–2 in December–April, and increased gradually from May onward, reaching 10.0 ± 2.6 W m–2 by mid-June 2020. Subsequently, under-ice melt ponds formed at most sites in connection with increasing ice permeability. Our analysis provides crucial information on the Arctic sea ice mass balance for future studies related to MOSAiC and beyond.

No abstract

Forests are crucial in carbon sequestration and oxygen release. An accurate assessment of forest carbon storage is meaningful for Chinese cities to achieve carbon peak and carbon neutrality. For an accurate estimation of regional-scale forest aboveground carbon density, this study applied a Sentinel-2 multispectral instrument (MSI), Advanced Land Observing Satellite 2 (ALOS-2) L-band, and Sentinel-1 C-band synthetic aperture radar (SAR) to estimate and map the forest carbon density. Considering the forest field-inventory data of eastern China from 2018 as an experimental sample, we explored the potential of the deep-learning algorithms convolutional neural network (CNN) and Keras. The results showed that vegetation indices from Sentinel-2, backscatter and texture characters from ALOS-2, and coherence from Sentinel-1 were principal contributors to the forest carbon-density estimation. Furthermore, the CNN model was found to perform better than traditional models. Results of forest carbon-density estimation validated the improvements effectively by combining the optical and radar data. Compared with traditional regression methods, deep learning has a higher potential for accurately estimating forest carbon density using multisource remote-sensing data.