Multi-Frequency Active and Passive Microwave Interactions of Arctic Sea Ice

Objectives

The 5.52 GHz C-band radar scatterometer and HUTRAD (Helsinki University of Technology Radiometer) brightness temperature measurements from the 6.9 GHz microwave radiometer captured two warm, moist air intrusions on the MOSAiC floe between 15 and 20 April 2020. For the first time, near-coincident measurements of Ka, Ku, X, C, and L-band microwave data collected in January-February 2020 during Leg 2 show a detailed understanding of how radar waves at multiple wavelengths penetrate through snow-covered sea ice. These measurements will further improve forward models of the radiative transfer in the snow and sea ice. The Ka- and Ku-band scatterometer captured meteorological events such as strong winds (9-15 November 2019 during Leg 1) and rain-on-snow (11-16 September 2020 during Leg 2) events that modified critical snow and sea ice geophysical variables, including snow depth and sea ice thickness. These changes are effective on a satellite footprint scale, which will allow comparison to the temporal evolution of the ground-based and satellite scatterometer/radar/radiometer measurements. The MOSAiC campaign has provided us with critical and unprecedented validation data of snow-covered sea ice over the annual cycle, which will be essential for us to timely monitor a rapidly evolving Arctic sea ice regime from satellites. These observations can also be compared in a rapidly changing Antarctic sea ice environment where 2023 has the lowest sea ice extent.

Warm Air Intrusions on Winter Arctic Sea Ice: Impact on Active and Passive Microwave Signatures

The MOSAiC sea ice floe experienced two significant warm air intrusions between 15 and 20 April 2020, which were detected by the 6.9 GHz HUTRAD surface-based passive radiometer and the 5.52 GHz C-band radar scatterometer (Figure 1, left). During both events, the floe experienced large-scale surface glazing, a thin ice crust forming on the snow surface (Figure 1, right). The lower-frequency C-band radar and HUTRAD radiometer channels were found to be less sensitive (Figure 2, bottom) to snow surface changes. This could be exploited in future through the synergy of active/passive data integration (backscatter + brightness temperature) methods to effectively retrieve sea ice concentration from upcoming satellite missions like the Copernicus Imaging Microwave Radiometer (CIMR) and Sentinel-1 Synthetic Aperture Radar (SAR) Legacy missions, both operating at C-band microwave frequencies.

Ka, Ku, X, C, and L-band radar signals originate from different depths on snow-covered sea ice

Ka, Ku, X, C, and L-band data collected during winter show varied penetration of radar signals at multiple frequencies on snow-covered sea ice. Figure 3 illustrates the first-ever, high-resolution multi-frequency range vs HH-polarized backscatter measurements of a daily-averaged Ka, Ku, X, C, and L-band dataset at 25 (top) and 45 degrees (bottom) incidence angles, acquired on the winter MOSAiC floe on 25 January (Ka, Ku, X, and L-band) and 3 February (C-band) 2020. Our analysis clearly demonstrates strong frequency-dependent backscatter separability between Ka and L-band as a function of penetration depth. Higher backscatter and smaller ranges (dotted vertical lines) at Ka-band suggest dominant backscatter from the snow surface, while much lower backscatter and longer ranges from the lower frequency L-band suggest backscatter originating from the sea ice volume. The classical shift in ranges to a greater depth with increasing incidence angle is noteworthy, strongly demonstrating the shift of microwave scattering from surface scattering (at 25 degrees) to volume scattering (at 45 degrees).

Wind and Rain on Snow-Covered Sea Ice: Impact on Ka- and Ku-band Radar Altimetry Waveforms

The MOSAiC sea ice floe experienced significant and numerous winter storm events and a summer rain-on-snow event. Nandan et al. (2023) show the impact of two wind events from November 2019 dynamically that impacted snow geophysical properties, which were effectively detected using the Ka- and Ku-band surface-based radar and a high-resolution terrestrial laser scanner. During both events, Ka and Ku-band radar waveforms and backscatter at nadir underwent significant changes in response to surface topography changes that were detected by the laser scanner. Radar waveforms detected the presence of previous air/snow interfaces buried beneath newly deposited snow (Figure 4). The additional scattering from these previous interfaces can significantly impact the range retrievals from Ka and Ku-band satellite altimeters, affecting the accuracy of snow depth retrievals on sea ice.

The September 2020 rain-on-snow event introduced liquid water into the snow and amplified melting, as documented by Stroeve et al. (2022). During rain, the surface-based Ka and Ku-band backscatter from KuKa radar decreased 4-fold compared to the non-rainy period. Snowpack refrozen after the rain stopped increased the backscatter by up to 6-fold. Microwave emissivity at 19 and 89 GHz increased during rain and decreased during refreezing. Analysis of Ku-band CryoSat-2 satellite waveform shape and backscatter (Figure 5, left), as well as brightness temperatures from AMSR2 (Figure 5, right), were clearly impacted by this rain-on-snow event. The study also shows how rain followed by refreezing permanently changed the snow's geophysical properties, which can further affect the accuracy of sea ice thickness (from satellite radar altimetry) and concentration (from satellite radiometry) retrievals and melt onset timing (from satellite imaging radars).