Evaluation of Operation IceBridge quick‐look snow depth estimates on sea ice

Abstract

We evaluate Operation IceBridge (OIB) “quick‐look” snow depth on sea ice retrievals using in situ measurements taken over immobile first‐year ice (FYI) and multiyear ice (MYI) during March of 2014. Good agreement was found over undeformed FYI (−4.5 cm mean bias) with reduced agreement over deformed FYI (−6.6 cm mean bias). Over MYI, the mean bias was −5.7 cm, but 54% of retrievals were discarded by the OIB retrieval process as compared to only 10% over FYI. Footprint scale analysis revealed a root‐mean‐square error (RMSE) of 6.2 cm over undeformed FYI with RMSE of 10.5 cm and 17.5 cm in the more complex deformed FYI and MYI environments. Correlation analysis was used to demonstrate contrasting retrieval uncertainty associated with spatial aggregation and ice surface roughness.

Intercomparison of snow depth retrievals over Arctic sea ice from radar data acquired by Operation IceBridge

“…If samples with rough-ness > 10 cm were not included in the comparisons, there would be a consistent decrease in the standard deviation of the differences for all retrievals. This suggests that all the retrievals seem to be affected by surface roughness, consistent with the results reported by King et al (2015). However, the correlation values did not increase for all algorithms (NSIDC: 0.29 to 0.34; GSFC-NK: 0.45 to 0.61; SRLD: 0.66 to 0.71; Wavelet: 0.72 to 0.67; JPL: 0.63 to 0.60).…”

“…Increased snow depth was associated with local regions of deformed ice where drifted accumulation was found in proximity to convergence features (i.e., rafting, rubble, and pressure ridging). Those areas were described as deformed FYI in King et al (2015) and were shown to have a higher mean depth of 20.7 ± 11.4 cm. Transect variations in density were conservative, with a mean of 306 kg m −3 and standard deviation of 50 kg m −3 , which is comparable to the assumed climatological mean of ∼ 320 kg m −3 near the end of the winter.…”

“…Snow depths within the radar footprints were sampled after the overflights. An extended description of the 2014 Eureka study site and field campaign can be found in King et al (2015). As part of the field campaign, a linear transect was established to characterize local-scale variations in snow depth and density in proximity to the OIB snow-radar footprint.…”

Linking Regional Winter Sea Ice Thickness and Surface Roughness to Spring Melt Pond Fraction on Landfast Arctic Sea Ice

“…Areas of thick ice, such as ridges in an otherwise smooth FYI zone can be underestimated if they are proximal to areas of thin ice and occur with the footprint of the EM sensor. Unfortunately, it was not possible to obtain co-located in situ snow depth observations in our study area and due to the uncertainties associated with snow depth estimates in complex ice environments Operation IceBridge (OiB) and CryoVex data were not included in the analyses [40]. We were able to obtain in situ snow depth measurements, which were collected using a MagnaProbe near Eureka in the CAA between 8 and 15 April 2016 (Figure 2).…”

“…However, MYI f p is not significantly correlated with surface roughness. Surface roughness can be defined for scales ranging from meters to kilometers, with footprint-scale uncertainties such as snow roughness and errors associated with sensor position during airborne measurement acquisition (yaw and pitch angles, and slant range distortion) exhibiting a stronger effect on the measurements than at larger spatial scales [26,40]. In order to minimize the influence of measurement uncertainty on surface roughness measurements we performed a moving average on the surface roughness data (~2400 m) and found stronger correlations between surface roughness and f p for both FYI (r s = −0.84) and MYI (r s = −0.34).…”