Influence of snow depth and surface flooding on light transmittance through Antarctic sea ice
Stefanie Arndt, Klaus M. Meiners, Robert Ricker, Thomas Krumpen, Christian Katlein, Marcel Nicolaus
ocean Antarctic
snow ice
2
Temporal evolution of surface properties
Year-round snow cover Seasonal changes in snow properties dominated by e.g.
§ Diurnal freeze-thaw cycles
§ Internal snowmelt
ice
ocean
snow lead
atmosphere
melt pond
Ice and snow
transport (dri3)
Lateral mel6ng Bo9om
mel6ng/ freezing
Internal mel6ng
Ice thickness
Snow depth Snowfall
Flooding Snow-ice forma2on Internal snowmelt
Superimposed ice forma2on
snow
ice
Internal ice layers
Surface energy budget
lead
Incoming short-
wave radia4on Reflected short- wave radia4on Absorp4on
Transmission Sca<ering
Ocean heat flux Incoming/
Outgoing long- wave radia4on Turbulent
heat flux
melt pond
Conduc4ve heat flux
ice
ocean snow
atmosphere Energy budget
Incoming short- wave radia3on
Reflected short- wave radia3on
Absorp3on
Transmission Sca;ering
Ocean heat flux
ice
ocean snow
atmosphere Energy budget
Importance of transmitted heat fluxes
Mass budget of sea ice
bottom melt
Energy budget of the upper ocean
warming of the upper ocean
Under-ice ecosystem
changing habitat conditions for ice-associated organisms
lead
Incoming/
Outgoing long- wave radia5on Turbulent
heat flux
melt pond
ice
ocean snow
atmosphere Energy budget
Study side and measurements
Measurements:
§ Spectral solar radiation measurements: Remotely Operated Vehicle (ROV)
§ Total sea-ice thickness: Multi-frequency electromagnetic induction (GEM-2)
§ Snow depth: Magna Probe WISKEY
= Winter study on Sea ice and KEY species 14 August to 16 October 2013
Arndt et al., 2017 (under review, JGR)
Physical properties of the pack ice floe
Sea-ice thickness, I:
mean(I) = 0.93 ± 0.45 m
Snow depth, S:
mean(S) = 0.39 ± 0.13 m
Transmittance, T:
mean(T) = 0.0024 (0.24%) mode(T) = 0.0008 (0.08%)
Antarctic pack ice transmits less than 0.1% of the incoming solar radiation during early spring
Physical properties of the pack ice floe
Sea-ice thickness, I:
mean(I) = 0.93 ± 0.45 m
Snow depth, S:
mean(S) = 0.39 ± 0.13 m
Ice freeboard, F:
mean(F) = -0.08 ± 0.10 m
Physical properties of the pack ice floe
Ice freeboard, F:
mean(F) = -0.08 ± 0.10 m
ice water
S
I
snow (dry)
ice water
Sdry
I
slush
Swet
= -F
F ρs S
ρi
ρs
ρi ρsl=ρi
snow (dry)
Non-flooded Flooded
ice water
S
I
snow (dry)
ice water
Sdry
I
slush
Swet
= -F
F ρs S
ρi
ρs
ρi ρsl=ρi
snow (dry)
Non-flooded Flooded
Spectral optical properties
Spectral optical properties
Normalized difference indices (NDI) of under-ice irradiance spectra:
!"# = !! !! − !!(!!)
!! !! + !!(!!)
λ1, λ2: wavelength pairs (Mundy et al., 2007)
Correlation surfaces of normalized difference indices (NDI) for snow depth
Non-flooded Flooded
correlation: 0.52 correlation: 0.57
The heterogeneous snow on Antarctic pack ice obscures a direct correlation between the under-ice light field and snow depth
Katlein, Arndt et al., 2015
Summer&
Melt& Freeze&
Winter& Winter&
snow% snow%
ice% melt%pond%
ocean%
Optical properties highly correlated with snow surface properties (e.g.
melt ponds)
Light transmittance significant higher (summer FYI: 0.09, summer MYI: 0.05)
Comparison with Arctic studies
Conclusions
Antarctic pack ice transmits less than 0.1% of the incoming solar radiation during early spring
Ice freeboard and related flooding at the snow/ice interface dominates the spatial variability of the under-ice light regime Limitation in the use of snow-NDI prevents estimating light transmission from snow depth and vice versa
In contrast to Arctic sea ice, the dependency of light
transmittance of Antarctic sea ice on its surface properties is more obscure
Ice freeboard
Transmittance
Snow depth correlation
Arndt et al., 2017 (under review, JGR)
Outlook
New field data sets for improved process understanding of the vertical snow layer
Comparison of relations of surface properties and (spectral) light transmittance in the Weddell Sea (WISKEY) with East Antarctic (e.g. SIPEX-2)
Antarctic-wide up-scaling approaches of the under-ice light field require more detailed field data and analysis
Application of existing Chlorophyll-a –NDI for Weddell Sea on WISKY data set to investigate spatial variabilities in Chlorophyll-a (Meiners, Arndt et al., in prep.)