The role of diurnal freeze-thaw cycles for summer melting on sea ice

Onset and duration of the summer melt period represent important variables for the seasonality of global sea-ice coverage and the analysis of climate variability in polar regions. The results of the former DFG project “Surface melting / Antarctic” provide 20 years of data that can be used to investigate the long-term variability of the summer melt period on Antarctic sea ice. Our recent findings show diurnal freeze-thaw cycles to be significantly responsible for physical changes at the upper sea-ice surface during summer. Motivated by a lack of attention to this process in former studies, we will investigate the large-scale spatial and temporal characteristics of the occurrence of diurnal freeze-thaw cycles by means of satellite-derived microwave data with high temporal resolution. A bihemispheric comparison of the character of diurnal variations and their role in sea-ice surface changes at the transition from spring to summer is meant to provide new insights into differences in surface melting between the Arctic and Antarctic. The analysis of diurnal freeze-thaw cycles and their causes in this context is innovative and will significantly enhance our understanding of processes involved in the metamorphosis of snow and sea ice. The scientific links to other research fields are manifold and include the possibility to improve parameterizations of surface albedo and microwave emissivity as well as a better understanding of the surface energy and sea-ice mass balances at the commence of summer.


Fig. 1. Amplitude of the diurnal variation of brightness temperatures (37 GHz, vert. pol.) on a) Arctic sea ice, Jun 24, 2004 and b) Antarctic sea ice, Dec 25, 2004; respectively 86 days after the beginning of the summer period (Apr 1/ Oct 1), for sea-ice areal coverage of at least 15%.

To account for the lack of long-term observations of melting on Antarctic sea ice, the MEDAntS datset needs to be analysed intensively. Moreover, diurnal freeze-thaw cycles, recognized as a central process contributing to snow melt in the Antarctic have not been included in large-scale melt investigations in the Arctic. In general, bi-hemispheric differences in the evolution of melt processes during summer are still to be understood in detail. These shortcomings will be compensated for with this proposed project.

Our primary objective is to analyze the spatial and temporal characteristics of surface melting, which is monitored through completely new methods, namely a diurnal microwave brightness temperature index and the MeDeA-algorithm. Moreover, the importance of diurnal cycles of snow wetness and snow temperature will be studied and assessed in both hemispheres through remote sensing and modelling. The results allow for a more detailed detection of climate variability and change in polar environments.

The main goals are as follows:

1. to detect and analyse the long-term variability of snowmelt characteristics that were identified with the MeDeA-algorithm. (first-time for the Antarctic and in comparison with existing methods for the Arctic).

2. to give insight into the spatial distribution, the temporal characteristics and potential trends of diurnal freeze-thaw cycles.

3. to identify external forcings for the spatial and temporal variability of freeze-thaw cycles and the superimposed melt dynamics.

4. to improve the understanding of the diurnal cycle of snow properties associated with melt processes at the transition to summer.

5. to explain the bihemispheric differences in the summer melt period and to interpret hemispheric particularities in superimposed ice formation and the surface energy balances.

Parallel to the preparation of long-term remote sensing data for an accurate comparison of snowmelt signals in both hemispheres, a modelling of snow properties with the SNTHERM model was performed during the first project year (07/2008 – 06/2009). As itemized in the work schedule of the current project (Wi 3314/-1), snow properties were simulated in the background of atmospheric forcing from NCEP reanalysis-2 data. The ouput of these simulations was used to derive snow microwave emissivities from the MEMLS2 microwave emission model. The SNTHERM model was adjusted to snow on sea ice by Nicolaus et al. [2006].

During the first project stage, technical implementation and adjustment of the combined thermodynamic/emission modelling was emphasized. Preliminary result are shown in Figure 2. The models are able to resolve the hemispheric contrasts observed with satellite data. A weaker increase in snow wetness accompanied by a much stronger temporal variability as compared with the Arctic induces decreasing emissivities and increasing diurnal emissivity amplitudes on sea ice in the Antarctic.


Fig. 2. Preliminary results of combined thermodynamic/emission modelling for selected snow profiles in the Arctic (80°N, 155°E, left panel) and Antarctic (70°N, 50°W, right panels), showing the seasonal variation of snow wetness and density (upper panels) and the resulting microwave emissivities at 19 and 37 GHz (vert. pol.) as well as the diurnal emissivity amplitude.

Currently, model results from different areas within the Arctic and Antarctic are compared with satellite data from the respective region. Examining high-resolution (4 times daily) emissivity simulations in comparison with SNTHERM snow property changes is expected to reveal distinct processes within the snow that are responsible for observed brightness temperature changes (see science plan of follow-up proposal No. 2). Results of these investigations will be used to evaluate the applicability of new satellite data channel combinations for the large-scale and long-term identification of early melt snow processes.

Scientists

Dr. Sascha Willmes
Dpt. Environmental Meteorology
University of Trier, Germany

Prof. Dr. Christian Haas
Department of Earth & Atmospheric Sciences
University of Alberta, Edmonton, Canada

Dr. Marcel Nicolaus
Polar Environmental Centre
Norwegian Polar Institute, Tromsö, Norway

Research areas

Antarctic and Arctic sea ice

Publications

Willmes S, Haas C, Nicolaus M, Bareiss J, 2009. Satellite microwave observations of the interannual variability of snowmelt on sea ice in the Southern Ocean. Journal of Geophysical Research, 114, C03006, doi:10.1029/2008JC004919.

Willmes S, Haas C, Nicolaus M, 2009. High radar-backscatter regions on Antarctic sea ice and their relation to sea-ice and snow properties and meteorological conditions. International Journal of Remote Sensing (in press).

Homepage

Find more about The role of diurnal freeze-thaw cycles for summer melting on sea ice at http://www.uni-trier.de/index.php?id=10722#c29609

Research funding organisation

German Research Foundation

Project numbers: WI 3314/1-1
Funding period: July 2008 - June 2009