The HIGHTSI model results have been validated against field measurements in fast ice areas in several field campaigns where it has yielded consistently accurate snow/ice thickness values. Also our compar-isons show that HIGHTSI produces useful sea ice information. However, the initial ice thickness values from the IC’s seem to underestimate the ice thickness. In the next winter we are going to run two opera-tional versions of HIGHTSI in parallel with 10-days ECMWF forecasts. One is using the IC ice thickness as its initial information, and the other is run throughout the whole winter without this information, only taking the previous model outputs as its inputs. Additionally the ice edge and ice concentration input information is derived from SAR data. These model versions will be compared to each other and also to the operational model based on Zubov’s equation. Additionally we will study the joint use of the parallel HIGHTSI model results.
recent SAR data), and the HIGHTSI five-day forecast for the same day (right).
The HIGHTSI snow thickness forecasting seems to give reasonable results. This can be explained by the good quality of weather forecasts (external forcing of HIGHTSI) from ECMWF and by that HIGHTSI models snow melting with a relative good accuracy (a new surface albedo parametrization used). In the future we hope that we can also utilize the snow cover information produced by HIGHTSI in sea ice SAR data interpretation.
Comparison to operational ITC’s also provided good results. One natural reason for this is that both HIGHTSI and the ITC’s use the same initial information from IC’s. However, our evaluation results from earlier winters have clearly shown that ITC’s give good estimates of the ice thickness compared to the measured values [7].
We are going to continue the evaluation of HIGHTSI yearly using all possible reference data sets available, the most important ones being the ice thickness measurements made by ice breakers and our operational ITC’s. In 2007–2008 the starting of the model was delayed because of technical reasons.
There also were some breaks in the service because of computer problems (full hard disk). In general the operational HIGHTSI model setup seems to work reliably and the results are reasonable. The evaluation methods are now ready and tested for the next season and HIGHTSI can be started from the beginning of the ice season, and yearly evaluation reports will be delivered, and HIGHTSI parametrization will be developed and adjusted accordingly.
References
[1] J. Launiainen, B. Cheng, Modeling of ice thermodynamics in natural water Bodies, Cold Reg. Sci Technol., v. 27, n. 3, pp. 153–178, 1998.
[2] B. Cheng, T. Vihma, J. Launiainen, Modeling of the superimposed ice formation and sub-surface melting in the Baltic Sea, Geophysica, v. 39, n. 1-2, pp. 31–50, 2003.
[3] B. Cheng, T. Vihma, R. Pirazzini, M. Granskog, Modeling of superimposed ice formation during spring snowmelt period in the Baltic Sea, Ann. Glaciology, v. 44, pp. 139–146, 2006.
and ice thermodynamics in the Arctic Ocean with CHINARE 2003 data, J. Geophys. Research, v.
113, 2008.
[5] J. Karvonen, M. Simila, I. Heiler, Ice Thickness Estimation Using SAR Data and Ice Thickness History, Proceedings of the IEEE International Geoscience and Remote Sensing Symposium 2003 (IGARSS’03), v. I, pp. 74-76, 2003.
[6] J. Karvonen, B. Cheng, M. Simila, Baltic Sea Ice Thickness Charts Based on Thermodynamic Ice Model and SAR Data, Proc. of the International Geoscience and Remote Sensing Symposium 2007 (IGARSS’07), pp. 4253-4256, 2007.
[7] J. Karvonen, J. Haapala, J. Lehtiranta, A. Seina, Polarview@FIMR: WWW-based Delivery of Baltic Sea Ice Products to End-Users, Proc. of the International Geoscience and Remote Sensing Sympo-sium 2007 (IGARSS’07), pp. 1242-1245, 2007.
[8] J. Karvonen, B. Cheng, M. Simila, M. Hallikainen, BalticSea Ice Thickness Charts Based on Ther-modynamic Snow/Ice Model, C-Band SAR Classification and Ice Motion Detection, Proc. of the International Geoscience and Remote Sensing Symposium 2008 (IGARSS’08).
Susann Haase
Department of Biological and Environmental Sciences PL 65 (Viikinkaari 1), 00014 University of Helsinki susann.haase@helsinki.fi; phone+358-445155646
Abstract
Sea ice plays an important role in heat flux mechanisms of the oceans such as air-ocean interaction, thermohaline mixing and albedo feedback. In addition, the brine channels of sea ice serve as a habitat for various organisms. A main factor controlling the biology of sea ice, but also optical properties, is dissolved organic matter (DOM) which is strongly influenced by the initial freezing process. Therefore, the behavior of DOM during the ice formation needs to be described in detail to allow predictions about the character and concentration of DOM in sea ice.
This study about the freeze fractionation of DOM from the Gulf of Finland used 450 liter tanks to grow sea ice at -5°C. The initial conditions were compared to the ice and under-ice water by measuring the absorption coefficient of chromophoric DOM, fluorescent DOM and the size distribution of DOM. Excitation-Emission Matrices (EEMs) of DOM fluorescence were used for parallel factor analysis (PARAFAC).
In order to investigate the behavior of DOM relative to salts, an enrichment factor was calculated for CDOM and FDOM for the initial water samples and one week after ice formation. The mean enrichment factor was found to be 1.4 in ice whereas 1 in under-ice water which means that enrichment only took place within the ice. The size exclusion chromatography shows a shift in size distribution towards higher molecular size in ice. This change is less obvious for young ice and hence depends on physical processes during the ice formation and growth. The components identified by PARAFAC behave differently during the freeze fractionation, but were always enriched in the ice relative to salts.
The tank experiment leads to the assumption that the freeze fractionation of DOM depends on molecular size, composition of DOM and physical processes such as diffusion and convection.