SEA ICE MICROWAVE EMISSION MODELLING APPLICATIONS

Transcription

1 SEA ICE MICROWAVE EMISSION MODELLING APPLICATIONS R. T. Tonboe, S. Andersen, R. S. Gill Danish Meteorological Institute, Lyngbyvej 100, DK-2100 Copenhagen Ø, Denmark Tel.: , L. Toudal Pedersen Ørsted, bld. 348, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark ABSTRACT Applications such as: ice concentration mapping using microwave radiometer data, snow cover mapping on sea ice and microwave sounding over sea ice are affected by the variable sea ice surface emissivity. A sea ice version of MEMLS (Wiesmann & Mätzler, 1999) is used to study the sea ice emissivity. The output from a thermodynamic model using ECMWF reanalysis data is used as input to the emissivity model in order to study the seasonal variability in the Arctic Ocean. These simulations are discussed with respect to the different applications. This ongoing work is done in the EU 6 th framework programme project DAMOCLES. INTRODUCTION The Arctic sea ice cover is undergoing changes (Wadhams, 1990). The actual sea ice extent especially in summer is retreating and the area covered by sea ice is getting smaller. This has been confirmed by microwave radiometer satellite observations since At the same time the Arctic atmosphere and the sea ice surface properties are changing. Different satellite microwave radiometer ice concentration algorithms used to map sea ice extent and area have different sensitivities to the atmosphere and the ice surface properties. The sea ice trend mapped with different algorithms is therefore different. The ice surface emissivity variability is the primary error source for the microwave radiometer ice concentration estimate over the near 100% ice cover in the Arctic Ocean. The ice concentration estimate error is in fact much larger than the actual ice concentration variability so that for high ice concentrations typical for the winter ice cover in the Arctic Ocean (>98%) there is no correlation between actual ice concentration and the radiometer ice concentration estimate. This has been confirmed by extensive comparisons with SAR data. The variability of the microwave emissivity is simulated using a sea ice version of MEMLS (Wiesmann & Mätzler, 1999). The aim with our model experiments is to start understanding the microwave emission from sea ice in applications such as ice concentration mapping, snow cover mapping and correlations between sounding and window channels. The microwave emission model is used to simulate the sea ice brightness temperature (Tb) variability as a function of a seasonal snow cover. The model is a sea ice version of MEMLS (Wiesmann & Mätzler, 1999) described in Mätzler et al. (2006) and hereafter called the emission model. Snow and ice profiles collected during the GreenIce project are used as input to the emission model and output from a mass and thermodynamic sea ice model (Tonboe, 2005) is used to assess the seasonal variability of these parameters in the central Arctic Ocean. This model is hereafter called the thermodynamic model. SEASONAL VARIABILITY OF THE MICROWAVE BRIGHTNESS TEMPERATURE Figure 2 show the simulated snow surface density on multiyear ice at 82.5ºN; 0.0ºE between Fram Strait and the North Pole during the 2000/2001 winter season using ECMWF reanalysis data as input to the thermodynamic model. The simulations begin with a bare ice surface on Sep. 1., which is approximately the end of the melt season. Precipitation events less than 1kg/m 2 (<1mm SWE) are not included. The emission model is coupled to the thermodynamic model and the seasonal variability of both the brightness temperature are shown. On Jan. 23. snow precipitation combined with winds about 14m/s deposits a surface

2 snow layer of 290kg/m 3 on top of the existing 130kg/m 3 surface layer. The new surface layer gradually compacts to 330kg/m 3. Later on Feb. 13. light snow fall combined with winds about 5m/s deposits a new surface snow layer of 190kg/m 3. These snow surface density variations explains the simulated polarisation (Tbv-Tbh) variability during this period. Figure 1. The simulated snow cover depth and equilibrium free-board at 82.5N 0.0E 2000/2001. The colours of the snow surface line indicate the snow surface density. Figure 2. The simulated brightness temperatures using sea ice MEMLS and the output from the thermodynamic model shown in figure 1.

3 Figure 3. The simulated polarisation (Tbv-Tbh) using sea ice MEMLS and the output from the thermodynamic model shown in figure 1. SNOW COVER MAPPING Passive microwave radiometer data contain information on dry snow volume on land. High correlations are found locally between snow water equivalent (SWE) and microwave brightness temperature signatures (Mätzler et al., 2006). Markus & Cavalieri (1998) further derived an empirical relationship between snow cover depth on Antarctic sea ice and the spectral gradient between 19 and 37GHz in space borne SSM/I radiometer data. However, a universal SWE algorithm for snow on ice does not exist because the brightness temperature signature is also affected by layering, crusts and volume scattering (Mätzler et al., 2006). Pulliainen et al. (1999) demonstrated how physical models including several snow parameters might be inverted to derive single snow parameters (SWE) using space borne radiometer data. This approach seems promising also for future sea ice snow cover mapping. A significant effort is needed to bring these algorithms up to operational standard. WINDOW SOUNDING CHANNEL RELATIONS The assimilation of atmospheric parameters derived from microwave satellite data e.g. AMSU has a significant impact on both global (ECMWF) and regional (HIRLAM) weather prediction models (Prigent et al., 2004). The principle is to separate the atmospheric emissivity from the surface emissivity in the Tb measurements from space, thus the atmospheric part is parameterised in terms of temperature or water vapour. The surface emissivity, which is high for sea ice compared to open water, can be determined at frequencies where the atmosphere is largely transparent in the atmospheric windows. This allows referring the emissivities at sounding frequencies to those at the window frequencies by interpolation using emission models. Temperature sounding uses frequencies around 50GHz and humidity sounding uses frequency >85GHz. Figure 4 shows the relations between different channels. The data are simulated seasonal variability with the coupled thermodynamic and emissivity model described above.

4 Figure 4. Simulated relations between channels using the simulated ice floe in figure 1. SEA ICE CONCENTRATION The observed sensitivity of the different ice concentration algorithms using SSM/I satellite data e.g. NASA Team, Bootstrap and Near90 GHz to the atmosphere or surface brightness temperature is different as shown in Figure 5 for a case in the Arctic Ocean. SAR images show that the real ice concentration is stable. The sea ice emissivity normally varies during winter because of ice growth, snowfall, diurnal cycling and snow/ice metamorphism. Warm air outbreaks (Figure 5) over consolidated sea ice pack in the Arctic Ocean during winter offer a possibility to investigate the sensitivity of the ice concentration estimate to changes in the snow and ice cover emissivity in the course of days. While the actual ice concentrations remain close to 100% during and after the advection of warm air followed by snowfall and possibly rain, the formation of depth hoar and icy layers in the snow pack and in general the metamorphism accelerates as described in e.g. Drinkwater et al. (1995) and Garrity (1992). The changes are persistent but recover in the course of months with new snow accumulation, ice drift and new-ice formation. Emission models can be used to compute and analyse the sensitivity of the retrieved ice concentration to the microphysical properties of the snow and ice and to select and develop algorithms with low sensitivity to variations in the surface emissivity. Figure 6 shows the modelled sensitivity of the NASA Team ice concentration to the density of the upper snow layer and correlation length (grain size) of the bottom 1cm snow layer.

5 Figure 5 (left). The measured ice concentration estimate using 3 different algorithms during a warm air intrusion in the Arctic Ocean. The real ice concentration is approximately constant near 100% and the apparent changes are due to the variable atmospheric and surface emissivity. Figure 6 (right). The simulated ice concentration using the NASA Team algorithm. Here the upper snow layer density and the snow grain size (correlation length) are varied in the model. CONCLUSIONS Several important applications such as sea ice concentration mapping using microwave radiometers, sea ice snow mapping, atmospheric sounding of the semi-transparent atmosphere suffer from sea ice emissivity variations. A sea ice microwave emission model has been tested and the seasonal variability simulated by coupling the emission model to a thermodynamic model for sea ice. Improvements to the models have been identified and the further developments will be conducted in the EU 6 th framework programme project DAMOCLES which is at the same time the EU commisions contribution to IPY ( ). REFERENCES Drinkwater, M. R., R. Hosseinmostafa, & P. Gogineni (1995). C-band backscatter measurements of winter sea ice in the Weddell Sea, Antarctica. International Journal of Remote Sensing 16(17), Garrity, C. (1992). Characterisation of snow on floating ice and case studies of brightness temperature changes during the onset of melt. In: F. D. Carsey (Ed.). Microwave remote sensing of sea ice, Geophysical monograph 68 (pp ). Washington DC: American Geophysical Union. Jordan, R., E. Andreas, & A. Makshtas. Heat budget of snow covered sea ice at North Pole 4. Journal of Geophysical Research 104(C4), , Marbouty, D. An experimental study of temperature gradient metamorphism. Journal of Glaciology 26(94), , Markus, T. & D. J. Cavalieri. Snow depth distribution over sea ice in the southern ocean from satellite passive microwave data. Antarctic Sea Ice, Antarctic Research Series 74, 19-39, Mätzler et al., (Eds.), Thermal Microwave Radiation - Applications for Remote Sensing, IEE Electromagnetic Waves Series, London, UK, Pulliainen, J., J. Grandell & M. Hallikainen. HUT snow emission model and its applicability to snow water equivalent retrieval. IEEE Transactions on Geoscience and Remote Sensing 37, , Sturm, M. & J. Holmgren. Difference in compaction behaviour of three climate classes of snow. Annals of Glaciology 26, , Tonboe, R. A mass and thermodynamic model for sea ice. Danish Meteorological Institute Scientific Report 05-10, 2005.

6 Ulaby, F. T., R. K. Moore & A. K. Fung. Microwave Remote Sensing, from Theory to Applications, vol. 3. Dedham MA: Artech House, Wadhams, P. Evidence for thinning of the Arctic ice cover north of Greenland. Nature 345, , Warren, S. G., I. G. Rigor, N. Untersteiner, V. F. Radionov, N. N. Bryazgin, Y. I. Aleksandrov & R. Colony. Snow Depth on Arctic Sea Ice. Journal of Climate 12, , Wiesmann, A. & C. Mätzler (1999). Microwave emission model of layered snowpacks. Remote Sensing of Environment 70, Prigent, C., F. Chevallier, F. Karbou, P.Bauer, G. Kelly. AMSU-A land surface emissivity estimation for numerical weather prediction. NWP SAF, document no. NWPSAF-EC-TR- 009, EUMETSAT, 2004.