C12B-01
Sea ice thickness retrieval from L-band radiometry
Soil Moisture and Ocean Salinity (SMOS) is an earth observation mission developed by the European Space Agency to be launched in 2009. The main objective is to provide global measurements of soil moisture over land and sea surface salinity over ocean from L-band (1.4 GHz) radiometric observations. An exciting spin-off is the retrieval of sea ice thickness which we demonstrate to be possible due to the large penetration depth at L-band. SMOS will provide sea ice thickness information complementary to those from altimeters because of the expected sensitivity for thin ice thickness variations. Moreover, SMOS will provide data with an almost global coverage every second day. A three layer (ocean-ice-atmosphere) dielectric slab model is used to calculate the brightness temperature as a function of ice thickness and the dielectric properties. The dielectric properties depend on the relative brine volume as a function of bulk salinity and temperature. A model for the brightness temperature of a mixture of open water and sea ice reveals that the parameters ice concentration and thickness can hardly be retrieved both simultaneously. With the assumption of a closed ice cover the retrieval of ice thickness is feasible. The model calculations suggest a thickness sensitivity of up to 150 cm for low salinity (multi year or brackish) sea ice at low temperatures. At temperatures approaching the melting point the thickness sensitivity reduces to a few centimeters. For first year ice the modeled thickness sensitivity is roughly half a meter. The brightness temperature at 1.4 GHz (L-band) was measured in the Bothnia in Bay in March 2007 as part of the SMOS Sea-Ice campaign. The research aircraft was equipped with the Technical University of Denmark (TUD) Electromagnetics Institute Radiometer (EMIRAD). The EMIRAD measurements were coordinated with helicopter EM ice thickness measurements. The campaign was conducted under non- favorable conditions with temperatures around the melting point and there was only a relatively small overlap of EMIRAD and EM data. However, these measurements agree in general very well with the model results. The ice thickness from the L-band radiometer is in good agreement with the EM ice thickness in the range of 20 cm to 150 cm. The retrieval failed for a region of ice with a melted surface.
C12B-02
Enhanced Resolution Passive Microwave Sea Ice Motion Fields: Impacts on the 2008 Summer Melt Season and Long-Term Circulation Patterns
Passive microwave sea ice products are well-suited to tracking sea ice motion because of their frequent all- sky coverage. However, the low spatial resolution of historical passive microwave data limits the ability to capture smaller-scale motions. The NASA Earth Observing System Advanced Microwave Scanning Radiometer (AMSR-E) has significantly improved spatial resolution, allowing it to obtain more accurate and more detailed ice motion information. This improved resolution allows better characterization of sea ice circulation patterns and key divergence/convergence events. Winter 2007-2008 sea ice motions from AMSR-E show significant drift across the Arctic. This circulation advected thicker multiyear ice toward Greenland and Fram Strait and was replaced by first-year ice. This resulted in a situation at the beginning of the melt season where areas over much of the Arctic, including the North Pole, were covered primarily by first-year ice. Additionally, there was significant divergence and breakup of multiyear ice westward out of the Beaufort Sea. This led to a more dispersed multiyear pack in the Beaufort that was more vulnerable to summer melt. Both of these events influenced the character of the summer 2008 sea ice melt season in the Arctic. However, the AMSR-E record only extends back to 2002, limiting its ability to track long-term change and variability in the sea ice circulation. The lower resolution Special Sensor Microwave/Imager (SSM/I) and Scanning Multichannel Microwave Radiometer (SMMR) provide ice motions since 1979, but at a much reduced spatial resolution. Resolution enhancement techniques have been developed to obtain greater accuracy and higher spatial resolution. The enhanced resolution SSM/I products can potentially produce a long-term sea ice motion product that is more accurate and more detailed than currently available. This will allow the greater detail of the AMSR-E sea ice motions to be put in a long-term context. Resolution enhancement techniques have also been applied to AMSR-E data to yield even more detailed ice motion information, particularly through narrow straits and channels.
C12B-03
Satellite Remote Sensing of Fram Strait Sea Ice Volume Flux for Years 2003 to 2008
In the light of a reduced Arctic summer time sea ice extent it is of particular interest to monitor the sea ice volume flux out of the Arctic Ocean. Sea ice export through Fram Strait is by far the largest portion of the total Arctic ice export and amounts to about 10% of the total sea ice mass of the Arctic Ocean. Interannual perturbations in the sea ice transport through Fram Strait can modify the major water mass formation processes in the Greenland Sea and further downstream with consequences for the deep water formation and global ocean circulation. This paper is concerned with the determination of sea ice volume transports from satellites. To obtain the sea ice volume flux from space the three quantities sea ice area, motion, and thickness have to be known. The sea ice thickness is estimated using ICESat sea ice freeboard observations. Under the assumption of free floating ice, freeboard heights are converted to ice thickness estimates using QuikSCAT scatterometer data to discriminate between multi- and first-year ice in order to apply different ice density values. Sea ice concentrations and drift are calculated subsequently from AMSR-E 89 GHz data. Both are combined with the ICESat sea ice thickness estimates to obtain the sea ice volume flux in the Fram Straight region for the period 2003 to 2008. Results demonstrate the high interannual variability of the sea ice volume flux in that region. Different source regions of the sea ice leaving the Arctic through Fram Strait and the influence of atmospheric low pressure systems on the sea ice volume flux can be observed. Results are also compared to modeled sea ice volume fluxes from two coupled sea ice-ocean models. The comparison reveals an reasonable agreement for the total Fram Strait sea ice export between models and observations but large differences in the spatial distribution of the sea ice volume flux. In comparison to former Fram Strait sea ice volume flux estimates obtained during the 1990s no evident change of the total amount of sea ice export can be observed.
C12B-04
Some Inconsistencies Between MODIS and IMS Snow Products
We are now one decade into the new generation of satellite-based information about snow cover at the earth's surface. The two most widely used global scale products are from the Interactive Multisensor Snow and Ice Mapping System (IMS) produced by NOAA's National Ice Center, and from NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) satellite flown on board Earth Observing System (EOS) Aqua and Terra platforms. While a number of studies have evaluated or compared these products or their predecessors with each other and with other sources of information at a variety of spatial and temporal scales, a great deal of work remains in identifying the inconsistencies, diagnosing their causes, and identifying solutions. Here the focus is on daily observations across Northern Hemisphere lands with relatively low relief under clear sky conditions, when the most robust results are expected. While each product on occasion identifies snow when the other does not, IMS identifies snow more often than MODIS particularly during spring.
C12B-05
High resolution ground-based snow measurements during the NASA CLPX-II campaign, North Slope, Alaska
Active microwave radar has been shown to have great potential for estimating snow water equivalent (SWE) globally from space, however application has been limited by a lack of experiments covering a wide range of radar parameters and snow conditions. The NASA Cold Land Processes Experiment (CLPX) II, building on results from the first CLPX in 2002/03, focused on calibration/validation of PolSCAT measurements for estimating SWE in 2007 and 2008. We made detailed measurements with several different high-resolution instruments in five separate 10-meter trenches during CLPX II on the North Slope of the Brooks Range, near the Toolik Research Station in Alaska in February 2008. Measurements were made with 3 different ground- based radar systems, covering the frequency range of the PolSCAT instrument as well as the majority of space-borne active radars currently operational (2-18 GHz), at 0 and 35 degrees incidence and multiple polarizations, at cm resolution. In-situ measurements of hardness and microstructure were made with a high- resolution penetrometer, at 20 cm horizontal and 0.04 mm vertical resolution. Dielectric properties were independently measured with a probe, at 50 cm horizontal and 5 cm vertical resolution. The snow stratigraphy was documented in the trench with a series of overlapping near-infrared photographs, and standard manual stratigraphy measurements were made every 20 cm. Snow density, grain size, and hardness were manually measured in each stratigraphic layer. These detailed, high resolution coincident measurements provide a new view of the Arctic snowpack at a resolution previously unavailable, using state- of-the-art tools that are all sensitive to the snow properties that control backscatter as measured by microwave remote sensors.
C12B-06
Multi-scale analysis and synthesis of combined active-passive microwave data over snow covered areas: implications for current and future swe products.
Passive microwave data at 19 and 37 GHz (or similar frequencies) have been historically used to retrieve snow parameters such as snow water equivalent (SWE) and snow depth. Also, studies using active microwave data collected from space for snow applications have been concentrating on the separation between wet and dry snow. Nevertheless, a sensitivity of active data exists to other parameters such as mean grain size, SWE and it has been demonstrated both theoretically and experimentally. The combination of active and passive microwave data for remote sensing of snow offers therefore a strong potential for improving the retrieval of snow parameters. In this talk, we review the sensitivity of both active and passive microwave data to snow depth evolution at large spatial scale and analyze their dynamic range during the snow season. Trends of backscattering coefficients recorded by the SeaWinds scatterometer flying onboard the QuikSCAT satellite and SSM/I and AMSR-E microwave brightness temperatures are analyzed in conjunction with snow depth at different spatial scales. Airborne active and passive microwave data collected during the Cold Land Process Experiment – 1 (CLPX- 1) are co-registered and re-gridded to a resolution of the order of hundreds of meters to study their correlation with snow parameters. These are obtained over the CLPX-1 domain at high spatial resolution (30 m) from a snow evolution modeling system forced with data from meteorological stations and atmospheric analyses grid-points distributed using a micrometeorological model. The outputs are then aggregated to march the spatial resolution of airborne and satellite data. Finally, an electromagnetic model is used to synthesize the active and passive microwave observations.
C12B-07
Implications of the representation of snowpack stratigraphy for large-scale passive microwave remote sensing
Passive microwave satellite observations have been available for more than 30 years and have been used to map snow cover extent and water equivalent globally. The layered character of snowpacks increases the difficulties in deconvolving the return microwave signal, but it also offers the opportunity to infer the metamorphic signature of the snowpack and to extract additional information by combining different frequencies and polarizations. Implementation of this approach requires knowledge of the stratigraphy of snowpack microphysical properties (temperature, density, and grain size), which as a practical matter can only be produced by predictive models. Given the relatively large spatial scales of existing satellite passive microwave sensors (tens of km); this prior information requires large-scale modeling of layered snowpacks, including representation of spatial variability of vegetation and snow characteristics. A multi-layer snow model designed for such applications is described, which is computationally efficient enough to be implemented at the scale of large watersheds, or even over continents. The model's ability to replicate large-scale snowpack layer features and their effect on passive microwave emissivity is evaluated using observations from the Cold Land Processes Experiment (CLPX) and a 2002 Nome-Barrow transect (SNOWSTAR). Anomalous behavior of brightness temperatures is reproduced using the coupled model, and frequency and polarization differences are evaluated. The sensitivity of model predictions to the model layering parameterization at large scales is also examined. Finally, examples of how brightness temperatures at different frequencies reflect snowpack properties at different depths and how this information can be exploited (e.g. through data assimilation) are discussed.
C12B-08
Ice Phenology on Great Bear Lake and Great Slave Lake, Canada, from AMSR-E: Algorithm Development and Analysis
Lake ice phenology encompasses: freeze-up in the autumn; a long period of growth and thickening in winter; a short period of ice melting and thinning, and finally, break-up and the complete disappearance of the ice cover in spring. The sensitivity of ice phenology (freeze-up/break-up dates, and ice cover duration) to climate conditions has been demonstrated in several investigations. Freeze-up/break-up dates have shown to be strongly correlated with air temperatures in the preceding months, usually 1-2 months, leading to the events. Observations of freeze-up/break-up dynamics on large northern lakes are very limited. Moreover, dates of ice formation from optical satellite sensors such as MODIS and AVHRR are difficult to obtain on high-latitude lakes due to long periods of obscurity and extensive cloud cover especially, during the fall/winter freeze-up period. There is therefore significant interest in developing algorithms for the retrieval of ice phenology parameters from microwave satellite imagery. In this paper, polarization difference algorithms from AMSR-E 18.7, 23.5, and 36.5 GHz brightness temperatures are assessed to determine freeze-up and break-up dates on two large lakes in northern Canada, Great Bear Lake (GBL) and Great Slave Lake (GSL). The algorithms are then applied to map and analyze the spatiotemporal variability in ice phenology on the two lakes for ice seasons 2002-2003 to 2006- 2007, and to contrast their ice regime. Preliminary results show a difference of approximately two to three weeks in freeze-up/break-up dates between cold and warm winters. Freeze-up and break-up dates differ by several weeks between GBL and GSL, with the largest difference observed during fall/winter freeze-back. The primary driver for these differences appears to be air temperature. Being located at higher latitude, GBL freezes first.