C23A-0927
Sea Ice Characteristics from the West Antarctic ODEN Cruise Compared to Satellite Analyses
In this study, AMSR-E's (The Advanced Microwave Scanning Radiometer - Earth Observing System) geophysical product for ice concentration and ice edge location was compared to ice observations made aboard ship in the Antarctic pack ice during ODEN expedition (Swedish icebreaker) sponsored by United States National Science Foundation (NSF) as of the first collaborative activities of the International Polar Year (IPY) 2007 - 2008. Ice observations consist of ASPECT (Antarctic Sea Ice Process and Climate) protocol that provides a standardized and quantifiable method for observing sea ice that is now accepted as the international standard. Estimates of ice concentration, ice type with its thickness, floe size, and snow type with it is thicknesses are conducted along with meteorological observations. Observations involved making half an hour observations from the ship's bridge within a radius of approximately 1 km of the ship. Our observations made along 67° 53 S latitude and 102° 97 W longitude and along 73° 52 S latitude and 178° 54 E longitude in December 2006. Different ice types and temperature conditions were encountered during the observations. Sea ice concentration varied between 10% and 90%. During the expedition the minimum and maximum ice thickness varied between 10 cm and 3m. The results from the observations analyzed using a GIS platform and provided detailed data for comparison with satellite products. AMSR-E microwave remote sensing data contains brightness temperatures (TBs) at 6.9 GHz, 10.7 GHz, 18.7 GHz, 23.8 GHz, 36.5 GHz, and 89.0 GHz. and uses various algorithms to estimate the snow depth, sea ice concentration, and sea ice temperature. The study region for the comparison includes the West Antarctic sea ice zone which is particularly interesting since the sea ice of the area is highly sensitive to global climate change.
C23A-0928
An Assessment Of The Correlation Between Radar Backscatter And Statistical Parameters Of Snow Microstructure
Snow microstructure irregularities are one of the numerous factors that affect radar backscatter. In this work, the correlation is retrieved between the volume component of radar backscatter and the statistical parameters of snow microstructure. This study uses the "slice" approach: snow particles close to the front of incident radar wave are considered to be approximately at the same distance from the radar. Therefore, these particles reflect incident electromagnetic wave coherently. Each slice is much narrower than the radar wavelength in the wave propagation direction. In the transverse plane the slice has spherical configuration (in accordance with the spherical wave front) within the cross section of the backscatter volume, which, in turn, is formed by main lobe of the antenna pattern. The snow radar equation, which is based on the slice model of backscatter volume, is derived. The equation includes contributions to the backscatter from uncorrelated small-scale fluctuations of snow reflectivity. It is shown that the average power of an echo signal is proportional to radar reflectivity, i.e. to the sixth moment of the snow particles' size distribution function, for the case of Poisson fluctuation of particle concentration within the slices and uncorrelated small-scale snow reflectivity fluctuations within a backscatter volume. In this case the known semi-empirical model of volume backscatter is valid. Overall, the volume component of the backscatter coefficient depends on the first three moments of the particle size distribution function. For completely homogeneous microstructure the volume component of return signal is zero. Preliminary estimates of microstructure parameters of dry snow zone of the Greenland Ice Sheet, which can cause the observable very low radar backscatter, is presented.
C23A-0929
Ground-Based Radar Measurements of the Northern Colorado Snowpack at CLPX- II
A stationary, laboratory-grade network analyzer (NA)-based (stepped frequency) radar system and a mobile FMCW (frequency-modulated continuous wave) radar instrument both acquired measurements of the Northern Colorado snowpack during NASA's Cold Land Processes Experiment (CLPX-II, February 2007). These ground- based radar measurements were complimented by manual snow samples and detailed scientific snow pit data. This study concentrates on the preliminary comparison of the two ground-based radar systems while incorporating the associated manual measurements (e.g. stratigraphy, temperature/density profile, grain size estimation, etc.) into the interpretation, on three separate days during Intensive Observation Period 3 of CLPX-II. The stationary radar is based on an Agilent® PNA series vector network analyzer (N5230A) with a 10 MHz to 20 GHz frequency range. A configurable test set was used which allowed for dual linear polarization combinations (HH, VV, HV and VH) to be acquired with the same antenna configuration. Dual-polarized 2-18 GHz horns were connected directly to the NA test set with high phase stability cables at a height of approximately 2 meters, within the far field and with the ability to adjust the incidence angle from nadir to horizontal. Measurements were made from a tripod, with a boom sweeping over an arc of approximately 3 meters. During the same field campaign, a mobile FMCW radar acquired measurements from 4-18 GHz, at HH/HV and VV/VH polarizations, at incidence angles of 30 and 45 degrees, and at a height of 2.3 meters (far-field). An additional portable radar was mounted at a height of 50 cm and 0 degrees incidence, and can be used to estimate snow depth, stratigraphy, and SWE. UNAVCO provided precision differential GPS equipment (Trimble 5700/R7, base station and rover) for the mobile FMCW radar, and the radar measurement and control software was adapted to sync with the cm-level positions, which were recorded every second. These measurements cover a wide range of sensor parameters; and in addition, the variation of in-situ snowpack properties within these radar footprints are less complicated to characterize, in contrast with air- and space-borne radar footprints. Manual measurements of snow depth, stratigraphy, and snow water equivalent (SWE) were made frequently throughout radar surveys and are also available for comparison. Future research will incorporate LiDAR and Jet Propulsion Laboratories' PolSCAT airborne scatterometer surveys that were flown during CLPX-II. These ground based measurements may be used to simulate the PolSCAT/QuickSCAT equivalent backscatter at 13-14 GHz, in addition to other frequency ranges not currently flown by airborne or satellite-based systems.
C23A-0930
A handheld device for measuring snow grain size at centimeter resolution using IR- and visible- spectrum sensors
The estimation of snow grain size from multiple-wavelength satellite imagery is well established as a glaciological technique. These techniques use the fact that large-grained snow absorbs strongly in the infrared to obtain grain estimates based on the difference between reflectance of sunlight in the visible spectrum to that in the infrared. Similar techniques hold promise for estimating grain size variations in firn core, in snow-pits and in boreholes. However, reflectance measurements with small light sources and detectors are complicated by the scattering behavior of snow, in which visible-wavelength photons propagate for much longer distances than infrared photons before being absorbed. This means that density variations as well as grain size variations can change the relative reflectances in the visible and infrared spectra. We present modeling and experimental results in the development of a handheld instrument that will allow grain- size measurements to be made at centimeter resolution. This instrument uses multiple visible-light detectors to estimate the spreading distance of visible photons. This gives an estimate of the scattering density, which in turn allows us to correct the infrared-visible spectral differences, giving a robust estimate of grain size. We present monte-carlo modeling of the measurement, validation experiments using laboratory standards, and measurements on firn cores.
C23A-0931
Examination of Snow Mapping Methods Over the Weber River Basin, Utah Using MODIS Observations
The MODIS instruments on the Terra and Aqua spacecraft provide snow cover observations on a daily basis when cloud cover permits over the globe as well as regional and local areas. This study reports some ongoing work using Terra MODIS snow cover observations centered on the Weber River Basin. The basin covers 2500 square miles/6400 square kilometers within the Great Salt Lake Basin in Utah. This report is focused on comparing various snow mapping methods in observing snow cover in this basin during the first six months (January to June) of 2001. The methods compared are the standard 8-day SNOWMAP results provided operationally by MODIS (designated MOD10A2 on lists of MODIS products), the fractional snow cover product (MOD10L2) for specific days, and the spectral end-member approach using three end-members:snow, soil and rock, and vegetation. The results show the expected variation of snow cover with time over the Weber Basin, but with distinct differences in the extent of the snow cover ranging from the SNOWMAP typically showing the largest snow cover due to the 8-day product giving maximum snow cover and the spectral end-member approach showing the least. Comparisons with more detailed imagery tends to indicate that, as expected, the spectral end-member approach appears to be the most accurate. In non-snow covered regions of the Weber River Basin there are anomalies (e.g. snow appearing where there is obviously no snow) and these anomalies are under investigation.
C23A-0932
Evaluation of Passive Microwave Sea Ice Concentration in Arctic Summer and Antarctic Spring by Using KOMPSAT-1 EOC
Spaceborne passive microwave sensors such as the Special Sensor Microwave/Imager (SSM/I) and the Advanced Microwave Scanning Radiometer – Earth Observing System (AMSR-E) have been observing polar sea ice concentration (SIC) playing important roles in the global climatic and environmental studies. To evaluate the passive microwave SIC we observed sea ice with the 6 m-resolution, Electro-Optical Camera (EOC) sensor onboard KOMPSAT-1 satellite. A total of 72 cloud-free EOC images, 18 km x 18 km each, of arctic sea ice edges were obtained from July to August and 68 images across the antarctic continental edges from September to November, 2005. We classified arctic sea ice types into land-fast ice, pack ice, and drift ice according to ice distribution and movement characteristics, and compared with SSM/I SIC calculated from NASA Team (NT) algorithm. The correlations between EOC and SSM/I SICs were related to the spatiotemporal stability of sea ice in arctic summer showing high correlation for stable land-fast ice and low for less stable pack ice and drift ice. In case of pack ice, SSM/I SIC were lower than EOC SIC by 19.63% in average due to the underestimation problem of NT algorithm for ice ridge and new ice. For drift ice, SSM/I SIC showed 20.17% higher than EOC SIC in average partly due to the wider IFOV of SSM/I than EOC swath resulting in the insertion of pack ice nearby and partly due to the condition of wet snow on drift ice causing overestimation in NT algorithm. In the Antarctic spring, sea ice types in the EOC images, mostly of land fast ice or pack ice, were classified into white ice (W), grey ice (G), and dark-grey ice (D) and then compared with SSM/I NT and AMSR-E NT2 SICs. EOC SIC of W and G, excluding D, showed best fit to SSM/I NT SIC suggesting that the SSM/I NT algorithm responds to young ice in addition to multi-year ice and first-year ice while AMSR-E SIC responds to all types of sea ice.
C23A-0933
New Methods for Estimating Sea-Level Rise Contributions From Svalbard Glaciers
Quantifying glacier contribution to sea level rise is difficult due largely to a paucity of long-term mass balance observations. For example, of more than 160,000 glaciers worldwide only about 40 have mass balance records longer than 20 years. Since 2003, the SLICES project has been developing novel methods for significantly increasing the distribution and spatial/temporal resolution of mass balance records. Because of their sensitivity to changes in climate, we have focused on the glaciers of the Svalbard archipelago. Small mountain glaciers like those in Svalbard constitute only about 3 percent of the glacierized area on Earth. However, of all the world's ice masses, they are currently thought to be contributing most to eustatic sea level rise. We use archived historical stereo photography to create a time-series of high-resolution digital elevation models (DEMs) which are controlled using contemporary lidar data sets. The large number of ground control points that can be extracted from the lidar means the quality of the resultant photogrammetric DEMs can approach those collected using traditional ground-based control. This method has yielded high-quality, high-resolution topographic data for a well distributed sample of Svalbard glaciers dating back to the since the early 1960s. These data provide the first long-term mass balance record for Svalbard of this quality, spatial resolution and areal distribution and includes glaciers on the more remote eastern side of the archipelago. Results have demonstrated that glaciers in the west of the archipelago experienced a clear acceleration in thinning over the entire period, which is consistent with long-term mass balance records. In the east of the archipelago, which is characterised by a cooler climate, thinning has occurred but the rate has been lower. Under the new GLIMPSE project, the methods presented are now being applied to improve our understanding of the thinning of outlet glaciers in southeast Greenland. http://geography.swan.ac.uk/glaciology/slices/
C23A-0934
Spatio-temporal Variability of Snow Cover in Svalbard Derived From a Combination of Spaceborne Scatterometer, Glaciological Ground Truth and Meteorological Data
Evidence for ongoing large changes in the Arctic climate has been accumulating during the last decade, particularly with regards to sea ice, permafrost and glacier mass balance. Due to the effect of global warming dramatic changes in the Arctic snow and ice coverage are currently observed, expressed by a reduction of 10% over the last 30 years associated with an extended and longer lasting melting season. Previous studies have demonstrated the capacities of active microwave instruments for the detection of surface melt and freeze-up due to the high sensitivity of radar backscatter to snow wetness. Spaceborne scatterometers provide data at low spatial but high temporal resolution allowing consistent observations on a daily time scale. This study focuses on the Svalbard region characterized by a highly variable climate throughout the year due to its position within a zone that includes both the polar ocean and atmospheric fronts between the Arctic Ocean, Nordic Seas and Barents Sea. Nevertheless, over a long enough period of time we expect to see a general climatic trend by monitoring the spatio-temporal variability of snow distribution and melt all over Svalbard. For this we utilize microwave backscattering measurements continuously carried out by the Ku-band scatterometer QSCAT since fall 1999. Furthermore, we tested retrieval algorithms for the assessment of single snow fall events and the amount of snow accumulation during winter. A set of glaciological ground-truth data serves us for interpretation of the backscattering signatures, i.e. snow accumulation measurements by means of snow probing, snow pit sampling, drilling of shallow ice cores as well as ground penetrating radar measurements. In addition, meteorological data from weather stations around Svalbard are available to investigate links between atmospheric circulation, moisture transport, near surface air temperature and corresponding deposition and melting of snow, respectively. The analysis demonstrates that coarse resolution scatterometer data can be usefully applied to trace consequences of a warming climate in Svalbard.
C23A-0935
MODIS albedo and regional mass balance of high Arctic glaciers
MODIS albedo products are compared to glacier mass balance measured on eight glaciers in the high Arctic, in northwestern Svalbard. The glaciers range in size from ca. 5-500 km2. We use MODIS L3 albedo products (MOD43B3), which have a nominal resolution of 1 km, and for which data cover the spring to autumn months of 2000-2006. We compare the albedo data to four glaciers for which there are mass balance measurements in each of the years in the study period (Austre Br\oggerbreen, Midtre Lov\´{e}nbreen, Kongsvegen, and Waldemarbreen), as well as four glaciers with some years of mass balance measurements (Holtedahlfonna, Irenebreen, and Elisebreen). While six of the glaciers are only represented by a few MODIS pixels, there is still good correlation between parameters derived from annual albedo minima and the equilibrium line altitude (ELA) of the study glaciers. However, there is no simple threshold method which can be used to directly extract the ELA from the albedo data.
C23A-0936
Satellite Interferometry and Antarctic Ice Thickness Estimation
In regions of sparse measured ice thickness data the BEDMAP dataset can be vastly inaccurate (up to a factor three). This study estimates ice thickness using measured velocity data in an attempt to assess their potential for future use in sparse data regions of BEDMAP. Four methods, all using satellite interferometry ice surface velocity data and based on conservation of mass, have been employed to produce estimates over approximately 65 % of the Antarctic ice sheet. Methods range from simple point approximations to two dimensional velocity divergence methods. Other datasets, such as balance flux and accumulation, were also used in the calculations. Estimates of ice thickness were compared to the measured ice thickness input data from BEDMAP to assess their accuracy. Another study was performed to aid in the explanation of the contribution of errors associated with the balance flux and accumulation datasets. This study involved estimating fluxes into box areas using ice thickness measurements from the BEDMAP dataset, the surface velocity data and accumulation datasets. Fluxes were then compared to balance flux estimates calculated for the box areas and also converted to give elevation change estimates. Estimates of elevation change were compared with elevation change data from satellite radar altimetry. The methods produced estimates of varying accuracy and ultimately it is the conclusion of the study that inaccuracies in current accumulation datasets are too large to allow accurate estimation of ice thickness.
C23A-0937
Towards predicting streamflow based on SWE, melt timing, and topography in subarctic heterogeneous terrain
Snowmelt onset date and snow water equivalent (SWE) are major factors that influence the spring runoff in high latitude, snow dominated basins. We combine AMSR-E L3 daily SWE from March to June 2003-2006, daily hydrological records from 3 sites on the Pelly and Ross Rivers, Yukon Territory [Pelly Crossing N62.82°, W136.58°, Faro N62.22°,W133.37°, and Ross River N61.99°,W132.37°], and a 1:250,000 DEM to develop a technique to predict streamflow in subarctic heterogeneous terrain. The AMSR-E L3 SWE algorithm was developed for global snow cover distributions; it is not optimized for heterogeneous terrain. Field data suggest that it underestimates the SWE in this area. We assume it represents the minimum SWE per pixel. SWE variations of the Pelly River basin (49,000 km2) and its two nested sub-basins (22,100 km2 and 7,250 km2), show that SWE had an apparent drop shortly after the snowmelt onset date determined from Tb and diurnal amplitude variations (DAV), which are also correlated with temperature change. During the early stage of snowmelt, high and low elevations have no significant SWE difference. After mid-April, the most intense melt period at lower elevations, low elevation SWE drops far below high elevation SWE, which is just beginning the melt process. Initial melt and the drop in low elevation SWE likely cause the first small discharge peak in the hydrograph. When the SWE throughout the basin approaches 0 mm for more than 3 days, it is followed by the peak flow. The largest basin has an about 14 day lag between the SWE drop and the flow increase, while the smaller basins have an about 9 day lag. Snow distribution, melt, runoff, and the lag times vary due to diverse terrain and microclimate factors such as: forest cover, permafrost, temperature and precipitation. By combining topography, snow distribution and melt timing, we have developed an understanding of basin-specific stream discharge response to spring thaw. Passive microwave derived daily SWE data combined with terrain and melt timing have significant potential for constraining and predicting stream flow timing and magnitude during the melt season in subarctic regions.
C23A-0938
The Third Generation of the Interactive Multisensor Snow and Ice Mapping System (IMS V3)
Since its inception in 1998, the NOAA Interactive Multisensor Snow and Ice Mapping System (IMS) has sought to provide the most timely, accurate and reliable snow and ice cover gridded data to a broad user community. The output of this product has improved incrementally in resolution, timeliness, and accuracy, with the most notable upgrade of IMS Version 2 (V2) in February 2004. New output requirements, the need for metadata, satellite advancements, and the desire for a broader suite of snow and ice variables mandates another significant overhaul in the IMS system requirements. The intent of the IMS Version 3 (V3) is to improve the output accuracy, timeliness, resolution, metadata, cost, and add flexibility for future enhancements. The IMS production is scheduled for an operational transition to the National Ice Center (NIC) in March 2008. The NIC currently uses a vector-based GIS production system for sea ice charting, which is in contrast to the current raster-based production of the IMS. Integration of the IMS into a vector-based system should allow many of the previously mentioned improvements in the output, while integrating NIC systems. While production will be vector-based, output products will include a more diverse yet integrated collection of raster products. Raster products remain the preferred input source for weather and climate models which are the primary users of the IMS. While this alteration from raster-to-vector based analysis will likely provide significant improvements, care must be taken in the transition from one production methodology to another to preserve its benefit to numerical weather prediction and climate monitoring. http://www.natice.noaa.gov
C23A-0939
Empirical Retrieval of Surface Melt Intensity using coupled MODIS Optical and Thermal Measurements over the Greenland Ice Sheet
An optical-thermal feature space partitioned as a function of melt intensity was derived using a one-dimensional thermal snowmelt model (SNTHERM). Cloud-free dates of June 10, July 5, 7, 9, and 11, 2001 MODIS daily reflectance Band 5 (1.230-1.250um) and surface temperature images rescaled to 1km over western Greenland were used in a retrieval algorithm. SNTHERM was forced by hourly meteorological data from the Greenland Climate Network (GC-Net) at reference sites spanning dry, percolation, and wet snow facies in the Jakobshavn drainage basin. Liquid water fractions (LWF) were derived for satellite composite periods covering May, June, and July displaying low fractions (0-1%) at elevations greater than 2500m and fractions at or greater than 15% at elevations lower than 1000m assessed for only the upper 5 cm of the snow surface. Comparison of LWF to mapped dry and wet zones determined from Quickscat backscatter indicated higher intensities distributed in wet snow zones, while lower intensities were grouped in dry zones.
C23A-0940
Ice, Cloud, and land Elevation (ICESat) Satellite Digital Elevation Models (DEMs) of Greenland and Antarctica at NSIDC
The Geoscience Laser Altimeter System (GLAS) instrument aboard the Ice, Cloud, and land Elevation (ICESat) satellite launched on 12 January 2003. The primary objective of the ICESat mission is to provide global measurements of polar ice sheet elevation to discern changes in ice volume and ice sheet mass balance over time. Secondary objectives of the mission are to measure sea ice thickness, cloud and atmospheric properties, land topography, vegetation canopy heights, ocean surface topography, and surface reflectivity. The GLAS instrument has three lasers, each of which has a 1064 nm laser channel for surface altimetry and dense cloud heights, and a 532 nm lidar channel for the vertical distribution of clouds and aerosols. Here we present information about 2 DEMs now available at the National Snow and Ice Data Center, one for Antarctica and one for Greenland. The ICESat/GLAS Antarctic DEM has a 500m grid spacing and the Greenland one has 1km grid spacing. They are both derived solely from GLAS/ICESat laser altimetry profile data from the first seven operational periods (from February 2003 through June 2005) of the GLAS instrument. The ICESat Antarctic DEM provides unprecedented detail for regions between 81.5 S and 86.0 S, because of the higher orbital inclination of ICESat relative to past satellite altimeters. However, some artifacts remain in these early DEMs as a result of gridding, cloud effects, and the relatively large spacing between orbit track lines at lower latitudes (e.g., > 20km at 70 N and S) We will discuss the strengths and weaknesses of these laser altimetry derived DEMs.
C23A-0941
Ice Penetrating Radar Surveys in the Fourcade Glacier of King George Island, Antarctica
To determine subglacial topography and internal features of the Fourcade Glacier of King George Island in Antarctica, helicopter-borne and ground-towed GPR data were recorded along four profiles in November 2006. Signature deconvolution, F-K migration velocity analysis, and finite difference depth migration applied to the mixed-phase, single-channel, ground-towed data, were effective in increasing vertical resolution, obtaining the velocity function, and yielding clear depth images, respectively. The radar sections show rugged subglacial topography, englacial gliding surfaces, and localized scattering noise. The maximum depth to the basement is over 79 m in the subglacial valley adjacent to the southeastern slope of the divide ridge between Fourcade and Moczydlowski Glaciers. In the ground-towed profile, we interpret a complicated conduit above possible basal water and other isolated cavities, which are a few meters wide. Near the terminus, the GPR profiles image gliding surfaces, fractures, and faults, which will contribute to the tidewater calving mechanism forming icebergs in Potter Cove. The ice cliff at the terminus of the outlet glacier has retreated up to 100 m during the last two years.
C23A-0942
Improved Snow Cover Retrievals from Satellite Passive Microwave Data over the Tibet Plateau: The need for atmospheric corrections over high elevations
During the past four decades much important information on continental to hemispheric scale snow extent and variability has been provided by satellite remote sensing using both optical and microwave data. The Tibet Plateau is the only large geographic region in the Northern Hemisphere where the microwave retrievals tend to consistently disagree with snow cover climatologies derived from optical data. Initial speculation as to the cause of this significant overestimate by microwave retrievals has included the following: 1) soil type 2) frozen soil 3) widespread presence of frozen lakes. Our analysis has eliminated these explanations and we present an alternative explanation based on the influence of the atmosphere on the microwave retrievals. Numerous microwave snow cover algorithms have been developed with approaches ranging from the theoretical to the purely empirical. However, in most all cases, when these algorithms are actually applied, they are tuned or calibrated to return snow extents that match well with the optical satellite data over a particular study area. Therefore, although the algorithms might have been initially developed using brightness temperatures measured at ground level or from aircraft, they have been adjusted to provide the most accurate results possible when applying brightness temperatures measured from satellite. This simply means that these algorithms have implicitly accounted for the presence of an atmosphere because the surface or scene brightness temperature values applied in the algorithms have actually passed through the atmosphere along their path to reach the satellite sensor. Therefore, given that an algorithm is tuned to return favorable results across a relatively standard atmospheric thickness between the ground surface and the satellite, a potential problem arises when the same algorithm is applied to an extremely high elevation target where the atmosphere thickness is greatly reduced, such as on the Tibet Plateau. Using radiosonde data from both the more typical lower elevations and the exceptional Tibet Plateau, combined with a radiative transfer model, we derive the coefficients required to make the necessary corrections to retrievals over land surfaces at high elevations.
C23A-0943
Assessment of ICESat Repeat Track Estimation Techniques for Polar Elevation Change Detection
Crossover methods have been a standard for measuring polar elevation change, but they do not provide dense enough coverage for some applications. As repeat track methods use all of the available data from a given instrument, they provide denser coverage (by about an order of magnitude for ICESat), but include topographic slope errors due to cross-track spacing between tracks from different times. In this presentation we assess the accuracy of a least squares approach to repeat-track elevation change estimation where no data interpolation is required. ICESat elevation data over Antarctica and Greenland from ten laser acquisition periods spanning 2003- 2007 are employed. We present a survey of different parameter-set estimation methods, including a variety of multiple-parameter and weighting techniques. We assess the relative quality of these repeat track results by comparing to crossover results and to ICESat elevation profiles during the same time periods. Relative impacts on different spatial scales are examined, from ice stream, to ice sheet, to continental-scale areas, including an estimation error assessment for equivalent water gain/loss.
C23A-0944
Estimating snowfall patterns using timeseries of remote sensing images within a Bayesian framework
Snow water equivalent (SWE) reconstruction methods have been used previously to characterize seasonal SWE accumulation using mass and energy balance models. Recognizing that the spatial signature of the seasonal SWE accumulation is an integration of a series of snowfall events, we have formulated a Bayesian SWE reconstruction which utilizes the ensemble Kalman smoother (EnKS) to combine timeseries of remote sensing estimates of snow covered area (SCA) with a land surface model (LSM) to estimate snowfall distribution. An ensemble-based snow depletion curve (SDC) is used to relate SCA and SWE. We perform a series of synthetic tests to assess how much information concerning snowfall accumulation patterns can be extracted from a timeseries of SCA measurements during the ablation season. The test is performed using vegetation and meteorologic data at the 625 km2 Colorado Rabbit Ears pass area studied during the NASA Cold Lands Processes Experiment. We perform experiments to examine sensitivity to a range of physiographic variables (e.g. vegetation cover, magnitude of SWE accumulation, and fraction of total accumulation falling during the ablation season). Sensitivity to over- and underestimation of melt flux, measurement error, and error in the sub-grid precipitation coefficient of variation used to define the LSM SDC are also investigated. Predictions are made about the accuracy of the EnKS posterior SWE estimates (and, thus, the potential usefulness of the Bayesian reconstruction) under a variety of physiographic and uncertainty scenarios.