Chester J. Koblinsky

Modeled Seasonal Variations of Firn Density Induced by Steady State Surface Air Temperature Cycle

Abstract Seasonal variations of firn density in ice-sheet firn layers have been attributed to variations in deposition processes or other processes within the upper firn. A recent high-resolution (mm-scale) density profile, measured along a 181 m core from Antarctica, showed small-scale density variations with a clear seasonal cycle that apparently was not related to seasonal variations in deposition or known near-surface processes (Gerland and others, 1999). A recent model of surface elevation changes (Zwally and Li, in press) produced a seasonal variation in firn densification, and explained the seasonal surface elevation changes observed by satellite radar altimeters. In this study, we apply our one-dimensional time-dependent numerical model of firn densification that includes a temperature-dependent formulation of firn densification based on laboratory measurements of grain growth. The model is driven by a steady-state seasonal surface temperature and a constant accumulation rate appropriate for the measured Antarctic ice core. The modeled seasonal variations in firn density show that the layers of snow deposited during spring to mid-summer that have the highest temperature history compress to the highest density, and the layers deposited during later summer to autumn that have the lowest temperature history compress to the lowest density. The initial amplitude of the seasonal difference of about 0.13 reduces to about 0.09 in 5 years and asymptotically to 0.0 at greater depth, which is consistent with the core measurements.

Correlation and Trend Studies of the Sea Ice Cover and Surface Temperatures in the Arctic

Abstract Co-registered and continuous satellite data of sea-ice concentrations and surface ice temperatures from 1981 to 2000 are analyzed to evaluate relationships between these two critical climate parameters and what they reveal in tandem about the changing Arctic environment. During the 19 year period, the Arctic ice extent and actual ice area are shown to be declining at a rate of –2.0±0.3% dec –1 and 3.1 ±0.4% dec–1, respectively, while the surface ice temperature has been increasing at 0.4 ±0.2 K dec–1, where dec is decade. The extent and area of the perennial ice cover, estimated from summer minimum values, have been declining at a much faster rate of –6.7±2.4% dec–1 and –8.3±2.4% dec–1, respectively, while the surface ice temperature has been increasing at 0.9 ±0.6K dec–1. This unusual rate of decline is accompanied by a very variable summer ice cover in the 1990s compared to the 1980s, suggesting increases in the fraction of the relatively thin second-year, and hence a thinning in the perennial, ice cover during the last two decades. Yearly anomaly maps show that the ice-concentration anomalies are predominantly positive in the 1980s and negative in the 1990s, while surface temperature anomalies were mainly negative in the 1980s and positive in the 1990s. The yearly ice-concentration and surface temperature anomalies are highly correlated, indicating a strong link especially in the seasonal region and around the periphery of the perennial ice cover. The surface temperature anomalies also reveal the spatial scope of each warming (or cooling) phenomenon that usually extends beyond the boundaries of the sea-ice cover.

Modeling of Ice Flow and Internal Layers Along a Flow Line Through Swiss Camp in West Greenland

Abstract An anisotropic-ice flowline model is applied to a flowline through Swiss Camp (69.57° N, 49.28° W), West Greenland, to estimate the dates of internal layers detected by radio-echo sounding measurements. The effect of an anisotropic-ice fabric on ice flow is incorporated into the steady-state flowline model. The stress–strain-rate relationship for anisotropic ice is characterized by an enhancement factor based on the laboratory observations of ice deformation under combined compression and shear stresses. By using present-day data of accumulation rate, surface temperature, surface elevation and ice thickness along the flowline as model inputs, a very close agreement is found between the isochrones generated from the model and the observed internal layers with confirmed dates. The results indicate that this part of the Greenland ice sheet is primarily in steady state.

The Potential of Using Landsat 7 Data for the Classification of Sea Ice Surface Conditions During Summer

Abstract During spring and summer, the surface of the Arctic sea-ice cover undergoes rapid changes that greatly affect the surface albedo and significantly impact the further decay of the sea ice. These changes are primarily the development of a wet snow cover and the development of melt ponds. As melt ponds generally do not exceed a couple of meters in diameter, the spatial resolutions of sensors like the Advanced Very High Resolution Radiometer and Moderate Resolution Imaging Spectroradiometer are too coarse for their identification. Landsat 7, on the other hand, has a spatial resolution of 30 m (15 m for the panchromatic band) and thus offers the best chance to map the distribution of melt ponds from space. The different wavelengths (bands) from blue to near-infrared offer the potential to distinguish among different surface conditions. Landsat 7 data for the Baffin Bay region for June 2000 have been analyzed. The analysis shows that different surface conditions, such as wet snow and melt-ponded areas, have different signatures in the individual Landsat bands. Consistent with in situ albedo measurements, melt ponds show up as blueish, whereas dry and wet ice have a white to gray appearance in the Landsat true-color image. These spectral differences enable areas with high fractions of melt ponds to be distinguished.

Postprocessing of satellite altimetry return signals for improved sea surface topography accuracy

Large off-nadir attitude deviations and high surface wave heights cause an alteration in the ocean return signal from a satellite radar altimeter. This leads to an error in the on-board calculation of the height measurement. The error can be removed by reprocessing the full radar return signal on the ground. In the ground processing, the correct tracking point in the return signal is recomputed through a procedure called retracking. There has been a controversy over whether or not all altimeter data would be retracked. This study analyzes retracked southern ocean data from the first 34 repeat cycles of the Geosat Exact Repeat Mission (ERM), covering November 1986 through April 1988. The final data set consists of over 2.5 million smoothed one-per-second measurements of the ocean surface. The significant wave height (SWH) distribution as given on the NOAA geophysical data records (GDRs) for these measurements peaks at around 2.1 m (19% of the measurements) and drops down almost linearly to 2% of the measurements at 5.8 m. There are over 1100 observations with SWH greater than 15 m. The difference between the surface heights calculated from the retracked data and the original on-board estimates is less than 10 cm for SWH less than 10 m but increases to approximately 1.0 m at a SWH of 18 m. In general, the electromagnetic (EM) bias coefficient calculated using the retracked data is slightly less than that using the unretracked data and does not decrease as much with SWH as do the EM bias coefficients calculated from the unretracked data. A map of the sea surface height variability of the southern ocean created using the retracked data shows differences from variability maps created using the unretracked data in regions of high wave heights. Retracking can be done efficiently on modern UNIX work stations at 0.064 times real-time acquisition. This study shows that retracking will improve altimeter precision.

A global mean sea surface based upon GEOS 3 and Seasat altimeter data

A mean sea surface relative to the International Union of Geodesy 1980 Geodetic Reference System reference ellipsoid has been derived from Seasat and GEOS 3 altimeter measurements. This surface, called MSS-9012, has been computed on a grid of 1/8° resolution. Each elevation value was calculated by fitting all data within 111 km to a local biquadratic surface using Bayesian least squares. Individual data points were weighted inversely to the square of the distance to the grid location in the gridding process. The surface covers the global ocean between 70°N and 72°S. In the vicinity of sea ice the altimeter heights have been corrected for the on-board tracker error that occurs over non-Gaussian surfaces. Comparisons are made between MSS-9012 and ocean bathymetry. The eastern extent of the Chain Fracture Zone in the Gulf of Guinea is more apparent in the altimetry than in the bathymetry data, as are many other features. The combination of data from the two satellites has successfully retrieved more information about the sea surface than was previously possible using only Seasat data.

Interannual Variations of Shallow Firn Temperature at Greenland Summit

Abstract Firn-temperature profiles are calculated in a thermal model using continuous surface temperatures derived from automatic weather station data and passive-microwave data in the Greenland summit region during the period 1987–99. the results show that significant interannual variations of mean summer (June–August) and annual temperatures occur in the top 15 m, in addition to the normal seasonal cycle of firn temperature. At 5 m depth, the seasonal cycle is damped to 13% of the surface seasonal range, but even at 15m about 1% or 0.6˚C of the seasonal cycle persists. Both summer and mean annual temperatures decrease from 1987 to 1992, followed by a general increasing trend. Interannual variability is 5˚C at the surface, but is dampened to 3.2˚C at 5 m depth and 0.7˚C at 15 m depth. Dampening of the interannual variability with depth is slower than dampening of the seasonal cycle, because of the longer time constant of the interannual variation. the warmer spring and summer temperatures experienced in the top 5 m, due to both the seasonal cycle and interannual variations, affect the rate of firn densification, which is non-linearly dependent on temperature. During the 12 year period 1987–99, the annual mean surface temperature is –29.2˚C, and the annual mean 15 m temperature is –30.1˚C, which is >1˚C warmer than a 15 mborehole temperature representing the period around 1959 and warmer than the best-fit temperature history by Alley and Koci (1990) back to AD 1500.