Spaceborne scatterometers have provided continuous synoptic microwave coverage of the Earth for nearly a decade. Though these scatterometers were originally designed to measure oceanic surface winds, their data are also extremely useful in a broad range of ice and land applications, including the use of extensive scatterometer time series to determine seasonal and interannual variability and possible relationships to climate change. Under a NASA Earth Science Enterprise grant, the Scatterometer Climate Record Pathfinder (SCP) project has produced non-ocean scatterometer imagery and data products that are now publicly available for the first time (http://www.scp.byu.edu/). To date, four spaceborne scatterometers have flown on five different spacecraft (Table 1). The SCP project is providing imagery from the three NASA scatterometers in both a unique, enhanced resolution format and an intrinsic resolution format. In addition, new value-added data products from ESCAT are also available.

Table 1. Characteristics of four spaceborne scatterometers flown on Seasat (SASS), ERS-1/2 (ESCAT), ADEOS (NSCAT), and QuikSCAT (SeaWinds or QSCAT).

A scatterometer transmits radar pulses and receives backscattered energy, the intensity of which depends on the roughness and dielectric properties of a particular target. For ice, snow, soil, and vegetation, roughness properties and geometry that affect backscatter include surface roughness, moisture content, leaf size and density, branch orientation, and preferential alignment of surface scatterers. Dielectric properties are affected by physical characteristics of the effective scattering medium or layer—for example, snow grain size, brine concentration in sea ice, and canopy leaf density—as well as by the phase state of water (meltwater on sea ice and land ice, re-frozen percolated melt water in glacial ice, and whether trees are frozen or actively respiring). As scatterometers can be very accurately calibrated to generally to less than a few tenths of a decibel (dB), seasonal and interannual differences that result in changes as low as 1-2 dB may be confidently examined.

The various scatterometer configurations provide additional means for surface discrimination (Table 1). These include the use of different sensor frequencies to isolate surface types, the relative backscatter response between two polarizations (for example, to separate sea ice from open water), the varying backscatter gradient of different surfaces over a range of incidence angles, and the azimuthal response for features with preferred orientations. In general, the sensitivity to surface roughness increases with higher frequencies. The wide swath of scatterometers provides near-daily global coverage, particularly in the polar regions, at intrinsic resolutions generally between 25-50 km, over incidence angles ranging from 20-55^o. We note the general trend toward finer resolution with newer sensors.

The European Space Agency’s scatterometer, ESCAT, provides the longest scatterometer record. As such, it is invaluable for climate-related investigations. However, due to recent spacecraft attitude and orbital control problems, ESCAT data have been unavailable since January 17, 2001. It is expected that future planned corrections to the ERS-2 spacecraft control will eventually enable continued ESCAT operations later this year. The use of ESCAT in conjunction with the different frequencies of the NASA scatterometer (NSCAT) and QuikSCAT (QSCAT) provides improved discrimination of scattering surfaces, and hence, better understanding. The Seasat-A scatterometer system (SASS), from 1978 data, provides a unique, albeit brief, historical data set that can be compared with the remaining scatterometers to study decadal changes. A recent special issue on scatterometer applications includes 10 papers on ice and land studies, nearly all of which examine long-term variability and the possible relationship to climate change [Drinkwater and Lin, 2000: hereafter referred to as TGARS2000].

Polar Ice

Scatterometer polar imaging applications were first proposed using Seasat SASS followed by ERS wind scatterometer data and have been more recently summarized in the context of NSCAT [Long and Drinkwater, 1999]. The daily global coverage of the scatterometer in the polar regions and its ability to discriminate sea ice, ice sheets, and icebergs, despite the poor solar illumination and frequent cloud cover of the polar regions, make it an excellent instrument for large-scale systematic observations of polar ice (Figure 1a).

Fig. 1. a) This QSCAT image shows Antarctica and surrounding sea ice cover in July 1999. Iceberg B10A (50 km x 100 km) is identified in Drake Passage and eventually melted near South Georgia Island in March 2000. Iceberg B10 broke off Thwaites Glacier in 1992 and split into two in June 1995. The complicated backscatter over the continent is related to ice and snow characteristics, surface and subsurface topography, katabatic winds, and melt zones. The variations in sea-ice returns are due to the snow cover, thickness, and history of the ice since formation. b) The QSCAT ice product of the Barents and East Greenland Seas is shown for April 5, 2001. Overlaid in red is the 100% ice concentration line derived from QSCAT and the southern limit of all known ice line, produced through analysis of visible, infrared, and SAR data by the NIC. c) A prototype product shows the NIC’s analyzed ice edge (green is 100 % ice concentration), QSCAT-derived ice edge (light blue), and ocean surface wind vectors from April 5, 2001. The wind vectors are used to forecast expected ice edge movement over a 24-hour period (cyan).

Ice Sheet Applications. Studies of the great ice sheets of Greenland and Antarctica take advantage of the sensitivity of backscatter to the density and size of snow grains in the various ice facies, especially in the percolation zone and during summer melt. This approach provides a means of examining long-term variability over the ice sheets, particularly with ESCAT, including the extent of the seasonal snow melt zone over Greenland and Antarctica (see articles by Wismann on Greenland and Bingham and Drinkwater in TGARS2000). Estimates of the changes in accumulation rates over ice sheets have been made over Greenland [Drinkwater et al., 2001] from combined scatterometer data from multiple sensors. Finally, the azimuthal modulation that is obtained by the suite of scatterometer antennas designed to derive wind direction was found to correlate with directional snow surface features that align with Antarctic katabatic winds (see article by Long and Drinkwater in TGARS2000).

The onset of seasonal snow melt and freeze-up provides a significant contrast to winter, when conditions over sea ice in the polar regions are colder. Melt onset defines a significant transition in the radiative budget of the ice-covered region, while autumn freeze-up is a key period during which the residual fraction and characteristics of perennial sea ice can be assessed and preconditioning for winter sea ice dynamics is established via the distribution and orientation of newly forming ice in leads. Most recently, algorithmic approaches using scatterometer data to determine melt onset and freeze-up have been examined, including in the Antarctic over a several year period (see article by Drinkwater and Liu in TGARS2000).

Additional studies of sea ice include the derivation of ice velocity fields and ice-type classification. Ice velocity fields are important for estimating heat flux between the ocean and atmosphere, as well as the sea ice mass balance through estimates of ice deformation and growth. Motion fields have been derived using scatterometer data with algorithms based on wavelet analysis [Liu et al., 1999] and cross-correlation with feature tracking [Long and Drinkwater, 1999]. The determination of ice type is another means of estimating mass balance, in addition to being important for navigation. Several recent ice type studies have been undertaken that often use combined data from one or more scatterometers and other sensors to isolate scattering mechanisms for different ice types, and thereby improve discrimination [e.g.,. Kwok et al., 1999; see article by Remund et al. in TGARS2000].

Iceberg Tracking. Operationally, the National Ice Center uses QSCAT imagery as a primary source for tracking large icebergs in the Southern Ocean. For example, in 1999, iceberg B-10A—which was 38 km x 77 km in diameter—drifted north out of the pack ice and into the shipping lanes of the Drake Passage (Figure 1a). QSCAT was used to track this iceberg from July 1999 until its deterioration in March 2000.

Terrestrial Biosphere Applications

Scatterometer data have also been applied to land studies, making use primarily of changes related to moisture content over both vegetated and bare soil, as well as the seasonal freeze-thaw cycle. Several studies have used the extensive ESCAT time series to examine the derivation of monthly indexes of soil moisture over western Africa (see article by Wagner and Scipal in TGARS2000), seasonal trends in soil moisture content in Spain (see article by Woodhouse and Hoekman in TGARS2000), and the seasonal variability of backscatter over different types of vegetation and land surface covers (see articles by Abdel-Messeh and Quegan, and Frison et al. in TGARS2000). Other studies have used scatterometer data to examine the freeze-thaw cycle of boreal forests [Kimball et al., 2001; see article by Wismann on Siberia in TGARS2000].

Scatterometer data can provide valuable information about the seasonality of vegetation in regional-to-continental-scale studies of ecological processes. In particular, in tropical grassland and woodland savanna, the scatterometer signal is influenced by the loss of vegetation due to seasonal anthropogenic fire and its recovery (refer to Figure 2 in Section News article). A combination of multi-temporal data from NSCAT and QSCAT at high frequency (14.0 and 13.4 GHz, respectively) and the ESCAT at lower frequency (5.3 GHz) has the potential for identifying the long-term climatic impacts on moisture availability and vegetation loss or recovery.

Available Ice and Land Data Products

The SCP provides two basic forms of gridded image products: 1) a calibrated browse backscatter image at the intrinsic sensor resolution (Table 1); and 2) a unique enhanced resolution image product, which combines multiple overlapping passes over intervals of a few days, or just one using an algorithm called Scatterometer Image Reconstruction (SIR) [Early and Long, 2001]. The resulting enhanced resolutions are about 25 km for ESCAT, 8-10 km for NSCAT and SASS, and either 8-10 km or 5-6 km for QSCAT. In addition, each enhanced image product is decomposed into two sub-images: 1) a backscatter image normalized to the mid-swath incidence angle of 40^o (so-called A image); and 2) the gradient of backscatter over the range of incidence angles (so-called B image). The latter product is not produced for QSCAT because its conically scanning antenna operates at two fixed incidence angles.

Bertoia, C., J. Falkingham, and F. Fetterer, Polar SAR data for operational sea ice mapping, in Recent Advances in the Analysis of SAR Data of the Polar Oceans, edited by R. Kwok and C. Tsatsoulis, Springer Verlag, Berlin, 201-234, 1998.

Drinkwater, M. R., and C. C. Lin, Introduction to the special section on emerging scatterometer applications, IEEE Trans. Geosci. Remote Sens., 38, 1763-1764, 2000.

Drinkwater, M. R., D. G. Long, and A. W. Bingham, Greenland snow accumulation estimates from scatterometer data, J. Geophys. Res., Atmos., PARCA Special Issue, in press, 2001.

Early, D. S., and D. G. Long, Image reconstruction and enhanced resolution imaging from irregular samples, IEEE Trans. Geosci. Remote Sens., 39, 291-302, 2001.

Gohin, F., and A. Cavanié, A first try at identification of sea ice using the three beam scatterometer of ERS-1, Int. J. Remote Sens., 16, 2031-2054, 1995.

Kimball, J. S., K. C. McDonald, A. C. Keyser, S. Frolking, and S. W. Running, Application of the NASA scatterometer (NSCAT) for determining the daily frozen and nonfrozen landscape of Alaska, Remote Sens. Environ., 75, 113-126, 2001.

Kwok, R., G. F. Cunningham, and S. Yueh, Area balance of the Arctic Ocean perennial ice zone: October 1996 to April 1997, J. Geophys. Res., 104, 25747-25759, 1999.

Liu, A., Y. Zhao, and S. Y. Wu, Arctic sea ice drift from wavelet analysis of NSCAT and SSM/I data, J. Geophys. Res., 104, 11529-11538, 1999.