• Korean Journal of Remote Sensing
• /
• v.17 no.4
• /
• pp.297-305
• /
• 2001

Ocean satellite altimetry-implied free-air gravity anomalies have had the shortest wavelengths removed during the processing to generate the optimal solution between multiple radar altimeter missions. ERS-1 168day mission altimetry was residualized to a reference geoid surface generated by integrating Anderson & Knudsen’s free-air gravity anomalies for the Barents Sea. The altimetry tracks were reduced and filtered to extract the shortest wavelengths (between 4 and 111 km) from both ascending and descending tracks, respectively. These data were recombined using existing quadrant-swapping techniques in the wavenumber domain to generate a correlated, high frequency gravity field related to the local geologic sources. This added-value surface adjusted the reference free-air gravity anomalies to better reflect features in the gravity field at a wavelength related to the distance between altimetry ground tracks.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.969-972
• /
• 2006

The long-range High-Frequency (HF) ocean radar system has observed surface velocity field in the upstream of the Kuroshio north of Ishigaki Island and east of Taiwan since 2001. Applying a new method to extract geostrophic velocity component from the HF surface velocity data with the aid of satellite-born wind data, time series of daily surface geostrophic velocity field has been determined. Despite limited width of the study area of the HF radar, analysis of the sea surface height anomaly determined from the satellite altimetry data in a wider area can provide estimated dates of arrival of mesoscale eddies in the study area of the HF radar. Variations of the Kuroshio position and strength are studied in detail for these cases of interaction with mesoscale eddy, although number of occurrence of direct interaction with the Kuroshio in the study area is not statistically enough. For example, when an anticyclonic eddy approaches to the Kuroshio, the Kuroshio axis is found tend to move northward, keeping away from the approaching eddy from the east.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.672-675
• /
• 2006

The geostrophic current component is estimated from the sea surface velocity observed by the long-range High-Frequency Ocean Radar (HF radar) system in the upstream of the Kuroshio, by comparing with geostrophic velocity determined from along-track T/P and Jason-1 altimetry data. However, the sea surface velocity of the HF radar (HF velocity) contains not only the geostrophic current but also the ageostrophic current such as tidal current and wind-driven Ekman current. Tidal current component is first extracted by the harmonic analysis of the time series of the HF velocity. Then, the Ekman current is further estimated from daily wind data of IFREMER by applying the least-square method to the residual difference between the HF velocity and the altimetry geostrophic velocity. As a result, the Ekman current in the HF velocity is estimated as 1.32 % of the wind speed and as rotated 45$^{\circ}$ clockwise to the wind direction. These parameters are found almost common in the Kuroshio area and in the Open Ocean. After these corrections, the geostrophic velocity component in the HF velocity agrees well with the altimetry geostrophic velocity.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.17-20
• /
• 2006

Three HF ocean radar stations were installed at the Soya/La Perouse Strait in the Sea of Okhotsk in order to monitor the Soya Warm Current. The frequency of the HF radar is 13.9 MHz, and the range and azimuth resolutions are 3 km and $5^{\circ}$, respectively. The radar covers a range of approximately 70 km from the coast. It is shown that the HF radars clearly capture seasonal and short-term variations of the Soya Warm Current. The velocity of the Soya Warm Current reaches its maximum, approximately 1 m $s^{-1}$, in summer, and weakens in winter. The velocity core is located 20 to 30 km from the coast, and its width is approximately 50 km. The surface transport by the Soya Warm Current shows a significant correlation with the sea level difference along the strait, as derived from coastal tide gauge records. The cross-current sea level difference, which is estimated from the sea level anomalies observed by the Jason-1 altimeter and a coastal tide gauge, also exhibits variation in concert with the surface transport and along-current sea level difference.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.39-42
• /
• 2006

The paper reports on the first experimental evidence for space-observed manifestation of the open ocean tsunami in the microwave radar backscatter (in C- and Ku-bands). Significant variations of the radar cross section synchronous with the sea level anomaly were found in the geophysical data record of the altimetry satellite Jason-1 for the track which crossed the head wave of the catastrophic tsunami of 26 December 2004. The simultaneous analysis of the available complementary data provided by the satellite three-channel radiometer enabled us to exclude meteorological factors as possible causes of the observed signal modulation. A possible physical mechanism of modulation of short wind waves due to transformation of the thin boundary layer in the air by a tsunami wave is discussed. The results open new possibilities of monitoring tsunamis from space..

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.38-38
• /
• 2006

• Proceedings of the KSEEG Conference
• /
• 2000.04b
• /
• pp.140-142
• /
• 2000

• Ocean and Polar Research
• /
• v.24 no.3
• /
• pp.255-261
• /
• 2002

We investigated the distribution of sea ice using Topex/Poseidon (T/P) and ERS-1 .ada. altimeter data in the northwest Weddell Sea, Antarctica, between the area $45-75^{\circ}W\;and\;55-66^{\circ}S$. Using the Geo_Bad_1 flag of the Merged GDR of the T/P, we classified the surface into ocean, land, and sea. Total 257 cycles of altimeter measurements between Oct. 1992 and Sep. 1999 (for nearly 2570 days) were used to analyze the distribution of the Antarctic sea ice. We then calculated the surface area of ice coverage using SUTM20 map projection to monitor the periodic variations. Each year, the maximum and minimum coverage of the sea ice were found in late August and February in the study area, respectively. We also studied the sea ice distribution using ERS-1 altimeter data between $45-75^{\circ}W\;and\;55-81.5^{\circ}S$ to compare with the T/P Using the Valid/Invalid flag of the Ocean Product, we analyzed the sea ice distribution between March and August of 1995, which showed very good coherence with the T/P measurements. Our preliminary results showed that the altimeter measurements can be effectively used to monitor the distribution of the sea ice in the polar region. However, the size of radar footprint, typically 2-6km depending on the roughness of the sea surface, may be too big to monitor the sharp boundary between ice and water/land. If more other altimeter mission data with dense coverage such as Geosat GM are analyzed together, this limitation can be significantly improved. If we also combine other microwave remote sensing data such as radiometer, and SSM/I, the result will be significantly enhanced.

• Korean Journal of Remote Sensing
• /
• v.21 no.1
• /
• pp.73-81
• /
• 2005

We construct improved geocentric digital elevation model (DEM), estimate tidal dynamics and ice stream velocity over Sulzberger Ice Shelf, West Antarctica employing differential interferograms from 12 ERS tandem mission Synthetic Aperture Radar (SAR) images acquired in austral fall of 1996. Ice, Cloud, and land Elevation Satellite (ICESat) laser altimetry profiles acquired in the same season as the SAR scenes in 2004 are used as ground control points (GCPs) for Interferometric SAR (InSAR) DEM generation. 20 additional ICESat profiles acquired in 2003-2004 are then used to assess the accuracy of the DEM. The vertical accuracy of the OEM is estimated by comparing elevations with laser altimetry data from ICESat. The mean height difference between all ICESat data and DEM is -0.57m with a standard deviation of 5.88m. We demonstrate that ICESat elevations can be successfully used as GCPs to improve the accuracy of an InSAR derived DEM. In addition, the magnitude and the direction of tidal changes estimated from interferogram are compared with those predicted tidal differences from four ocean tide models. Tidal deformation measured in InSAR is -16.7cm and it agrees well within 3cm with predicted ones from tide models. Lastly, ice surface velocity is estimated by combining speckle matching technique and InSAR line-of-sight measurement. This study shows that the maximum speed and mean speed are 509 m/yr and 131 m/yr, respectively. Our results can be useful for the mass balance study in this area and sea level change.

• Journal of the Korean earth science society
• /
• v.23 no.4
• /
• pp.349-356
• /
• 2002

In an effort to properly assess the validity of spaceborne radar altimeter measurements, we made a direct comparison of two different sea surface heights (SSH) acquired by both Topex/Poseidon (T/P) satellite and in-situ tide-gauges (T/G). This comparative analysis was conducted using the data sets collected from three locations along the eastern coast of Korea which include: Ulleungdo, Pohang, and Sokcho. In the course of the analysis of satellite altimeter, information of SSH was extracted from the T/P MGDR data sets through the application of both atmospheric and geophysical corrections. To compare the T/P data sets in parallel basis, the T/G data sets were averaged using the measured values within the peripheral radius of 55km. When compared among different locations, the compatibility between the two methods was much more significant in an offshore location (Ulleungdo) than the two onshore locations (Pohang, Sokcho). If the low-pass filtered results were compared among the sites, the offshore site exhibited the best correlations between the two methods (correlation coefficient of 0.91) than those of the onshore sites. These large differences in the strength of correlations among different locations are due to the deformation of M2, S2, and K1 tidal components used in the tidal model. In case of the offshore location, the compatibility of the two different methods were improved systematically by the low-pass filtering with an increase of the filtering duration such as up to 200 days.