Gravity model development for TOPEX/POSEIDON: Joint Gravity Models 1 and 2

340

197

Abstract. The radial orbit error has long been the major error source in ERS-1 altimetry, crippled by having only satellite laser ranging for precise tracking and relying on insufficiently accurate general-purpose gravity field models. Altimeter crossovers are used very effectively as additional tracking data to laser ranging. The ERS Tandem Mission even provides the unique possibility to simultaneously determine orbits of two similar satellites flying the same orbit.Altimeter crossovers between the two satellites then link the two orbits into a common reference frame. Tailoring of the Joint Gravity Model 3 (JGM 3) is another step to reduce orbit errors. This technique is aimed at the reduction of the geographically anticorrelated orbit error (observed in the crossover height differences) through the adjustment of selected gravity field parameters. The resulting Delft Gravity Model (DGM)-E04 has reduced this part of the orbit error by a factor of 2, performs even better with respect to the ESA-provided orbits, and also outperforms the recent Earth Geopotential Model EGM96 in this respect. ERS-1 and ERS-2 orbits for the entire Tandem Mission are computed and studied in detail, and orbit errors due to the gravity field and nonconservative forces are identified. Analyses systematically show that the orbits computed with JGM 3 have a radial root-mean-square orbit accuracy of 7 cm, with DGM-E04 5 cm. IntroductionThe Tandem The bulky satellites ERS-1 and ERS-2 were never designed for high-accuracy orbit determination, and the loss of the Precise Range and Range-Rate Equipment (PRARE) tracking system left ERS-1 even more poorly equipped for orbit determination. Yet, subdecimetric orbit accuracy is not of academic interest only. The ERS altimetric system has performed well above expectations and is unique because of its multidisciplinary character, sampling not only ocean but also land and ice surfaces, in combination with the suite of instruments on board, providing, e.g., simultaneous measurements of wet tropospheric content and surface temperature. Undoubtedly, ERS will always lag behind on the 2-cm root-meansquare orbit accuracy of the TOPEX/POSEIDON altimeter mission [Marshall et al., 1995], so only when the precise orbit determination is stretched to its very limits, ERS altimetry will be regarded a reliable source of information. Only then will ERS be able to demonstrate its additive value in ocean research and its unique capabilities in, e.g., monitoring of the ice sheet mass balance.In section 2 we will discuss the numerous advances made in ERS operational and precise orbit determination over the years up to the current state-of-the-art modeling. An important step in this is the development of the ERS-tailored Delft Gravity Model (DGM)-E04 (section 3). In section 4, DGM-E04 demonstrates that it constitutes a remarkable improvement on the ERS orbits but also has a more general applicability. Section 5 examines the effect of nonconservative forces acting on the satellite. Section 6 combines all results and attem...

Section: Historymentioning

confidence: 99%

Precise orbit determination and gravity field improvement for the ERS satellites

Scharroo¹,

Visser²

1998

340

197

“…Finally, we provide a tour of the gravity field of the oceans and point out a few of the new and interesting features. We do not attempt to review other methods for recovery of shortwavelength gravity information from satellite altimetry [e.g., [Nerem et al, 1994] is used to establish the height of the satellite above the reference ellipsoid (dashed curve). We assume the sea surface height above the ellipsoid is equal to the geoid height so permanent sea surface slopes associated with currents will appear as false anomalies in our gravity solution.…”

Section: Introductionmentioning

confidence: 99%

Marine gravity anomaly from Geosat and ERS 1 satellite altimetry

Sandwell¹,

Smith²

1997

1,617

1,228

Abstract. Closely spaced satellite altimeter profiles collected during the Geosat Geodetic Mission (-6 km) and the ERS 1 Geodetic Phase (8 km) are easily converted to grids of vertical gravity gradient and gravity anomaly. The long-wavelength radial orbit error is suppressed below the noise level of the altimeter by taking the along-track derivative of each profile. Ascending and descending slope profiles are then interpolated onto separate uniform grids. These four grids are combined to form comparable grids of east and north vertical deflection using an iteration scheme that interpolates data gaps with minimum curvature. The vertical gravity gradient is calculated directly from the derivatives of the vertical deflection grids, while Fourier analysis is required to construct gravity anomalies from the two vertical deflection grids. These techniques are applied to a combination of high-density data from the dense mapping phases of Geosat and ERS 1 along with lower-density but higher-accuracy profiles from their repeat orbit phases. A comparison with shipboard gravity data shows the accuracy of the satellitederived gravity anomaly is about 4-7 mGal for random ship tracks. The accuracy improves to 3 mGal when the ship track follows a Geosat Exact Repeat Mission track line. These data provide the first view of the ocean floor structures in many remote areas of the Earth. Some applications include inertial navigation, prediction of seafloor depth, planning shipboard surveys, plate tectonics, isostasy of volcanoes and spreading ridges, and petroleum exploration.

“…The long-wavelength (>2000 km) component of the field is first removed using the JGM-3S [Nerem, 1994;Tapley et al, 1996] satellite tracking field. Using splines in tension [Smith and Wessel, 1990], the ascending and descending slopes are independently interpolated onto a grid with intervals of 0.025 ø latitude and 0.10 ø longitude (--•2.5 km x 2.5 km grid).…”

Section: Ers Satellite Gravitymentioning

confidence: 99%

“…Long-wavelength errors in the ERS gravity field would require (1) substantial errors in the long-wavelength (,k > 2000 km) in JGM-3S [Nerem, 1994;Tapley et al, 1996] satellitetracking gravity field used in the remove-restore procedure for ERS gravity computation [Laxon and McAdoo, 1998]; (2) systematic errors in the ERS orbits; (3) unspecified limitations in the gravity computation algorithms or the flat Earth assumption in the transform; (4) the uncertain effects of the lack of detailed surface gravity north of 82øN as far field input in the geodetic transformation computations of ERS gravity, or (5) the mapping of steady state ocean currents or regional changes in ice freeboard into the gravity field. Large errors in requirement 1 or 2 have not been identified and seem unlikely.…”

Section: Visual Inspection Reveals the Similarity Between Ers 1998mentioning

confidence: 99%

New gravity data in the Arctic Ocean: Comparison of airborne and ERS gravity

Childers¹,

McAdoo²,

Brozena³

et al. 2001

Abstract. New gravity fields from airborne gravimetry and from ERS-1 and -2 satellite altimetry cover extensive portions of the Arctic Ocean. These two data sets may constitute as much as 60% of the data contributions to the Arctic Gravity Project compilation. Here we evaluate the accuracy and resolution of these data and quantify their impact on the compilation. Both gravity determinations compare favorably with Geological Survey of Canada surface measurements in the Beaufort Sea (airborne, 1.86-2.09 mGal rms; ERS, 2.64-3.11 mGal rms). Comparisons between the airborne and ERS data over the Chukchi Borderlands reveal a 4.38 mGal rms difference over the smoother region of the field and 7.36 mGal rms over the rugose field generated by the shallow ridges and deep troughs. Coherency between the two data sets in the Chukchi region implies a resolution of 19 km. Comparison with Science Ice Expedition submarine measurements over Chukchi Plateau suggests that the ERS field resolves even shorter-wavelength signal than the airborne data, whereas in the Beaufort Sea the airborne data showed better coherence to ground truth data. Long-wavelength differences exist between the two data sets, expressed as a 2-3 mGal offset over the Chukchi region. This study highlights the respective strengths of the two data sets. The ERS gravity field has the advantage of ubiquitous coverage of the ocean south of 81.5øN, a denser sampling of the gravity field, and a recovery of signal down to -15 km. The airborne data cover a significant portion of the polar hole in the satellite coverage, have lower measurement noise, and recover somewhat higher anomaly amplitudes in the 25-100 km wavelength range. IntroductionThe origin and evolution of the Arctic Ocean are two of the largest remaining uncertainties in plate tectonics. Whereas the opening histories of the world's other major ocean basins have been understood for more than 20 years, the formation of the Arctic's Amerasia Basin remains a mysteu. To date, the primary impediment to a fuller tectonic understanding has been the dearth of available geophysical data in the Arctic Ocean. Logistical problems, due to the year-round ice cover, were compounded by Cold War political sensitivities; as a result, the ocean basin was poorly sampled with bathymetry, gravity, and magnetics data. Moreover, much of the data actually collected , of which these two data sets will constitute a significant contribution, it is important to understand how these data sets compare in terms of accuracy and resolution. This study compares these two data sets over two regions of overlap in the Amerasia Basin. First, both sets are compared separately with high-quality ice surface gravity measurements in the Beaufort Sea region (Figure 1) of the Canada Basin, and then with each other and with Science Ice Expedition (SCICEX) submarine gravity measurements over the rugose gravity field of the Chukchi Borderland.