Home / Sci. Teams / OSTST 2012 / Abstracts
Print  PDF

Abstract

Preliminary studies on empirical roll angle error reduction and tidal detection for COMPIRA

Dr Osamu ISOGUCHI(1), Dr Masato OHKI(2), Mr Akihiko SHINOZAKI(3)

(1) Remote Sensing Technology Center of Japan, Japan
(2) Japan Aerospace Exploration Agency (JAXA), Japan
(3) Remote Sensing Technology Center of Japan (RESTEC), Japan

Session theme: Others (Posters only)

  • poster

    Abstract

    Japan Aerospace Exploration Agency (JAXA) is working on a conceptual study of altimeter mission named Coastal and Ocean measurement Mission with Precise and Innovative Radar Altimeter (COMPIRA), which will carry a wide-swath altimeter named Synthetic aperture radar (SAR) Height Imaging Oceanic Sensor with Advanced Interferometry (SHIOSAI).
    As preliminary studies on sea level anomaly (SLA) measurements by COMPIRA, we conducted simulations about 1) the reduction of roll angle errors which is peculiar to the SAR-typed altimeter and 2) the effect of the wide-swath SSH observation on tide detection in the marginal and coastal seas where the tide model derived by traditional nadir-type altimeters is not, under existing conditions, accurate enough to derive SLAs.
    Empirical correction methods of roll angle errors have been developed as studies toward SWOT by some researchers (i.e., Dibarboure et al., 2011). We performed, by following them, the crossover method to examine the effect of the proposed COMPIRA orbit configuration on the roll angle error reduction. The current orbit configuration of COMPIRA is as follows: the revisit time is 9.8671 days and the inclination is about 51?, which makes acquisition time differences of crossover points relatively short.
    Simulated sea surface heights (SSH) were obtained from Japan Coastal Ocean Predictability Experiment (JCOPE) 2 produced by JAMSTEC. Error factors such as roll angles, orbit, baseline length, sensor thermal noise, and wet tropospheric delays were added to the SSHs. Their dependence on cross-track distance was defined as linear, constant, quadratic, random, and 50km cycle, respectively. The amplitude and cycle of the roll angles were given as 1arcsec and 800km, respectively, giving rise to a 27cm SSH error at the 80km cross-track distance.
    Roll angles were locally estimated at each crossover by a least square method from the SSH differences and the cross-track distances. The result showed that the error of the estimated roll angles was 0.07arcsec corresponding to 2.9cm error at 80km. It was thus demonstrated that the proposed roll angle error reduction method could reduce the SST error caused by low frequency roll angle to several centimeters which is equivalent to those caused by the principal error factor.
    Next, we examined the effect of wide-swath SSH observations on tide detection. We simulated 3-year ocean tide measurements using the proposed COMPIRA orbit configuration. The tide model we applied is NAO.99b tidal prediction system (Matsumoto et al., 2000). Amplitude and phase of the major 8 tidal constituents were calculated by a harmonic analysis from the simulated tide time series at each grid. The derived amplitudes and phases were then interpolated onto 0.25deg. grid for two cases: one used SHIOSAI plus nadir altimeters and the other nadir altimeter only. The interpolated tidal parameters were evaluated by comparing the model tidal parameters. The result showed significant improvement in SHIOSAI plus nadir altimeter case especially in the marginal and coastal seas. In the East China Sea, whereas the total error of the major 8 tidal constituents was 2.6cm for the SHIOSAI plus nadir altimeter case, it was 9.9cm for the only nadir altimeter case.

     
    Corresponding author:

      Dr Osamu Isoguchi
      Remote Sensing Technology Center of Japan
      JAXA Tsukuba Space Center
      2-1-1, Sengen
      305-8505 Tsukuba
      Japan
      E-mail:
    •