Received: 6 October 2008 / Accepted: June 2009
Download 5.1 Mb. Pdf ko'rish
|
radon review
Partial Stacking Window (a) (b) Fig. 5 a Unstacked seismic traces after aligning on SS. The shaded region marks a moving- average (partial stacking) window. The averages of the sliding windows are plotted at the distance of the original seismograms. b Seismic traces after partial stacking. Weak signals corresponding to S410S and S660S are greatly enhanced Surv Geophys 123 equalize the source, as the SS-SdS relative times are less affected by origin time uncertainty or source complexity. For consistency the ray parameter p of a given signal of interest (e.g., S660S) is expressed as the differential ray parameters to SS; p is approximately constant for the appropriate distance range (An et al. 2007 ; Gu et al. 2009 ). The process of aligning SS is equivalent to setting the reference ray parameter to a value of zero. Unlike time-domain analyses (see review by Deuss, this issue), Radon-based methods preserve the relative move-out between SS and SdS. Figure 6 a shows a record section of transverse-component synthetic data aligned and normalized by SS. The Radon model after applying HRT (Fig. 6 b) recovers three well- defined energy maxima with relative p values of -0.10, -0.23 and -0.50 deg/s for S220S, S410S and S660S, respectively. Signals outside of the immediate s-p window of interest, for instance ScSdScS and sdsS/sdsS diff , are effectively decoupled from the reconstructed time series (Fig. 6 c). Mechanisms that allow for Radon-domain windowing and post- conditioning prior to data reconstruction underscore a crucial advantage of inversion-based RT methods (e.g., LSRT and HRT) over classical RT methods. The post-conditioning criteria/algorithms are empirically determined from signal properties and tradeoff curves. Measurement uncertainties are estimated entirely from the Radon domain. One can adopt a bootstrapping procedure (e.g., Shearer 1993 ) to determine s and p values from random subsets of the seismic traces used in the final inversion. The standard deviation of the automatically determined Radon solutions is a reasonable estimate of the measurement uncertainty. The resulting depth uncertainty is usually less than 3 km for the measurements shown in Sect. 4 . 4 Mantle Reflectivity Imaging This section briefly reviews recent observations of the mantle reflectivity structure based on LSRT of SS precursors. Some images are modified from An et al. ( 2007 ) and Gu et al. ( 2009 ) to testify the power of LSRT and HRT in delineating the seismic reflectivity structure within the Earth’s crust and mantle. We highlight the difference between advanced RT approaches and classical time- or Radon-domain approaches whenever appropriate, and refer the reader to the referenced manuscripts for in-depth discussions and interpretations of the observations. We make the following abbreviations to improve succinctness: (1) An et al. 2007 (to An07), (2) Deuss 2007 (to Deuss07), (3) Flanagan and Shearer 1998 (to FS98), (4) Upper mantle transition zone (to MTZ), and (5) 410, 520 and 660 km discontinuities (to 410, 520 and 660, respectively). 4.1 Role of Ray Parameter in Depth Estimation A key advantage of solving for ray parameters (in addition to time) is that it provides information on the behavior (e.g., slope and continuity) of a move-out curve for a given seismic phase. This information is critical in validating the nature of the arrival. For instance, substantial deviations of measured p from the expected value would raise questions about the true identity of the phase, whereas relatively minor variations may be evidence of a dipping interface or a heterogeneous velocity structure. Another important contribution of the ray parameter information, which sometimes goes unnoticed, is that it improves the accuracy of the reflector depth computation. For example, previous studies (e.g., Gu et al. 1998 ) adopted a simple time-to-depth conversion formula based on Per- turbation Theory (Dziewonski and Gilbert 1976 ), Surv Geophys 123 dr ¼ dt r 2 r v ðrÞ p 2 1=2 ð13Þ where r is the radius of the Earth in kilometers up to the reflector, dr is the perturbation in radius relative to PREM prediction, dt is the perturbation of reduced time, p is the mea- sured ray parameter (not perturbation), and v ðrÞis the shear velocity beneath the reflector. S220S S660S S410S e d uti l p ma de zil a mr o n Radon Model Download 5.1 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling