行政院國家科學委員會專題研究計畫 期中進度報告
台灣西南海域大陸與海洋相互作用的特性--海陸岩圈交界
與重力異常及磁力異常之小波譜值相關(2/3)
期中進度報告(精簡版)
計 畫 類 別 ： 整合型 計 畫 編 號 ： NSC 95-2611-M-002-004- 執 行 期 間 ： 95 年 08 月 01 日至 96 年 07 月 31 日 執 行 單 位 ： 國立臺灣大學海洋研究所 計 畫 主 持 人 ： 喬凌雲 報 告 附 件 ： 國外研究心得報告 處 理 方 式 ： 本計畫可公開查詢中 華 民 國 96 年 07 月 16 日
Crustal magnetization equivalent source model of Mars
constructed from a hierarchical multiresolution inversion
of the Mars Global Surveyor data
Ling-Yun Chiao,1 Ja-Ren Lin,1 and Yuan-Cheng Gung2
Received 28 March 2006; revised 13 July 2006; accepted 21 July 2006; published 29 December 2006.
[1] Several magnetic field models of Mars have been constructed since the Mars Global Surveyor data became available. Three distinct schemes formulated through spherical harmonic functions, discrete equivalent dipoles, and the continuous magnetic field kernels have yielded results that are grossly compatible but with very different details. Models of internal potential function in terms of spherical harmonics tend to yield divergent high-degree Mauersberger-Lowes spectra, whereas crustal magnetization models exhibit flat but still significant spectra up to high degrees. To have a better fitting to the observed data seems to have dominated previous efforts that have yielded fine details with wavelengths shorter than the lateral track spacing. The variance-reduction versus model-variance tradeoff analysis is invoked in this study for the determination of the appropriate regularization. Taking advantage of the recently developed multiscale inversion, we are able to conservatively retain only the model components that are robustly constrained by the data rather than unilaterally pushing for a higher degree of fitting. With the variance reduction around 82%, we find that to reach a reasonably fair data fitting without high model variance, the high-degree power spectra of our preferred model exhibit an obvious decaying trend, implying that a lot of the short-wavelength energy embedded within established models is either not robustly resolvable or is of external origin or is simply reflecting the nonuniform distribution of sampling at short scales. The reason that models based on spherical harmonics have greater high-degree power is attributed to the spectral leakage due to the truncated representation.
Citation: Chiao, L.-Y., J.-R. Lin, and Y.-C. Gung (2006), Crustal magnetization equivalent source model of Mars constructed from a hierarchical multiresolution inversion of the Mars Global Surveyor data, J. Geophys. Res., 111, E12010, doi:10.1029/
2006JE002725.
1. Introduction
[2] A magnificent magnetic field variation in the southern
hemisphere of Mars has been discovered owing to the compilation of the Mars Global Surveyor (MGS) data [Acuna et al., 1999]. There is, however, no significant field intensity observed for the northern lowlands although there are as many observations for the northern hemisphere as for the south. The significant magnetic signature demarcating an extensive part in the south has been attributed to ancient magnetization of the Martian crust [e.g., Hood et al., 2005]. Consequently, there have been considerable efforts to con-struct models of the Martian crustal magnetic field, not only for delineating potential tectonic features but also to system-atically summarize the robust information of the precious MGS data as completely as possible.
[3] The nature and the quality of the magnetic observations
as well as features of the main phases of MGS have been documented previously [e.g., Albee et al., 2001]. Essentially, the MGS satellite observed magnetic data consists in vectorial, three-component magnetic field observations at different altitudes, from below 200 km to 367 – 435 km, during different mission phases. The three-component data set we used in this study is the same set previously used to construct the spherical harmonics degrees 90 internal poten-tial model [Cain et al., 2003] as well as the spapoten-tially continuous magnetization model [Whaler and Purucker, 2005]. There are in total three-component measurements at 111,274 points, with altitudes from 102 to 426 km, composing the 333,822 field intensity data. The specific parameters and the adopted coordinate system of the data are described by Cain et al. [2003].
[4] The consensus established from previous works
attributes the current major contributor of the Martian magnetic field to the lithospheric magnetization of a layer about 40 km thick, and that there is presently no dominant dipole field for the planet [e.g., Voorhies et al., 2002; Langlais et al., 2004; Whaler and Purucker, 2005]. One school of approach is to find the scalar internal potential
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, E12010, doi:10.1029/2006JE002725, 2006
Click Here for Full Article 1
Institute of Oceanography, National Taiwan University, Taipei, Taiwan.
2Department of Geosciences, National Taiwan University, Taipei, Taiwan.
Copyright 2006 by the American Geophysical Union. 0148-0227/06/2006JE002725$09.00
function at the surface of Mars in terms of spherical harmonics such that the gradient vectors of the model will fit the observations [Connerney et al., 2001; Arkani-Hamed, 2001, 2002, 2004; Cain et al., 2003]. Discussions about the effects of the variation of the attitudes and the lateral sampling have been raised. Interestingly, although assump-tions on the internal origin of the magnetic field have been made in these studies, the divergent Mauersberger-Lowes power spectra [Backus et al., 1996] toward high degrees, however, implies significant contributions from external sources. Other studies assume a continuously varying mag-netization vector function M(r), where r stands for the position vector, such that the theoretical magnetic field intensity vector observed at robs,
B rð obsÞ ¼ rrobs m0 4p Z V M rð Þ rr 1 r robs j j d3r 8 < : 9 = ;; ð1Þ
fits the field observation. This linear data rule states that the data functional is of the form of an inner product between the model function and the data kernel. To evaluate expression (1) for a given magnetization model, a numerical scheme based on parameterizing the model function must be implemented. Langlais et al. [2004] use the equivalent source dipole technique that attributes the magnetic field to the contribution from 4840 dipoles with spatially varying magnetization intensity and direction, uniformly distributed across the globe and 20 km below the Martian surface. Whaler and Purucker [2005], on the other hand, expand the model function in terms of the data kernels. One advantage of this expansion is that it automatically avoids annihilators [Parker, 1994], since any component expressible in terms of the data kernels will not be orthogonal to all data kernels. That is, there will be no model components of this form that make no contribution to the data. One of the major disadvantages, however, is that the resulting Gram matrix is too sizeable and thus computationally demanding, although the matrix is usually sparse. Whaler and Purucker [2005] take advantage of the sparseness and indicate that an effective computation can usually be performed with only the 0.21% largest elements of the Gram matrix retained. Both these two studies obtain models that reveal power spectra similar to the former studies under degree 40. The higher-degree power spectra become considerably lower but are still significant. There have been concerns that crustal magnetic features with wavelengths shorter than the altitude of the observation might not be robustly resolvable [Connerney et al., 2001; Arkani-Hamed, 2002]. Noticeably, since the north-south trending track spacing of the MGS has a width of 2° – 5°, that is, 100 –300 km at the equator [Arkani-Hamed, 2001], it has been argued that the highest harmonics degree corresponding to twice the lateral resolvable wavelength is thus about 65 [Arkani-Hamed, 2004]. In spite of these discussions, recent models tend to have significant power spectra contributions from much higher degrees.
2. Method
[5] We basically follow the approach that inverts for the
spatial variation of the equivalent source crustal
magnetiza-tion. We build the spherical tessellation initiated from a spherical icosahedron. Midpoints on the edges of each of the 20 spherical triangles are then connected to form 4 children triangles. The refinement of the spherical meshes is then executed successively until we have 10242 nodes marking the vertices of the 20480 (= 20 45) triangular faces. Summation of the integrand of equation (1) evaluated at finite Gaussian integration points [e.g., Zienkiewicz and Taylor, 1991] within each triangle is then computed to numerically approximate the inner product of the data rule. Let m be the vector with M (= 3 10242) magnetization model components, then the N (= 333,822) dimensional data vector d is constrained by
Gm¼ d: ð2Þ
[6] Notice that in the current formulation, the degrees of
freedom of the model, 3 10242, is more than twice as much over the previous formulation based on the equivalent source dipoles, 3 4840 in the work of Langlais et al. [2004]. The parameterization of Langlais et al. [2004] assumes a finite amount of equivalent dipoles located on the vertices of the spherical triangular meshes. We, on the other hand, assume that the magnetization varies linearly within each of the 20480 triangles such that the magnetization is a globally continuous vector function. This further enables much better capability of resolving short-wavelength features. Elements of each row of the sensitivity matrix G specify the depen-dency of a particular datum upon the M dimensional model vector. An example of the spatial variations of selected observations reveals the localized constraints and the effects of the distinct altitudes (Figure 1). Conventionally, model estimates, ^m, can then be solved by the damped least squares (DLS) [e.g., Lawson and Hanson, 1974] algorithm,
^
m¼ G TGþq2I 1GTd: ð3Þ
[7] The value of the nonnegative damping factor q2
controls the strictness of the imposed preference of the minimum model norm. It is also a knob for tuning the variance reduction (vr) versus model variance (sm) tradeoff.
Briefly, the variance reduction is defined to indicate the capability of a model (m) to reconstruct the observed data (d). It can be calculated by vr¼ 1 jjGm djj2 jjdjj2 ! 100%: ð4Þ
[8] On the other hand, the model variance is a measure of
the uncertainty of a model manifested from noises contam-inated to the data; it is computed [Paige and Saunders, 1982] by sm¼ XM l¼1 s2l; s2l ¼ jjd Gmjj 2 sll; sll¼ diag G TGþ q2I 1 h i : ð5Þ
[9] It is noted that a heavier damping setup by a larger
value ofq2usually leads to a robust model (lowersm), but
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
2 of 12
Figure 1. Spatial variation of the amplitude of the sensitivity matrix G (equation (2)). It can be visualized as the variation of the discrete data kernel function. In this figure a particular example is shown for the data observed at (180°E, 30°N) and an altitude of 370 km, as well as another southern hemisphere observation at (180°E, 45°S) (marked by small green open circles, respectively) but at a lower altitude of 130 km. Notice that for a lower-altitude observation, since the sources to observation distances are shorter relative to those of a higher-altitude observation, there will be a wider area such that the sensitivity of sources within it are above an effective threshold.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
3 of 12
sacrifices the data fitting (lower vr) at the same time. We
will show in the following how tradeoff between the model robustness and the data fitting helps to determine an appropriate value ofq2and an optimal model.
3. Multiscale Inversion Based on the Spherical Wavelet Basis
[10] It has been pointed out that minimum norm solutions
obtained from DLS generally lack interpolation capabilities into sparsely sampled areas and tend to yield fragmented and fractured models [e.g., Chiao and Liang, 2003]. Reg-ularizations based on enforcing model smoothness or rough-ness penalization have also been conventional practices in handling geophysical inverse problems [e.g., Menke, 1984; Delprat-Jannaud and Lailly, 1993]. The implementations, however, usually presume that the model smoothness [e.g., Meyerholtz et al., 1989], or the intrinsic model correlation length [Tarantola and Nercessian, 1984], is spatially uni-form or stationary. It has been shown that this is not a realistic presumption and has led to devices of multiscale regularization based on wavelet representations of models such that spatially nonstationary smoothness enhancement is automatically invoked depending on the in situ density of model constraints offered by the data [Chiao and Kuo, 2001; Chiao and Liang, 2003]. We follow the same ratio-nale and transform the aforementioned spherical meshes into a stage to build spherical wavelet bases.
[11] To briefly summarize the algorithm, a simplified
single triangle is taken as an example (Figure 2). To discretely describe a function f(x) across the interior of the triangle, we can specify the spatial variation of f at uni-formly distributed nodes, such as f1 = f(r1), f2 = f(r2),
f3,. . .. . . where r1,r2 are position vectors at the internal
nodes 1,2 (Figure 2). These nodes are vertexes of internal triangles through successive levels of refinement of the original triangle. That is, connecting midpoints on the edges, the parent triangle D123 is subdivided into four
children triangles D456, D536, D146, D425 (Figure 2). Each of the resulting triangles can be further subdivided accordingly. Now instead of representing f(x) by [f1, f2, f3,
f4. . ..f9. . ...] distributed uniformly throughout the triangle,
there are ways to build hierarchical representations of f(x). A naı¨ve example is cast in the following sense:
Level1 : h11¼ f1;h21¼ f2;h13¼ f3 Level2 : h21¼ h11;h 2 2¼ h 1 2;h 2 3¼ h 1 3;h 2 4¼ f4 h1 1þ h 1 2 2 ; h25¼ f5 h1 2þ h 1 3 2 ; : : Level3 : h3 4¼ h 2 4;h 3 2¼ h 2 2;h 3 5¼ h 2 5;h 3 7¼ f7 h2 4þ h 2 2 2 ; h3 8¼ f8 h2 2þ h 2 5 2 ; : : . . . : ½h3 1;h 3 2;h 3 3;h 3 4;h 3 5; . . . : ¼ W fð½1;f2;f3;f4. . . :f9. . . : : Þ: ð6Þ
[12] That is, on the fundamental level, level_1, there are
3 degrees of freedom hi 1
= fi,i = 1,2,3 to be specified where
the upper index marks the refinement level and the lower indices are for the locations of nodes. On the next refine-ment, there are 6 degrees of freedom, hi2,i = 1..6. As
specified in equation (6), the first 3 degrees of freedom that are used to characterize the large-scale variation are inherited from the lower level of representation whereas the additional 3 degrees of freedom are obtained by the in situ deviations of f(x) from the expected values predicted by linearly interpolated from larger-scale variation at each midpoint, for example, h4
2
= f4 (h1 1
+ h2 1
)/2. That is, the original in situ variations, f4 = (h11 + h21)/2 + (h42), are
replaced by the combination of a low-passed portion (the contribution interpolated from a larger scale) and a high-passed detail. Fast wavelet transforms [e.g., Mallat, 1998] are efficient schemes that accomplish the transformation W in equation (6) that maps the strictly spatial representation fi
to a localized hierarchy representation hil of this sort. In
addition, lifting schemes [Sweldens, 1996] can be incorpo-rated to further improve the quality of the multiresolution representation. In this study, we transform the representation based on the original spherical mesh into an expansion utilizing spherical wavelet bases [see also Chiao and Kuo, 2001]. That is the reason why our construction subdivides the edges of each spherical triangles of the icosahedron by 25 segments instead of any integer as in the work of Langlais et al. [2004]. In fact, starting from the formulation (1) based on the direct spatial representation; we devise a bi-orthogonal wavelet transform [Cohen et al., 1992] directly on each row of the coefficient matrix G, that is GW*, such that the solution model vector to be solved for is now automatically the wavelet representation of the original spatial function for the crustal magnetization. That is, equation (2) is replaced by
GW*
ð Þ Wmð Þ ¼ d: ð7Þ
[13] The new solution becomes, instead of equation (3),
^
m¼ W1m^¼ W1 ðGW*ÞTðGW*Þ þ q2I
h 1
GW* ð ÞTd; ð8Þ
Figure 2. Triangular configuration as an example of multiresolution representation of a two-dimensional lateral variation (see the text and equation (6)).
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
4 of 12
where W1 is the inverse wavelet transform that reverses the operation of the forward wavelet transform W. The advantage of solving for m in the wavelet domain is that with the same amount of degrees of freedom, parameters in the wavelet representation are grouped into a natural hierarchy of local scales such that the damping regulariza-tion acts to sort through successive scales depending on the local data constraints. In short, sites with dense constraints are capable of resolving more details robustly whereas robust long-wavelength components are still available for sparsely constrained area.
4. Results
[14] We execute two different groups of inversions based
on the simple damping scheme (equation (3)) and the multiscale inversion (equation (8)), each with several dif-ferent values for the damping factor q2. The variance-reduction (vr) versus model-variance (sm) tradeoff curves
(Figure 3) clearly indicate how an appropriate model might be selected. We first notice that with comparable variance reduction, the results obtained via the multiscale inversions (marked by solid triangles on Figure 3) have model var-iances that are in general an order of magnitude lower than the simple damping results (marked by open circles on Figure 3). As mentioned in the previous section, this is due
to the way the model variation is assembled through the scales hierarchy from the longer wavelengths that have more accumulated constraints in the multiscale inversion. For both the simple damping results and the multiscale inversion results, high model variances are associated with the solutions that best fit the observational data (solutions marked by group 3 on Figure 3, that are located on the high-variance-reduction extreme on the tradeoff curves), imply-ing that there are significant unreliable components poorly constrained by the data embedded in such solutions to reach high data fitting. In other words, these lightly damped solutions are overinterpreting the information content of the data. On the other hand, solutions approaching the knees of tradeoff curves (marked by group 2 on Figure 3) that exhibit almost similar variance reductions (over 92%) bear considerably lower model variances. Continuing the trend of decreasing the model variance, conservative solutions with variance reduction around 82% (solutions group 1 around the knee of the tradeoff curves) reduce the model variance even more. Further model variance decreasing (along the reversed horizontal axis on Figure 3), however, sacrifices too much variance reduction to gain just barely significant decreases of model variance, and is thus under-fitting the precious observational data.
[15] For reasons discussed above, we believe that the
appropriate solutions worth exploring that will reveal robust model structure without sacrificing significant amount of data information are located in between group 2 and group 1. In fact, we prefer the conservative group 1 multiscale inversion solution (referred as solution_1 hereafter) that can be characterized as the most reliable model with a reasonably low data misfit. Simple damping group 2 solution (referred as solution_2 hereafter), on the other hand, can be treated as a reference conventional model that might be a little bit on the overinterpreting side. The overall patterns of the crustal magnetization revealed in these two solutions are similar (Figures 4 and 5). In fact, the general features are quite similar to previous works such as those obtained by Langlais et al. [2004] and Whaler and Purucker [2005]. However, the conservative multiscale solution solution_2 is dominated by long-wavelength structures at some places. Notice that this smoothing is not applied in a stationary sense, that is, the model is not the result of a uniform low-passing like in other conventional regularizations that enforce smoothness [Chiao and Kuo, 2001]. The relatively smooth model, solution_1, can fit the MGS data reasonably well (see also Figures 6, 7, and 8) although there are notable short-scale deviations from the observations. It is also worth pointing out that in Whaler and Purucker’s model, to build the minimum RMS magnetization model, short-scale fea-tures are required to enforce null magnetization within data gaps. These short-scale features have very little effects on modifying the data misfit or to increase the variance reduc-tion. Our solutions have considerably less and decaying high-degree power spectra but still retain reasonable data fitting. The reference simple damping solution, solution_2, has very similar Mauersberger-Lowes spectra up to degree 75 as compared to Whaler and Purucker’s model. However, our preferred robust multiscale solution, solution_1, has similar power spectra to almost all previous models only up to degree 40 and then starts to dive. We will show in the next section, through inversion experiments executed on Figure 3. Curves displaying tradeoff between variance
reduction of fitting (equation (4)) versus model variance (equation (5)) for different solutions. Notice that the scale of the model variance for simple damping solutions (marked by open circles and annotated on the upper horizontal axis) is almost an order of magnitude higher than the correspond-ing multiscale solutions (marked by solid triangles and annotations on the lower horizontal axis). The solutions marked as group 3 (solution_3, damping factor 104) are apparently underdamped since the variance reductions are not much better than solution_2 (damping factor 103), but the model variances are considerably higher. On the other hand, solution_1 (damping factor 3 103) represents a relatively conservative but reliable solution without sacrifi-cing too much variance reduction.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
5 of 12
Figure 4. Crustal magnetization of Mars obtained from the multiscale inversion, solution_1 (left column) and the conventional simple damping solution_2 (right column). The top row is for the radial component, Mr, whereas the middle and lower rows are for lateral components, Mqand Mf.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
6 of 12
data generated from a synthetic model, why such a conser-vative choice to pick a reliable solution is important.
5. Discussions
[16] The sampling geometry of a particular data set such
as the distribution of track spacing and the observing altitudes inevitably imposes natural limits on the shortest resolvable model wavelengths. Although the general con-sensus is to formulate inverse problems with enough model degrees of freedom to avoid the potential aliasing effect, the actual resolvable model components are intrinsically deter-mined by the sampling geometry and are usually much less than those implied from the resolution presumed by the formulation. The variance reduction versus model variance tradeoff analysis helps to locate the optimal model resolu-tion by offering the appropriate degree of strictness of regularization or damping. In principle, formulations based
on data kernels [e.g., Whaler and Purucker, 2005] are intrinsically free from the concern of nonuniqueness since there will not be annihilators embedded. There are, however, always the problem associated with the noise contamination or observation errors. In other words, proper regularization is still essential to avoid overinterpreting the data. Unlike other previous works that pursue the best data fitting only, Whaler and Purucker [2005] as well as Langlais et al. [2004] invoke the minimization of the RMS magnetization to regularize the inverse problem. However, the model with the least data misfit still seems to be the choice for the preferred model (e.g., Table 2 of Whaler and Purucker [2005]).
[17] Our solutions that fit the data reasonably well have
considerably less and decaying high-degree power spectra (Figure 9) although our spherical mesh, with a mean spacing of about 1.4°, is fully capable of resolving fine details beyond these higher degrees. These solutions are selected based on locating the optimal area around the knee of the variance-reduction versus model-variance tradeoff curve. That is, decaying high-degree power spectra is a consequence of having low model variance while retaining a reasonable data fitting. In other words, fine details corresponding to those high-degree power spectra are rela-tively less robustly constrained by the data.
[18] Notice that there are external as well as internal field
contributions to the data. An inverse problem formulated following equation (1) results in an equivalent source magnetization model that extracts crustal signals as far as it is permissible. Arkani-Hamed [2004] used the radial component of the mapping phase data alone that are believed to be least contaminated by the external field, as well as covariance analysis and comparison between models derived from two subsampled data sets, to suppress the time-varying and noncrustal parts of the models induced by external field. Although he concluded conservatively that the degree 62 is likely an optimum upper limit of the harmonic degrees of the crustal magnetic field that can be resolved by the high-altitude mapping phase MGS data, it is interesting to note that the resulting model, however, has high-degree power higher than Cain et al.’s [2003] model (Figure 9) that is based on a data set including all three components data from AB, SPO and MO phases. In other words, external field contaminations do not seem to be the main factors responsible for the differences of their high-degree power.
[19] We believe there are two major factors that result in
the apparent discrepancies among models established so far. The first factor that differentiates results based on the Crustal magnetization Model (CM), might it be discrete in nature such as the GSFC model [Langlais et al., 2004] or the continuous ones such as the WP model [Whaler and Purucker, 2005] and the model of this study, from those based directly on Spherical Harmonics (SH) can very likely be attributed to the effect of spectral leakage [Trampert and Snieder, 1996; Chiao and Kuo, 2001]. This effect is similar to the aliasing effect when truncated Fourier series is adopted to expand a function with high-degree energy. The high-degree energy that is not properly represented by their actual degrees owing to the truncated expansion will pile up near the truncated degree and distort the actual spectra especially close to the truncated degree. Instead of directly decomposing a function or a time series, the Figure 5. Comparison between different solutions
show-ing the magnified portion of part of the southern hemisphere crustal magnetization zoomed in from Figure 4.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
7 of 12
Figure 6. Comparison between different solutions similar to Figures 4 and 5 but showing the manifested magnetic field at the altitude of 200 km above Martian surface.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
8 of 12
spectral leakage is very similar in essence except that it occurs when less than enough degrees of freedom are adopted for a model parameterization of an inverse problem. It is obvious from this as well as other previous studies that it will take even higher degrees to get strictly numerically better fitting to the MGS data and that it is quite clear that
degree n = 90 is just an arbitrary level of truncation. In other words, when a SH representation truncated at n = 90 is adopted to fit the MGS data, spectral leakage onto those high degrees close to the truncation degree will be inevi-table. The CM models are however, truncated differently. In fact, one needs much higher degrees to completely represent these models in terms of the spherical harmonics expansion. That is, there are still significant power beyond n = 90 for CM models whereas SH models have their powers drasti-cally annihilated reaching beyond n = 90. We believe that this is the main reason that makes the SH models have higher power around n = 90 than the CM models.
[20] The second factor that makes some models having
lower high-degree power than others within the same group is regularization, the key issue that we have been discussing
Figure 7. Histograms of fitting residuals of three solutions of this study. (a) solution_2 using simple damping. (b) solution_2 using multiscale inversion. (c) solution_1
using multiscale inversion (see Figure 3 and the text). Figure 8.solutions of Figure 7 and the Mars Global Surveyor (MGS)Comparison of calculated values of the three observations for a segment of the AB2 low-altitude (shown in the lowest panel) collection period. Small open circles mark the MGS observations; thin line is the prediction from the solution_2 model using simple damping, whereas the dark thick line is the solution_1 using the multiscale inversion. All the variations are plotted as a function of the areodetic latitudes, but the longitudinal range is also shown on the top axis.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
9 of 12
in the current study. To ensure that the solution is reliable, we suggest adjusting the strictness of the imposed regular-ization or damping. Enforced regularregular-ization shaves off poorly constrained model components while sacrificing some degree of goodness of fit. That is, we have reasons other than pursuing just better fitting to choose our preferred model that has high-degree power even lower than other CM models. On the other hand, we believe that the reason for the FSU model [Cain et al., 2003] having much lower high-degree power than the MG model [Arkani-Hamed, 2004] and the NASA model [Connerney et al., 2001] is that the FSU model is based on a more complete data set that reduces the degree of nonuniqueness of the inverse problem. That is, for the same amount of degrees of freedom to be modeled, more data constraints behave similarly as regularization that reduces relatively poorly constrained components and results in less high-degree power. Further comparison of spatial patterns of the surface potential among our preferred models and those established previ-ously demonstrates the fundamental differences that might be results of the two factors mentioned above (Figure 10). Notice that the FSU model (Figure 10b) that is constrained with more data than the MG model (Figure 10a) appears to be much simpler along with much lower high-degree power (Figure 9). That is, it is very likely that a significant portion of those short-scale complexities in the MG model with high-degree power are not robustly resolvable model com-ponents. The WP model (Figure 10d) is in fact constrained by the same data set as the FSU model. So the reason why the WP model bears even less complicated structures than the FSU model is very likely due to its distinct formulation that avoids null space model components from scratch. It is interesting to note that our solution_2 model (Figure 10e) is very similar to the WP model. Whereas our solution_2 model reaches a variance reduction over 92%, the intrinsic model structure is much simpler than those previously established SH models (Figures 10a and 10b). Furthermore, we have reasons to believe that the even simpler structures
in the conservative solution_1 model (Figure 10f) might be more robust. It is also worth mentioning that although the difference between the surface potential models from solu-tion_2 and solution_1 seems to be subtle (Figures 10e and 10f), their manifestations on the crustal magnetization models are in fact significant (Figures 4 and 5).
[21] To further verify the interpretation of the finer detail
features discussed above, we execute recovery experiments with a known implanted synthetic magnetization model. A circular crustal model with constant 20 km depth and alternating positive and negative magnetization in the radial component, Mr, is implanted around the equator
(Figure 11a). There are no assumed lateral, Mq and Mf,
components. The same sampling geometry of the MGS data set is invoked as the observations. That is, the three-component magnetic field intensity data observed at Figure 9. Comparison of the Mauersberger-Lowes power
spectra at the Martian surface between selected previous models (following Whaler and Purucker [2005], we have NASA for the model of Connerney et al. [2001]; MG for the model of Arkani-Hamed [2004]; FSU for the model of Cain et al. [2003]; GSFC for the model of Langlais et al. [2004], and WP for the model of Whaler and Purucker [2005]) and the two solutions obtained in this study.
Figure 10. Comparison of the magnetic potential evalu-ated at the Martian surface among selected previous models and the two solutions obtained in this study: (a) MG, (b) FSU, (c) GSFC, (d) WP, (e) simple damping solution_2 model, and (f) solution_1 model obtained from multiscale regularization (see also the caption of Figure 9).
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
10 of 12
Figure 11. Inversion experiments executed using the same observational sampling geometry but with field intensity data generated from synthetic model and corrupted with noise with N/S about 60%. (a) The implanted synthetic magnetization model with alternating positive and negative magnetization in the radial direction centered at the equator overlaid upon two negative circular sources to the southeast and northwest quadrant. Notice that there are only implanted radial, Mr, component. (b) Inverted result if
solution_3, the simple damping solution of group 3 on the tradeoff curve (Figure 3), is selected. Notice the obvious corruption of the inverted model arisen from the contamination embedded within the data. (c) Similar to Figure 11b, except that it is a solution approaching solution_2 instead of solution_3. Notice the improvement on reducing the corruption of uncorrelated, nonphysical structure. (d) In the preferred multiscale solution_1 model, the effect of the multiscale regularization to annihilate unreliable model components while grouping correlated model structure is obvious. (e, f) Notice that the aliasing onto the Mqand the Mfcomponents is inevitable (check the sensitivity matrix G in Figure 1).
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
11 of 12
different altitudes across Mars are replaced by synthetic data generated from the implanted magnetization structure. A considerable amount of uncorrelated noise with peak amplitude as high as 60% of the peak amplitude of the model generated data is then randomly blended in the synthetic data. Since the sampling geometry is the same, there is no need to carry out a new tradeoff analysis for the synthetic data set. Damping factors for the three groups of solutions marked on the tradeoff curves on Figure 3 are tested to obtain corresponding solutions. Not surprisingly, the recovered, underdamped simple damping solution (solution_3 on the tradeoff diagram shown by Figure 3) is significantly corrupted by manifestation from the uncorre-lated noise added to the data (Figure 11b). The corruption reduces considerably toward solution_2, but it is still significant and interferes with the correct interpretation of the recovered model (Figure 11c). On the other hand, the noise corruption upon the recovered model that corresponds to the multiscale inversion solution_1 is obviously much lower (Figure 11d) and reasonable. What is worth cautioning is the significant aliasing effects onto the Mq
and the Mfcomponents that are not implanted (Figures 11e
and 11f). This is, however, inevitable for any formulation based on equation (1) and is simply unresolvable by data constraints alone.
[22] In summary, the reason to carry out tradeoff analysis
is to serve as an effective way of picking the right degree of regularization and thus the appropriate model components that are robustly constrained by the data. The quality of the actual MGS data is probably much better than the tested synthetic data such that the potential corruption might not be as serious as what is demonstrated in Figure 11. However, overinterpretation or overfitting data with unreli-able model components is prone to misleading results that can be avoided by giving up a small fraction of the relatively less reliable data information.
[23] Acknowledgments. We wish to acknowledge K. A. Whaler and M. E. Purucker for unselfishly sharing all supporting material of their model on the Web: http://planetary-mag.net/jgr_mars_whaler/ [Whaler and Purucker, 2005]. Constructive comments from Joseph Cain and anonymous reviewers have been more than helpful in making significant improvements. All graphs have been created using the Generic Mapping Tools package [Wessel and Smith, 1991]. This study is supported by the National Science Council of ROC under the contracts NSC 95-2611-M-002-004 and NSC 94-2116-M-002-023. See the auxiliary material1 for the computer files used.
References
Acuna, M. H., et al. (1999), Global distribution of crustal magnetization discovered by the Mars Global Surveyor MAG/ER experiment, Science, 284, 790 – 793.
Albee, A. L., F. Palluconi, and T. Thorpe (2001), Overview of the Mars Global Surveyor Mission, J. Geophys. Res., 106, 23,291 – 23,316. Arkani-Hamed, J. (2001), A 50-degree spherical harmonic model of the
magnetic field of Mars, J. Geophys. Res., 106, 23,197 – 23,208.
Arkani-Hamed, J. (2002), An improved 50-degree spherical harmonic model of the magnetic field of Mars derived from high-altitude and low-altitude data, J. Geophys. Res., 107(E10), 5083, doi:10.1029/ 2001JE001835.
Arkani-Hamed, J. (2004), A coherent model of the crustal magnetic field of Mars, J. Geophys. Res., 109, E09005, doi:10.1029/2004JE002265. Backus, G., R. Parker, and C. Constable (1996), Foundations of
Geomag-netism, 369 pp., Cambridge Univ. Press, New York.
Cain, J. C., B. B. Ferguson, and D. Mozzoni (2003), An n = 90 internal potential function of the Martian crustal magnetic field, J. Geophys. Res., 108(E2), 5008, doi:10.1029/2000JE001487.
Chiao, L.-Y., and B.-Y. Kuo (2001), Multiscale seismic tomography, Geo-phys. J. Int., 145, 517 – 527.
Chiao, L.-Y., and W.-T. Liang (2003), Multiresolution parameterization for geophysical inverse problems, Geophysics, 68(1), 1 – 11.
Cohen, A., I. Daubechies, and J.-C. Feauveau (1992), Biorthogonal bases of compactly supported wavelets, Commun. Pure Appl. Math., 45, 485 – 560. Connerney, J. E. P., M. H. Acuna, P. J. Wasilewski, G. Kletetschka, N. F. Ness, H. Reme, R. P. Lin, and D. L. Mitchell (2001), The global magnetic field of Mars and implications for crustal evolution, Geophys. Res. Lett., 28, 4015 – 4018.
Delprat-Jannaud, F., and P. Lailly (1993), Ill-posed and well-posed formu-lations of the reflection travel time tomography problem, J. Geophys. Res., 98, 6589 – 6605.
Hood, L. L., C. N. Young, N. C. Richmond, and K. P. Harrison (2005), Modeling of major martian magnetic anomalies: Further evidence for polar reorientations during the Noachian, Icarus, 177, 144 – 173, doi:10.1016/j.icarus.2005.02.008.
Langlais, B., M. E. Purucker, and M. Mandea (2004), Crustal magnetic field of Mars, J. Geophys. Res., 109, E02008, doi:10.1029/ 2003JE002048.
Lawson, C. L., and R. J. Hanson (1974), Solving Least Squares Problems, 340 pp., Prentice-Hall, Upper Saddle River, N. J.
Mallat, S. (1998), A Wavelet Tour of Signal Processing, 577 pp., Elsevier, New York.
Menke, W. (1984), Geophysical Data Analysis: Discrete Inverse Theory, 260 pp., Elsevier, New York.
Meyerholtz, K. A., G. L. Pavlis, and S. A. Szpakowski (1989), Convolu-tional quelling in seismic tomography, Geophysics, 54, 570 – 580. Paige, C. C., and M. A. Saunders (1982), LSQR: An algorithm for sparse
linear equations and sparse least squares, Trans. Math. Software, 8, 43 – 71.
Parker, R. L. (1994), Geophysical Inverse Theory, 386 pp., Princeton Univ. Press, Princeton, N. J.
Sweldens, W. (1996), The lifting scheme: A custom-design construction of biorthogonal wavelets, Appl. Comput. Harmon. Anal., 3, 186 – 200. Tarantola, A., and A. Nercessian (1984), Three-dimensional inversion
with-out blocks, Geophys. J.R. Astron. Soc., 76, 299 – 306.
Trampert, J., and R. Snieder (1996), Model estimations biased by truncated expansions: Possible artifacts in seismic tomography, Science, 271, 1257 – 1260.
Voorhies, C. V., T. J. Sabaka, and M. E. Purucker (2002), On magnetic spectra of Earth and Mars, J. Geophys. Res., 107(E6), 5034, doi:10.1029/ 2001JE001534.
Wessel, J. K., and H. F. Smith (1991), Free software helps map and display data, Eos Trans. AGU, 72(41), 445 – 446, 441.
Whaler, K. A., and M. E. Purucker (2005), A spatially continuous magne-tization model for Mars, J. Geophys. Res., 110, E09001, doi:10.1029/ 2004JE002393.
Zienkiewicz, O. C., and R. L. Taylor (1991), The Finite Element Method, 4th ed., 807 pp., McGraw-Hill, New York.
L.-Y. Chiao and J.-R. Lin, Institute of Oceanography, National Taiwan University, P.O. Box 23-13, Taipei 106, Taiwan. ([email protected])
Y.-C. Gung, Department of Geosciences, National Taiwan University, Taipei 106, Taiwan.
1Auxiliary materials are available in the HTML. doi:10.1029/ 2006JE002725.
E12010 CHIAO ET AL.: MAGNETIC FIELD MODEL OF MARS
12 of 12
赴國外研究心得報告
計畫編號 95-2622-M-002-004 計畫名稱 台灣西南海域大陸與海洋相互作用的特性-海陸岩圈交界與重力異常及磁 力異常之小波譜值相關(2/3) 出國人員姓名 服務機關及職稱 喬凌雲 (Chiao, Ling-Yun) 臺灣大學理學院海洋研究所教授出國時間地點 May 9-11 , 2007 ; Hotel Birke, Kiel, Germany
國外研究機構
International Data Exchange Workshop 2007 :Building a Global Data Network for Studies of Earth Processes at the World's Plate Boundaries
MARGINS, NSF (美國國家科學基金會大陸邊緣計劃辦公室)
工作記要：
本計劃核定之出國差旅經費原訂參加 2006AGU. 雖然以共同作者提出論文三篇(見附件 一);並擬以通訊作者口頭發表一篇。後接獲美國 MARGINS, NSF 邀請參加 International Data Exchange Workshop 2007 :Building a Global Data Network for Studies of Earth Processes at the World's Plate Boundarie. 由於該工作小組將研商各國學界研究用海域板塊邊緣之觀測資料的 分享共用。一方面涉及我國學者使用他國(尤指美、日)觀測資料;另一方面則由於我國科會海 科中心海洋資料庫已頗有成效,在未來海洋學門指定海研一號貴重儀器中心統籌運作,如何參 與國際地科學界合作將是重要的課題。而且本次會議受邀者俱為全球重要海洋科學中心,亞洲 地區僅邀請我國與日本,機會難得,因此決定參加 2007 年 5 月 9-11 日在德國的研討會。該研 討會參與者均為大型跨國地科計劃之 PI 或資料庫負責人(見附件二),會期雖然不長,但目標明 確討論集中,對於發展國際性板塊邊緣之觀測資料分享共用的文化頗有重大推展。而且我國學 界海域觀測資料之發展與管理也受到一定之矚目。對於嗣後發展改進方向很有啟發。研討會 初步共同工作報告如附件三。 由於研討會對於受邀參與者提供部分補助,其不足部分由海科中心支援。故本年度並未使 用計劃核定之出國差旅經費。
附件一
HR: 0800h AN: S51C-1283
TI: Multi-scale upper mantle tomography of the Eurasia using surface waveform data AU: * Gung, Y
AF: Department of Geosciences, National Taiwan university, P.O. BOX 13-318, Taipei, 106 Taiwan AU: Chiao, L
AF: Institute of Oceanography, National Taiwan university, P.O. BOX 23-13, Taipei, 106 Taiwan AB: We invert long period seismograms in the time domain in the framework of normal-mode-based nonlinear asymptotic coupling theory (NACT) [Li and Romanowicz, 1995] for the seismic structure
underneath Eurasia. While only Eurasia region, where the sensitivity is the highest in the selected data set, is inverted for its radial anisotropic structure, the effects from global 3D heterogeneous structure are taken into account in the forward stage. The implementation of multi-scale inversion is achieved by converting the partial derivative matrices of the initial model parameters, either spherical harmonics or globally distributed spherical triangle meshes, into a multi-resolution wavelet representation in our interested region. In the scheme of multi-resolution wavelet representation, model parameters are grouped into natural hierarchy of local scales such that the damping regularization acts to sort through successive scales depending on the local data constraints. In other word, the spatially nonstationary smoothness enhancement is automatically invoked depending on the in-situ rigors of the model constraints offered from the data. As a result, sites with strong constraints are capable of resolving more details robustly whereas stable long wavelength
components are still available for sparsely constrained area [Chiao and Liang, 2003]. DE: 7200 SEISMOLOGY
DE: 7208 Mantle (1212, 1213, 8124)
DE: 7255 Surface waves and free oscillations SC: Seismology [S]
MN: 2006 Fall Meeting
HR: 14:55h AN: T53F-06
TI: Possible eastern edge of the Meso-Tethyan slab in the lower mantle beneath southern Tibet AU: * Kuo, B
AF: Institute of Earth Sciences, Academia Sinica, POB 1-55 Nankang, Taipei, 10000 Taiwan AU: Lin, P
AF: Institute of Oceanography, National Taiwan University, 1 Roosevelt Road, Taipei, 10000 Taiwan AU: Chiao, L
AF: Institute of Oceanography, National Taiwan University, 1 Roosevelt Road, Taipei, 10000 Taiwan
AB: We analyzed slowness of P, ScP, and PcP recorded by Indepth III and HIMNT temporary arrays in Tibet from events in the Sumatra-Sunda subduction zones. Raypaths of P, ScP, and PcP constitute a wide range of incidence angles that could help sorting the position of an anomaly. Unlike the HIMNT data, the Indepth slowness residuals show significant discrepancy between P and the group of ScP and PcP. The mean of the former is 0.10 s/deg, while the mean of the latter is -0.46 s/deg. The residual vs. incidence angle pattern is robust and excludes systematic velocity variations in the upper mantle as the source of the discrepancy. The contribution from D”can be ruled out by the similarity between ScP and PcP and the extreme magnitude of the residuals. The favored depth range to suit the anomaly is the upper part of the lower mantle where ScP and PcP still remain close but together far enough from the P paths. Models that are consistent with our observations are (1) a horizontal gradient over a distance of 300 km with 1 percent increase in Vp per 100 km towards northwest; and (2) a volumetric anomaly of 2-3 percent with a 90-deg or obtuse corner that is sampled by the bundle of rays to the array. The combination of the two also gives the observed negative residuals. Both (1) and (2) represent a boundary of a subducted slab exposed sideways to the mantle. The broken Meso-Tethyan oceanic lithosphere that was last consumed ~150 Ma is the most likely candidate for this structure.
DE: 1213 Earth's interior: dynamics (1507, 7207, 7208, 8115, 8120) DE: 7203 Body waves
SC: Tectonophysics [T]
MN: 2006 Fall Meeting
HR: 15:10h AN: T53F-07
TI: Imaging Upper Mantle Structure Beneath the Tibetan Plateau and the Himalaya by Multiscale
Finite-Frequency Tomography
AU: * Hung, S
AF: Department of Geosciences, National Taiwan University, Taipei, 106 Taiwan AU: Wang, C
AF: Institute of Oceanography, National Taiwan University, Taipei, 106 Taiwan AU: Chiao, L
AF: Institute of Oceanography, National Taiwan University, Taipei, 106 Taiwan
AB: The Tibetan plateau and the Himalaya, created by the Indo-Asian collision started 50 million years ago, are the classic sites for studies of the evolution of continental orogeny. Determining seismic velocity structure of the underlying crust and mantle is essential for understanding how plate tectonics and mantle dynamics shape the towering Himalaya mountains and flat topography of Tibet. Recent development in finite-frequency tomography has been proven useful in imaging 3-D velocity variations on the scale comparable to the characteristic wavelength of the waves. Using available data from the INDEPTH and HIMNT experiments, we conduct finite-frequency traveltime tomography for compressional wavespeed heterogeneity of the upper mantle beneath Tibet. We measure relative delay times of P-wave arrivals between stations using
multichannel cross-correlation of bandpass-filtered waveforms in different frequency ranges. The measured traveltime delays of the same phase arrival at different frequencies are actually sensitive to individual unique volume of structural heterogeneity surrounding the ray path. Such frequency-dependent sensitivity is
naturally represented by 3-D banana-doughnut shaped Fréchet kernels for tomographic imaging. Moreover, multiscale wavelet-adaptive parameterization is invoked in the inversion and the resulting velocity models have spatially-varying resolutions subject to the quality of data sampling. The preliminary model reveals a region of relatively high P-wave velocity extending continuously from the uppermost mantle to the depth of 350 km beneath central Tibet (30°N--34°N). At depths above 200 km, there is a strong lateral gradient of 3--4% in P wavespeed from high velocity structure beneath the Himalaya to low velocity beneath the Tibetan plateau.
DE: 3260 Inverse theory DE: 7203 Body waves
DE: 7270 Tomography (6982, 8180)
DE: 8102 Continental contractional orogenic belts and inversion tectonics DE: 8120 Dynamics of lithosphere and mantle: general (1213)
SC: Tectonophysics [T]
附件二
Last Name First Name Institution
Abers Geoff MARGINS Office, Boston University (USA) Ahern Tim IRIS (USA)
Al-Habsi Harib Sultan Qaboos University (Oman) Arko Bob LDEO, Columbia University (USA) Bach Wolfgang University of Bremen (Germany)
Baker Maria National Oceanography Centre, Southhampton (UK) Barckhausen Udo BGR-German Geological Survey
Baumann Peter Jacobs University Bremen (Germany)
Beaudoin Yannick UNEP-Continental Shelf Programme (Norway)
Blackman Donna Ridge 2000 Office, Scripps institution of Oceanography (USA) Blower Jon University of Reading (UK)
Briones Katiusca Oceanographic Insitute of Ecuador
Cannat Mathilde Institut de Physique du Globe de Paris (France) Carbotte Suzanne LDEO, Columbia University (USA)
Chen Bob CIESIN, Columbia University (USA) Chiao Ling-Yun National Taiwan University (Taiwan, ROC) Clark Dru University of California, San Diego (USA) Condit Christopher University of California, San Diego (USA)
Cogan Christopher Alfred Wegener Institute for Polar and Marine Research (Germany) Damm Timo University of Kiel (Germany)
Devey Colin IFM-GEOMAR (Germany) Diviacco Paolo
Istituto Nazionale di Oceanografia e di Geofisica Sperimentale (Italy)
Dransch Doris GeoForschungsZentrum Potsdam (Germany)
Fox Christopher National Oceanic and Atmospheric Administration (USA) Galkin Anastasia GeoForschungsZentrum Potsdam (Germany)
Goldstein Steven LDEO, Columbia University (USA) Goodwillie Andrew LDEO, Columbia University (USA)
Graybeal John Monterey Bay Aquarium Research Institute (USA)
Haak Katherine Ridge 2000 Office, Scripps Institute of Oceanography (USA) Halvorsen Oystein UNEP-Continental Shelf Programme (Norway)
Hanafusa Yasunori JAMSTEC (Japan)
Haq Bilal National Science Foundation (USA) Hosseini Keivan Ferdowsi University of Mashhad (Iran) Huettmann Falk University of Alaska, Fairbanks (USA) Javidpour Jamileh IFM-GEOMAR (Germany)
Jones Craig GNS Science (New Zealand)
Kandel Cary MARGINS Office, Boston University (USA) Khan Shuhab University of Houston (USA)
Klump Jens GeoForschungsZentrum Potsdam (Germany) Le Bas Tim National Oceanography Centre (UK)
Lehnert Kerstin LDEO, Columbia University (USA)
Lezaeta Pamela MARGINS Office, Boston University (USA) Lowry Roy British Oceanographic Data Centre (UK) Matsuda Shigemi The Center for Deep Earth Exploration (Japan) Meier Thomas Bochum University (Germany)
Miller Stephen University of California, San Diego (USA) Miville Bernard International Ocean Drilling Program (Japan) Moussat Eric IFREMER (France)
Neben Soenke BGR-German Geological Survey Nygard Atle University of Bergen (Norway)
Pirenne Benoát University of Victoria (Canada) Ramirez-Llodr
a Eva CMIMA-CSIC (Spain) Ranero Cesar CMIMA-CSIC (Spain)
Ryan William LDEO, Columbia University (USA) Sadeghi Hossein Ferdowsi University of Mashhad (Iran) Salters Vincent Florida State University (USA)
Sarbas Baerbel Max-Planck-Institute for Chemistry (Germany) Schaap Dick SeaDataNet (Netherlands)
Schaefer Angela Jacobs University Bremen (Germany) Schirnick Carsten IFM-GEOMAR (Germany)
Schoolmeeste
r Tina UNEP-Shelf Programme (Norway)
Shiomi Katsuhiko National Research Institute for Earth Science and DisasterPrevention (Japan) Shipley Thomas University of Texas (USA)
Stransky Julia IFM-GEOMAR (Germany) Trueger Mickael IFREMER (France)
Tsuboi Seiji JAMSTEC-IFREE (Japan)
Unnithan Vikram Jacobs University Bremen (Germany) Venuti Fabio National Oceanography Centre (UK) Wallrabe-Ada
ms Hans-Joachim University of Bremen (Germany)
Weatherall Pauline British Oceanographic Data Centre (UK) Weinrebe Wilhelm IFM-GEOMAR (Germany)
附件三
Bui
l
di
ng
a
Gl
obal
Dat
a
Ne
t
wor
k
f
or
St
udi
e
s
of
Ear
t
h
Pr
oc
e
s
s
e
s
at
t
he
Wor
l
d’
s
Plate Boundaries
Draft Workshop Report
Executive Summary
An international group of scientists, data managers, and information technologists convened for a 2.5 day meeting in Kiel, Germany, to explore the opportunities for international data exchange and to address the cultural and political challenges to building a global data network that facilitates mid-ocean ridge and continental margin related research internationally. Workshop participants discussed technical, procedural, and organizational issues of global data sharing, and reached agreement on the following set of recommendations grouped broadly under the working group themes:
Science User Needs
Open access to data is fundamental to verifiable scientific progress. All data that are necessary
to reproduce published scientific results, including field data, processed data, and laboratory (derived) data products, need to be published and archived in accepted archives. We need to
advance a culture among scientists that is more open to data sharing. (T1-R11; T2-R5; T4-R4)
Scientists require access to multidisciplinary data and data integrated from both the marine and
terrestrial world. (T1-R2; T1-R3) Data Documentation and Publication
Uniform best practices and standards need to be developed and used routinely within the
international community for data acquisition, data submission to data centers, and data publication. Best practices should include formal submission agreements between individual institutions and respective inter/national data centers and the use of globally unique identifiers for data. Scientific societies should take on an active role in formulating best practice guidelines for the publications of data. In addition, new mechanisms are needed to track the use of data
sets both to ensure academic recognition and to support scientific collaborations. (T1-R4;
T2-R2; T2-R4; T2-R6; T2-R7; T2-R8; T4-R1)
The ultimate responsibility for ensuring adequate documentation of a field program lies with
scientists. Metadata creation and data submission should be made as easy as possible for ship
1
operators and scientistswith development of new tools for automation to support and further the implementation of best practices and standards. Funding agencies must be involved in enforcing
standard practices for data documentation and submission to data centers. (T2-R1; T2-R3;
T2-R4; T2-R6)
Data and Metadata Interoperability
The community must minimize the proliferation of metadata standards and work toward a
uniform approach for scientific metadata. Processes need to be defined regarding how to develop community-based standards, guidance, and profiles. New efforts to develop standards and protocols to support interoperability should build upon and take advantage of existing community-based projects. (T3-R1; T3-R2; T3-R3; T3-R4)
Development of a data discovery service across distributed marine geoscience data resources
within the international community is an achievable initial goal. Data centers should work to expose their data resources via web services using e.g. OGC or OAI protocols. (T3-R5; T3-R6)
Opportunities and Obstacles for International Data Sharing
International programmes and bodies such as GEOSS, the eGY and ICSU that stimulate the
development of global data sharing systems should be leveraged to promote an initiative for a global data network for marine and terrestrial geoscience data. (T4-R5)
A dedicated task group should be established to advance the implementation of a global data
network. In addition, special interest groups to share experience and solutions on issues concerning metadata and interfaces should be formed with tools to facilitate collaboration.
(T4-R6; T3-R7)
Based on these recommendations, the following next steps are identified; 1. Develop test bed sites for a data discovery service across globally distributed data resources; 2. Establish forums for guidance and development of best practises in the areas of data acquisition, metadata, vocabularies, and interfaces; 3. Formulate a dedicated task group to advance international alliances; 4. Convene a follow-up meeting in one year.
1. Motivation for the Workshop
Over the past decade, rapid advances have occurred in database technology for scientific research providing new access to data and new tools for data visualization and integration. Along with these advances in Information Technology has come the growth of digital data collections for a broad suite of data related to the marine geosciences. Developments in database connectivity provide new opportunities for open exchange of data across distributed data collections, greatly expanding the volume and diversity of data available to the scientist to address a particular scientific problem of interest. These advances hold great promise for the solid earth sciences, an inherently multi-national and multi-disciplinary field, which involves the collection of typically
unique data sets during oceanic and terrestrial expeditions conducted by research institutions around the globe.
The international marine geoscience community is actively engaged in scientifically aligned goals through the InterRidge-Ridge2000 programs and InterMARGINS-MARGINS programs. These programs represent broad multi-disciplinary initiatives focused on understanding the fundamental processes of crustal formation, modification and destruction at the world’s plate boundaries. InterRidge and InterMARGINS are international programs, which aim to coordinate efforts and priorities in mid-ocean ridge and continental margin research, respectively, across nations. Ridge2000 and MARGINS are aligned US-funded programs which conduct focused investigations in a few geographic locations, most of which involve international partners. At present there are no formal agreements for data sharing within these international communities, and data exchange occurs primarily by informal agreements between scientists directly involved in specific projects. However significant benefits to these linked marine-terrestrial geoscience research efforts internationally could be achieved if data collections maintained as national efforts could be better linked and if broader access were initiated. New database technologies are available that enable independent globally distributed sites to share, link, and integrate their data holdings and services while maintaining full ownership and credit for these holdings.
To explore current opportunities and challenges for international data exchange to support continental margin and mid-ocean ridge research,theworkshop entitled “Building aglobaldata network forstudiesofearth processesattheworld’splateboundaries”wasconvened with two primary goals:
1. To explore current data management efforts relevant to continental margin and mid-ocean ridge
science goals within partner countries.
2. To devise a strategy for building a global data network to support the sharing and exchange of
data of greatest scientific interest for continental margin and mid-ocean ridge studies.
The primary hoped for outcome of this meeting was the development of new partnerships between marine geoscientists and data centers within the international community to establish greater access and exchange of data sets of broad interest for continental margins and mid-ocean ridge research.
2. Workshop Structure
The workshop was convened by four scientists from Germany, Japan, and the United States from within the InterMARGINS and InterRidge communities, and jointly sponsored by InterMARGINS, MARGINS, InterRidge and Ridge2000. The US National Science Foundation and theGerman project“TheFutureOcean”provided additionalfinancialsupport.About70 people from 14 countries attended the workshop, including scientists from the InterRidge and
InterMARGINS communities, data managers representing data centers and data systems across a spectrum of geoscience data, and information technologists involved in various aspects of interoperability development.
The workshop was held at the meeting facilities of the Hotel Birke in Kiel, Germany. The official program started on May 9 in the morning and lasted for 2.5 days. Interested participants were invited to continue discussions on May 11 in the afternoon. The workshop ended with an informal field trip to the historical town of Lübeck on May 12.
The first 1.5 days of the workshop were devoted to presentations within three general areas: (a) Science Needs: Scientists outlined their needs for data access and defined data sets of broad interest for continental margin and ridge-related science.
(b) Data Resources: Representatives of data centers presented existing data systems available for academic research. These presentations were complemented by poster presentations and live demonstrations of the systems.
(c) Technologies: Information technologists reported about emerging technologies for interoperability and data sharing.
The afternoon of Day 2 and morning of Day 3 were devoted to working group sessions to discuss technological as well as organizational and cultural issues of global data exchange. The working group discussionswere structured into fourthemes,each ofwhich (exceptforthe‘Science UserNeeds’group)had two sessions:
1. Science User Needs & Concerns 2. Data Documentation and Publication
a. Standards for Data Documentation b. Data Publication
3. Data and Metadata Interoperability
a. Standards & Technologies for Metadata and Interfaces b. The Low-Hanging Fruit for Data Exchange
4. Opportunities and Obstacles for International Data Sharing
a. Archives and Data Contributions
b. Implementing an International Data Network
Each working group addressed a range of questions provided to the session leaders by the workshop conveners, and was charged to generate a set of recommendations working group leaders presented in plenary sessions. Questions and recommendations are outlined in the following section.
3. Working Group Discussions
3.1. Theme 1:
Science User Needs & Concerns
Scientists engaged in plate boundary research study the wide variety of active processes associated with the formation, modification, and destruction of the crustal layer of the earth, which supports life on the planet. Plate boundaries transect the oceans, hug the continental margins, and penetrate into continental interiors. They are the locus of most earthquake and volcanic activity on earth and of the pervasive fluid-chemical-thermal interactions associated with the development of unique ecosystems and the formation of economical metal deposits. Increasingly, these active environments are studied as integrated complex physical, chemical and biological systems, subject to a variety of influences, rather than as primarily geological structures. To address these interdisciplinary goals, scientists increasingly require access to multidisciplinary data sets and from both the terrestrial and marine setting. These needs represent unique requirements and challenges for scientific data access and exchange.
The science user working group considered the following questions;
What are science user needs and concerns with regard to data sharing? What are the key data sets needed for international exchange?
What links exist and are desired between the marine and terrestrial world?
What capabilities are desired that are currently lacking? What technologies are promising to
scientists?
There is strong endorsement within the science user community of the principle of fully open access to data. Scientists desire access to all existing data relevant for the problem they wish to address. For programs conducted in the open ocean, scientists desire access to everything collected in a geographic area of study. Closer to shore, along the continental margins, there may be economic or national security concerns that affect access to some kinds of data, but much data of value to basic science should be available. Easy access to a diverse suite of data is necessary for many studies (Table1). However, many of the data resources currently available represent disciplinary databases. More focus is needed on building data systems to support integrative science, providing access to multi-disciplinary data. Although the fundamental science questions associated with continental margin studies transect the shoreline, the shoreline represents a major boundary in how data are collected, organized and later archived. This disparity is a significant obstacle to scientific data access.
Recommendations
T1-R1 Open data access is fundamental to verifiable scientific progress. Full open access to
data is needed, first and foremost to support scientific progress but also very importantly, to enable the verification of research results. Geosciences in general relies on a unique set of field
observations, so differs from most experimental sciences in that most measurements are difficult to repeat. With the typically unique data sets used to support plate boundary studies, research results are often impossible to verify without open access to field observations and measurements.
Scientists want unrestricted access to all data as feasible within the framework of national requirements and proprietary periods of data collectors. National needs may require limitations for some data types and in some environments (eg. Ultra-high resolution bathymetry in shallow coastal waters, on-land gravity, reflection seismics in petroleum-rich basins), but every reasonable effort should be made to release such data in a reasonable time frame. For research data subject to proprietary hold periods, scientists would like access to metadata describing the existence and location of the data at an early stage with mechanisms that support interactions between data collectors and other scientists wishing to form collaborations.
T1-R2. Scientists require access to multidisciplinary data. The integrative science programs
that characterize modern studies at mid-ocean ridges and continental margins drive the need for integrated access to multidisciplinary data. More and more, scientists seek to work across traditional disciplinary boundaries either through developing collaborations or by acquiring interdisciplinary expertise. Data systems, which support and facilitate collaborations and multidisciplinary access are required. Scientists need access to multi-disciplinary databases of geographically referenced data as well as of physical property measurements such as experimentally-derived material properties. Derived data sets including images, and data-based models have tremendous value for interdisciplinary studies and these need to be preserved.
T1-R3. Integration of data resources from both the marine and terrestrial world is needed.
Research along the continental margins requires access to both terrestrial and marine data. However, available data resources typically stop at the shoreline with different agencies and organizations involved in terrestrial and marine studies. Significant obstacles to obtaining access to data across the shoreline relates to differences in how data are collected and organized. Whereas offshore work is usually defined and organized by cruise, onshore field studies are characterized in a variety of ways, by networks of instruments, by investigating group, by national or other geographic boundaries, or otherwise. Also, onshore and near-shore data sets tend to be spread through a wide array of national agencies with varying standards and missions. Data systems are needed which support the ability to search for and find related data objects in a variety of different frameworks that make sense for the problem at hand, not always dependent on the platform or group collecting the data. While geographic data access makes sense for many problems, time-series data inherently require the need for searches at a wide variety of time scales, and the wide variety of characteristics of different data sets indicate other primary search categories may be valuable as well.
T1-R4: Mechanisms are needed to track the use of data sets both to ensure academic
recognition and to support scientific collaborations. While the existence of open data collections
representing the accumulation of data from many individual studies provides important resources for scientists, an ongoing concern is how to ensure that credit to original data collectors is preserved.
