| DOI: | 10.1016/j.tecto.2019.228299
|
| 论文题名: | Sensitivity of gravity and topography regressions to earth and planetary structures |
| 作者: | Pivetta T.; Braitenberg C.
|
| 刊名: | Tectonophysics
|
| ISSN: | 00401951
|
| 出版年: | 2020
|
| 卷: | 774 | | 语种: | 英语
|
| 中文关键词: | Crustal units
; Global gravity field
; Isostasy
; Mars
; Planets
; Topography
|
| 英文关键词: | Gravitation
; Inverse problems
; Learning algorithms
; Machine learning
; Planets
; Regression analysis
; Topography
; Characteristic value
; Continental crusts
; Crustal units
; Gravity field
; Isostasy
; Lithosphere structure
; Mars
; Regression parameters
; Planetary surface analysis
; algorithm
; crustal structure
; gravity field
; isostasy
; machine learning
; Mars
; planetary surface
; regression analysis
; sensitivity analysis
; topographic effect
|
| 英文摘要: | The availability of global gravity fields and topography through calculation services like the International Centre for Global Earth Models, allows easy access to gravity data, greatly enlarging the spectrum of users. The applications extend much farther than the classic modeling through the gravity-specialist. We investigate the sensitivity of the joint analysis of topography and gravity data based on linear regression analysis and clustering of the response to particular characteristics of the lithosphere structure. The parameters of the regression analysis are predicted to have characteristic values, which allow to distinguish continental crust from oceanic crust, and signalize the presence of crustal inhomogeneity. Predictions are made through theoretical considerations and on synthetic models. We use the South Atlantic Ocean and the confining South American and African continents for illustration, where the regression parameters distinguish oceanic crust from the ridge up to the bathymetric inflection point, from the transitional crust and the continental crust, allowing to map these units. The general properties of the parameters are statistically relevant, since the errors on the parameters are <10% the amplitude of the parameters. We compare the regression parameters with those produced by a global crustal model (CRUST1.0), and find good correspondence between the observed and predicted fields. The analysis can be applied with machine learning algorithms, without the need of specific forward or inverse gravity modeling skills. It is therefore particularly useful in view of the enhanced access to the data through the calculation service, and could be implanted as an add-on tool, since it allows to efficiently distinguish isostatic contribution to the gravity field from crustal sources. Given the experience on the gravity field of the Earth, the analysis can be analogously extended to other planets. For illustration, we show that for Mars a coherent class of Martian crust can be identified. © 2019 |
| Citation statistics: |
|
| 资源类型: | 期刊论文
|
| 标识符: | http://119.78.100.158/handle/2HF3EXSE/170681
|
| Appears in Collections: | 气候变化与战略
|
|
There are no files associated with this item.
|
|
作者单位: | Department of Mathematics and Geosciences, University of Trieste, Via Weiss 1, Trieste, Italy
|
| Recommended Citation: | |
Pivetta T.,Braitenberg C.. Sensitivity of gravity and topography regressions to earth and planetary structures[J]. Tectonophysics,2020-01-01,774
|
|
|