Difference between revisions of "IC SG1"

From Icctwiki
Jump to: navigation, search
(Objectives)
(Introduction)
(9 intermediate revisions by 2 users not shown)
Line 1: Line 1:
<big>'''JSG 0.10: High-rate GNSS'''</big>
+
<big>'''JSG 0.1: Application of time-series analysis in geodesy'''</big>
  
Chair: ''Mattia Crespi (Italy)''<br>
+
Chair: ''W. Kosek (Poland)''<br>
Affiliation:''Commissions 1, 3 4 and GGOS''
+
Affiliation:''GGOS, all commissions''
  
 
__TOC__
 
__TOC__
 
 
===Introduction===
 
===Introduction===
  
Global Navigation Satellite Systems (GNSS) have become for a long time an indispensable tool to get accurate and reliable information about positioning and timing; in addition, GNSS are able to provide information related to physical properties of media passed through by GNSS signals. Therefore, GNSS play a central role both in geodesy and geomatics and in several branches of geophysics, representing a cornerstone for the observation and monitoring of our planet.
+
Observations provided by modern space geodetic techniques (geometric and gravimetric) deliver a global picture of dynamics of the Earth. Such observations are usually represented as time series which describe (1) changes of surface geometry of the Earth due to horizontal and vertical deformations of the land, ocean and cryosphere, (2) fluctuations in the orientation of the Earth divided into precession, nutation, polar motion and spin rate, and (3) variations of the Earth’s gravitational field and the centre of mass of the Earth. The vision and goal of GGOS is to understand the dynamic Earth’s system by quantifying our planet’s changes in space and time and integrate all observations and elements of the Earth’s system into one unique physical and mathematical model. To meet the GGOS requirements, all temporal variations of the Earth’s dynamics – which represent the total and hence integral effect of mass exchange between all elements of Earth’s system including atmosphere, ocean and hydrology – should be properly described by time series methods.
  
So, it is not surprising that, from the very beginning of the GNSS era, the goal was pursued to widen as much as possible the range in space (from local to global) and time (from short to long term) of the observed phenomena, in order to cover the largest possible field of applications, both in science and in engineering; two complementary, but primary as well, goals were, obviously, to get these information with the highest accuracy and in the shortest time.  
+
Various time series methods have been applied to analyze such geodetic and related geophysical time series in order to better understand the relation between all elements of the Earth’s system. The interactions between different components of the Earth’s system are very complex, thus the nature of the considered signals in the geodetic time series is mostly wideband, irregular and non-stationary. Therefore, the application of time frequency analysis methods based on wavelet coefficients – e.g. time-frequency cross-spectra, coherence and semblance – is necessary to reliably detect the features of the temporal or spatial variability of signals included in various geodetic data, and other associated geophysical data.
  
The advances in technology and the deployment of new constellations, after GPS (in the next years will be completed the European Galileo, the Chinese Beidou and the Japanese QZSS) remarkably contributed to transform this three-goals dream in reality, but still remain significant challenges when very fast phenomena have to be observed, mainly if real-time results are looked for.
+
Geodetic time series may include, for instance, temporal variations of site positions, tropospheric delay, ionospheric total electron content, masses in specific water storage compartments or estimated orbit parameters as well as surface data including gravity field, sea level and ionosphere maps. The main problems to be scrutinized concern the estimation of deterministic (including trend and periodic variations) and stochastic (non-periodic variations and random fluctuations) components of the time series along with the application of the appropriate digital filters for extracting specific components with a chosen frequency bandwidth. The application of semblance filtering enables to compute the common signals, understood in frame of the time-frequency approach, which are embedded in various geodetic/geophysical time series.
  
Actually, for almost 15 years, starting from the noble birth in seismology, and the very first experiences in structural monitoring, high-rate GNSS has demonstrated its usefulness and power in providing precise positioning information in fast time-varying environments. At the beginning, high-rate observations were mostly limited at 1 Hz, but the technology development provided GNSS equipment (in some cases even at low-cost) able to collect measurements at much higher rates, up to 100 Hz, therefore opening new possibilities, and meanwhile new challenges and problems.
+
Numerous methods of time series analysis may be employed for processing raw data from various geodetic measurements in order to promote the quality level of signal enhancement. The issue of improvement of the edge effects in time series analysis may also be considered. Indeed, they may either affect the reliability of long-range tendency (trends) estimated from data or the real-time processing and prediction.
  
So, it is necessary to think about how to optimally process this potential huge heap of data, in order to supply information of high value for a large (and likely increasing) variety of applications, some of them listed hereafter without the claim to be exhaustive: better understanding of the geophysical/geodynamical processes mechanics; monitoring of ground shaking and displacement during earthquakes, also for contribution to tsunami early warning; tracking the fast variations of the ionosphere; real-time controlling landslides and the safety of structures; providing detailed trajectories and kinematic parameters (not only position, but also velocity and acceleration) of high dynamic platforms such as airborne sensors, high-speed terrestrial vehicles and even athlete and sport vehicles monitoring.
+
The development of combination strategies for time- and space-dependent data processing, including multi-mission sensor data, is also very important. Numerous observation techniques, providing data with different spatial and temporal resolutions and scales, can be combined to compute the most reliable geodetic products. It is now known that incorporating space variables in the process of geodetic time series modelling and prediction can lead to a significant improvement of the prediction performance. Usually multi-sensor data comprises a large number of individual effects, e.g., oceanic, atmospheric and hydrological contributions. In Earth system analysis one key point at present and in the future will be the development of separation techniques. In this context principal component analysis and related techniques can be applied.
  
Further, due to the contemporary technological development of other sensors (hereafter referred as ancillary sensors) related to positioning and kinematics able to collect data at high-rate (among which MEMS accelerometers and gyros play a central role, also for their low-cost), the feasibility of a unique device for high-rate observations embedding GNSS receiver and MEMS sensors is real, and it open, again, new opportunities and problems, first of all related to sensors integration.
+
===Objectives===
  
All in all, it is clear that high-rate GNSS (and ancillary sensors) observations represent a great resource for future investigations in Earth sciences and applications in engineering, meanwhile stimulating a due attention from the methodological point of view in order to exploit their full potential and extract the best information. This is the why it is worth to open a focus on high-rate (and, if possible, real-time) GNSS within ICCT.
+
* To study geodetic time series and their geophysical causes in different frequency bands using time series analysis methods, mainly for better understanding of their causes and prediction improvement.
 +
* The evaluation of appropriate covariance matrices corre-sponding to the time series by applying the law of error propagation, including weighting schemes, regulariza-tion, etc.
 +
* Determining statistical significance levels of the results obtained by different time series analysis methods and algorithms applied to geodetic time series.
 +
* The comparison of different time series analysis methods and their recommendation, with a particular emphasis put on solving problems concerning specific geodetic data.
 +
* Developing and implementing the algorithms – aiming to seek and utilize spatio-temporal correlations – for geo-detic time series modelling and prediction.
 +
* Better understanding of how large-scale environmental processes, such as for instance oceanic and atmospheric oscillations and climate change, impact modelling strate-gies employed for numerous geodetic data.
 +
* Developing combination strategies for time- and space-dependent data obtained from different geodetic observa-tions.
 +
* Developing separation techniques for integral measure-ments in individual contributions.
  
===Objectives===
 
 
* To realize the inventories of: <br /> 1- the available and applied methodologies for high-rate GNSS, in order to highlight their pros and cons and the open problems, 2- the present and wished applications of high-rate GNSS for science and engineering, with a special concern to the estimated quantities (geodetic, kinematic, physical), in order to focus on related problems (still open and possibly new) and draw future challenges, 3- the technology (hw, both for GNSS and ancillary sensors, and sw, possibly FOSS), pointing out what is ready and what is coming, with a special concern for the supplied observations and for their functional and stochastic modeling <br /> with the by-product of establishing a standardized terminology
 
* To address known (mostly cross-linked) problems related to high-rate GNSS as (not an exhaustive list): revision and refinement of functional and stochastic models; evaluation and impact of observations time-correlation; impact of multipath and constellation change; outliers detection and removal; issues about GNSS constellations interoperability; ancillary sensors evaluation, cross-calibration and  integration
 
* To address the new problems and future challanges arised from the inventories
 
* To investigate about the interaction with present real-time global (IGS-RTS, EUREF-IP, etc.) and regional/local positioning services: how can these services support high-rate GNSS observations and, on reverse, how can they benefit of high-rate GNSS observations
 
  
 
===Program of activities===
 
===Program of activities===
 
+
Updating the webpage, so that the information on time series analysis and its application in geodesy (including relevant multidisciplinary publications and the unification of terminology applied in time series analysis) will be available.
* To launch a questionnaire for the above mentioned inventory of methodologies, applications and technologies.
+
Participating in working meetings at the international sym-posia and presenting scientific results at the appropriate sessions.
* To open a web page with information concerning high-rate GNSS and its wide applications in science and engineering, with special emphasis on exchange of ideas, provision and updating bibliographic list of references of research results and relevant publications from different disciplines.
+
Collaboration with other working groups dealing with geo-detic time-series e.g. Cost ES0701 Improved constraints on models of GIA or the Climate Change Working Group.
* To launch the proposal for two (one science and the other engineering oriented) state-of-the-art review papers in high-rate GNSS co-authored by the JSG Members.
 
* To organize a session at the forthcoming Hotine-Marussi symposium.
 
* To promote sessions and presentation of the research results at international symposia both related to Earth science (IAG/IUGG, EGU, AGU, EUREF, IGS) and engineering (workshops and congresses in structural and geotechnical engineering).
 
  
 
===Members===
 
===Members===
  
'' '''Mattia Crespi (Italy), chair''' <br /> Juan Carlos Baez (Chile) <br /> Elisa Benedetti (United Kingdom) <br /> Geo Boffi (Switzerland) <br /> Gabriele Colosimo (Switzerland) <br /> Athanasios Dermanis (Greece) <br /> Roberto Devoti (Italy) <br /> Jeff Freymueller (USA) <br /> Joao Francisco Galera Monico (Brazil) <br /> Jianghui Geng (Germany) <br /> Kosuke Heki (Japan) <br /> Melvin Hoyer (Venezuela) <br /> Nanthi Nadarajah (Australia) <br /> Yusaku Ohta (Japan) <br /> Ruey-Juin Rau (Taiwan) <br /> Eugenio Realini (Italy) <br /> Chris Rizos (Australia) <br /> Nico Sneeuw (Germany) <br /> Peiliang Xu (Japan) <br />''
+
'' '''W. Kosek (Poland), chair'''<br /> R. Abarca del Rio (Chile)<br /> O. Akyilmaz (Turkey)<br /> J. Böhm (Austria)<br /> L. Fernandez (Argentina)<br /> R. Gross (USA)<br /> M. Kalarus (Poland)<br /> M. O. Karslioglu (Turkey)<br /> H. Neuner (Germany)<br /> T. Niedzielski (Poland)<br /> S. Petrov (Russia)<br /> W. Popinski (Poland)<br /> M. Schmidt (Germany)<br /> M. van Camp (Belgium)<br /> O. de Viron (France)<br /> J. Vondrák (Czech Republic)<br /> D. Zheng (China)<br /> Y. Zhou (China)<br />''

Revision as of 09:59, 2 July 2012

JSG 0.1: Application of time-series analysis in geodesy

Chair: W. Kosek (Poland)
Affiliation:GGOS, all commissions

Introduction

Observations provided by modern space geodetic techniques (geometric and gravimetric) deliver a global picture of dynamics of the Earth. Such observations are usually represented as time series which describe (1) changes of surface geometry of the Earth due to horizontal and vertical deformations of the land, ocean and cryosphere, (2) fluctuations in the orientation of the Earth divided into precession, nutation, polar motion and spin rate, and (3) variations of the Earth’s gravitational field and the centre of mass of the Earth. The vision and goal of GGOS is to understand the dynamic Earth’s system by quantifying our planet’s changes in space and time and integrate all observations and elements of the Earth’s system into one unique physical and mathematical model. To meet the GGOS requirements, all temporal variations of the Earth’s dynamics – which represent the total and hence integral effect of mass exchange between all elements of Earth’s system including atmosphere, ocean and hydrology – should be properly described by time series methods.

Various time series methods have been applied to analyze such geodetic and related geophysical time series in order to better understand the relation between all elements of the Earth’s system. The interactions between different components of the Earth’s system are very complex, thus the nature of the considered signals in the geodetic time series is mostly wideband, irregular and non-stationary. Therefore, the application of time frequency analysis methods based on wavelet coefficients – e.g. time-frequency cross-spectra, coherence and semblance – is necessary to reliably detect the features of the temporal or spatial variability of signals included in various geodetic data, and other associated geophysical data.

Geodetic time series may include, for instance, temporal variations of site positions, tropospheric delay, ionospheric total electron content, masses in specific water storage compartments or estimated orbit parameters as well as surface data including gravity field, sea level and ionosphere maps. The main problems to be scrutinized concern the estimation of deterministic (including trend and periodic variations) and stochastic (non-periodic variations and random fluctuations) components of the time series along with the application of the appropriate digital filters for extracting specific components with a chosen frequency bandwidth. The application of semblance filtering enables to compute the common signals, understood in frame of the time-frequency approach, which are embedded in various geodetic/geophysical time series.

Numerous methods of time series analysis may be employed for processing raw data from various geodetic measurements in order to promote the quality level of signal enhancement. The issue of improvement of the edge effects in time series analysis may also be considered. Indeed, they may either affect the reliability of long-range tendency (trends) estimated from data or the real-time processing and prediction.

The development of combination strategies for time- and space-dependent data processing, including multi-mission sensor data, is also very important. Numerous observation techniques, providing data with different spatial and temporal resolutions and scales, can be combined to compute the most reliable geodetic products. It is now known that incorporating space variables in the process of geodetic time series modelling and prediction can lead to a significant improvement of the prediction performance. Usually multi-sensor data comprises a large number of individual effects, e.g., oceanic, atmospheric and hydrological contributions. In Earth system analysis one key point at present and in the future will be the development of separation techniques. In this context principal component analysis and related techniques can be applied.

Objectives

  • To study geodetic time series and their geophysical causes in different frequency bands using time series analysis methods, mainly for better understanding of their causes and prediction improvement.
  • The evaluation of appropriate covariance matrices corre-sponding to the time series by applying the law of error propagation, including weighting schemes, regulariza-tion, etc.
  • Determining statistical significance levels of the results obtained by different time series analysis methods and algorithms applied to geodetic time series.
  • The comparison of different time series analysis methods and their recommendation, with a particular emphasis put on solving problems concerning specific geodetic data.
  • Developing and implementing the algorithms – aiming to seek and utilize spatio-temporal correlations – for geo-detic time series modelling and prediction.
  • Better understanding of how large-scale environmental processes, such as for instance oceanic and atmospheric oscillations and climate change, impact modelling strate-gies employed for numerous geodetic data.
  • Developing combination strategies for time- and space-dependent data obtained from different geodetic observa-tions.
  • Developing separation techniques for integral measure-ments in individual contributions.


Program of activities

Updating the webpage, so that the information on time series analysis and its application in geodesy (including relevant multidisciplinary publications and the unification of terminology applied in time series analysis) will be available. Participating in working meetings at the international sym-posia and presenting scientific results at the appropriate sessions. Collaboration with other working groups dealing with geo-detic time-series e.g. Cost ES0701 Improved constraints on models of GIA or the Climate Change Working Group.

Members

W. Kosek (Poland), chair
R. Abarca del Rio (Chile)
O. Akyilmaz (Turkey)
J. Böhm (Austria)
L. Fernandez (Argentina)
R. Gross (USA)
M. Kalarus (Poland)
M. O. Karslioglu (Turkey)
H. Neuner (Germany)
T. Niedzielski (Poland)
S. Petrov (Russia)
W. Popinski (Poland)
M. Schmidt (Germany)
M. van Camp (Belgium)
O. de Viron (France)
J. Vondrák (Czech Republic)
D. Zheng (China)
Y. Zhou (China)