pp 1-12 | Cite as
Estimation of Lesser Antilles Vertical Velocity Fields Using a GNSS-PPP Software Comparison
- 253 Downloads
Vertical land motion in insular areas is a crucial parameter to estimate the relative sea-level variations which impact coastal populations and activities. In subduction zones, it is also a relevant proxy to estimate the locking state of the plate interface. This motion can be measured using Global Navigation Satellite Systems (GNSS), such as the Global Positioning System (GPS). However, the influence of the processing software and the geodetic products (orbits and clock offsets) used for the solution remains barely considered for geophysics studies.
In this study, we process GNSS observations of Guadeloupe and Martinique network (Lesser Antilles). It consists of 40 stations over a period of 18 years for the oldest site. We provide an updated vertical velocity field determined with two different geodetic software, namely EPOS (Gendt et al, GFZ analysis center of IGS–Annual Report. IGS 1996 Annual Report, pp 169–181, 1998) and GINS (Marty et al, GINS: the CNES/GRGS GNSS scientific software. In: 3rd International colloquium scientific and fundamental aspects of the Galileo programme, ESA proceedings WPP326, vol 31, pp 8–10, 2011) using their Precise Point Positioning modes. We used the same input models and orbit and clock offset products to maintain a maximum of consistency, and then compared the obtained results to get an estimation of the time series accuracy and the software influence on the solutions. General consistency between the solutions is noted, but significant velocity differences exist (at the mm/yr level) for some stations.
KeywordsLesser Antilles Guadeloupe Martinique GNSS Vertical velocity field Subsidence Processing comparison Time series analysis
Nowadays, the Global Navigation Satellite Systems (GNSS) have become an indispensable tool to monitor the Earth crust motion. Nevertheless, their use in some part of the world remains challenging. It is the case of the Lesser Antilles Subduction Zone, at the convergence of the Nord-American Plate under the Caribbean Plate. This subduction is singular on several aspects: it is one of the slowest in the world (∼2cm/yr), the lack of emerged lands prevents the determination of a complete deformation profile like it can be done in other areas. Moreover, the islands of the volcanic arc are located too far from the trench, prohibiting most often the detection of significant velocity gradients with respect to the stable plate. Because of these reasons, the locking state deduced from GNSS observations, and thus megathrust risk is uncertain. Symithe et al. (2015) estimated with GNSS observations all along the volcanic arc a low coupling rate but did not exclude a megathrust possibility either. Since the estimation of a horizontal deformation rate is a difficult task for this area, vertical motion observations can then become a proxy to help the assessment of a potential strain accumulation (e.g. Mouslopoulou et al. 2016). Moreover, island areas are also threatened by the sea level rise, and extra subsidence can be an aggravating factor (Ballu et al. 2011). For these two reasons, measuring vertical motion in the Lesser Antilles is crucial. Some vertical movement assessments in this area were performed in the past. Paleo-geodesy based on coral reef growth tends to show a subsidence trend in Martinique and Les Saintes Islands (south of Guadeloupe archipelago) (Weil-Accardo et al. 2016; Leclerc and Feuillet 2019). This subsidence is corroborated by GNSS observations for a few stations within the vertical velocity ULR6 solution (Santamaría-Gómez et al. 2017). However, an uplift with decreasing rate for la Désirade Island (West of the Guadeloupe Archipelago) was measured (Léticée et al. 2019). Even though GNSS exploitation in this area is challenging, the islands are generally well instrumented, especially the two islands of Guadeloupe and Martinique.
Therefore, this study has a double objective, technical and scientific: on one hand, we estimate a denser vertical velocity field using a maximum of GNSS data in the area. On the other hand, we compare and quantify the differences between the results (coordinates and velocities inferred from the time series) obtained with two GNSS processing software using homogeneous inputs.
The observation set described below was processed using two different GNSS processing software, namely EPOS (Gendt et al. 1998; Uhlemann et al. 2015) and GINS (Marty et al. 2011; Loyer et al. 2012), but using a similar strategy, equivalent models and identical product inputs. The underlying idea is to quantify the intrinsic differences due to different software. We used a Precise Point Positioning approach with float ambiguity resolution. A PPP processing is the most suitable one for this area because the reduced number of IGS reference stations prevents efficient differential processing. Moreover, those reference stations can be affected by the same tectonic processes as the geophysics stations. Thus, IGS stations in the area (namely ABMF and LMMF) are used here as “regular” stations. We considered only GPS observations since most of the stations recorded only this constellation during most of the period considered. The orbits and clock offset products have been generated beforehand by the GFZ Analysis Center in preparation of the IGS Repro3 campaign (Männel et al. 2020). These products are consistent with ITRF2014 (Altamimi et al. 2016). Regarding the models used, we kept a consistency between the two software configurations. The same antenna eccentricities are used for both processings based on station site logs.
Once the two daily coordinate sets are obtained, we select the intersection of both to get equivalent time series with the same daily coordinates. Indeed, some daily data were not properly computed by one or the other software. Stations STG0, SBL0 and PDB0 are completely excluded because of a lack of observations. For each time series, the corresponding velocities are determined using the trend estimation software HECTOR (Bos et al. 2013). We model the time series as combinations of a long term linear trend and an annual+semi-annual periodic signal, along with white and power-law noise. The term trend designates hereafter the linear component. The discontinuities introduced in the trend estimation are based on the material change site logs (an antenna change is systematically considered as a discontinuity) and on a supplementary visual detection (Sakic et al. 2019). The same discontinuities are applied for both equivalent EPOS and GINS solutions.
4 Coordinate and Velocity Differences
Mean, median and standard deviation in millimeters for the three topocentric coordinates and the planimetric distance for the common daily points of the two estimated solutions
5 Vertical Velocity Results
Vertical velocities estimated for the Guadeloupe and Martinique network, for both EPOS and GINS solutions
EPOS solution (mm/yr)
GINS solution (mm/yr)
The mean velocity rate measured for all stations on the archipelago, with the volcano area excluded, is − 1.60 ± 1.54 mm/yr (1σ) using EPOS solution, and − 2.17 ± 1.23 mm/yr (1σ) using GINS solution.
For the Martinique Island, the two solutions are also consistent and general subsidence is observed except for SAM0 station. The mean velocity rate measured is − 1.80 ± 1.36 mm/yr (1σ) using EPOS solution, and − 1.68 ± 1.23 mm/yr (1σ) using GINS solution.
6 Comparison with Existing Solutions
Comparison of vertical velocities for LMMF and ABMF of this study with existing solutions
−2.40 ± 0.46
−0.84 ± 0.40
−2.41 ± 0.41
−0.39 ± 0.42
−3.55 ± 0.48
−2.70 ± 1.33
−2.37 ± 1.15
−2.49 ± 0.54
−1.74 ± 0.88
−2.54 ± 0.21
−0.92 ± 0.37
Using two different solutions but based on the same geodetic products and homogeneous models, we obtain significant disparities in terms of coordinates difference repeatability, especially on the East and Up components with a standard deviation at the centimeter level. Regarding the estimated vertical velocities using the same set of points and the same discontinuities, the differences are also significant. This result tends to motivate investigation on velocity combination strategies between different processing centers, as suggested and tested by Ballu et al. (2019) for instance, where a joint least square modeling is developed to combine equivalent time series from different Analysis Centers. A combination based on a maximum likelihood estimation would be also an relevant method.
Nevertheless, for the studied area of the Guadeloupe and Martinique Islands, a negative velocity trend on the Up component is observed for most of the stations, which might suggest generalized subsidence of the area. This tendency is clear for the Martinique Island, but more complex trends for the Guadeloupe Archipelago can be observed, especially in the area around the Soufrière Volcano. This result can also be nuanced, since some stations have a positive trend, which might be due to local effects. A positive trend can also be due to an important number of discontinuities over the time series period, like the stations PAR1 (furthermore located inside the volcano area) and FFE0 (outside the volcano area). On another hand, a large number of discontinuities for the same station seem to lead also to an overestimated negative trend, like for instance the station LAM0, with a velocity estimated over −3 mm/yr for six discontinuities referenced. This statement reveals the necessity to maintain networks with a minimum of hardware discontinuities, i.e. by reducing the number of antenna changes. MGL0 station, located on the Marie-Galante Island, presents a singular behavior. It is the only station clearly uplifting, with a quasi-complete time-series of 3.5 years, without any visible discontinuity, while the other station on Marie-Galante (MAGA) presents a subsiding trend. Unfortunately, since this station belong to the commercial ORPHEON network, we have only a few metadata that prevent us to explain clearly this behavior.
We corroborate the paleo-geodesy studies carried out in the region. The coral reef records in Martinique (Weil-Accardo et al. 2016) and Les Saintes (Leclerc and Feuillet 2019) indicate also a subsidence but with a smaller order of magnitude of few tenths of a millimeter per year, which can be explained by the difference in the observation time spans (only a few years for GNSS, ca. one century for the coral records). According to those studies, long term subsidence can have multiple origins: volcanic activity, crustal faulting, subduction of the Tiburon ridge for the Saintes Islands (Leclerc and Feuillet 2019), and a potential deep interseismic loading for Martinique (Weil-Accardo et al. 2016).
We used only one software for velocity estimation since we mainly focussed on the GNSS processing itself, but some other velocity estimator software are available (e.g. Blewitt et al. 2016; Santamaría-Gómez 2019). The impact of the velocity estimation software on solutions have been analysed for instance by Mazzotti et al. (2020).
This work brings a comparison of the coordinate time series obtained for the same dataset with two different software but using consistent parameters. New homogeneously calculated vertical velocity fields are made available for geophysical modeling, with unprecedented density for the two Martinique and Guadeloupe Islands. A general subsidence trend is observed for both islands.
We thank the different scientific institutes for the continuously operating GNSS acquisitions, the instrumental maintenance and for providing publicly and freely the observations. Data of the IPGP network, operated on site by the Guadeloupe and Martinique seismic and volcanologic observatories, are available on https://volobsis.ipgp.fr. Data of the IGN and the SONEL networks are available on https://rgp.ign.fr and https://sonel.org respectively. ORPHEON data were provided to the authors for a scientific use in the framework of the agreement between the GEODATA company (https://orpheon-network.fr) and the RESIF-RENAG network maintained by the CNRS-INSU (https://renag.resif.fr). We acknowledge the CNES (Centre national d’études spatiales) for its geodetic processing software GINS. We thank A. Walpersdorf & an anonymous reviewer along with the editors J. Freymueller & L. Sanchez, for their constructive comments which improved the content of this paper.
- Ballu V, Bouin MN, Siméoni P, Crawford WC, Calmant S, Boré JM, Kanas T, Pelletier B (2011) Comparing the role of absolute sea-level rise and vertical tectonic motions in coastal flooding, Torres Islands (Vanuatu). Proc Nat Acad Sci 108(32):13019–13022. https://doi.org/10.1073/pnas.1102842108. http://www.pnas.org/lookup/doi/10.1073/pnas.1102842108
- Ballu V, Gravelle M, Wöppelmann G, de Viron O, Rebischung P, Becker M, Sakic P (2019) Vertical land motion in the Southwest and Central Pacific from available GNSS solutions and implications for relative sea levels. Geophys J Int 218(3):1537–1551. https://doi.org/10.1093/gji/ggz247. https://academic.oup.com/gji/article/218/3/1537/5499028
- Blewitt G, Hammond W, Kreemer C (2018) Harnessing the GPS data explosion for interdisciplinary science. Eos 99. https://doi.org/10.1029/2018EO104623. https://eos.org/project-updates/harnessing-the-gps-data-explosion-for-interdisciplinary-science
- Bos MS, Fernandes RMS, Williams SDP, Bastos L (2013) Fast error analysis of continuous GNSS observations with missing data. J Geod 87(4):351–360. https://doi.org/10.1007/s00190-012-0605-0. http://link.springer.com/10.1007/s00190-012-0605-0
- Gendt G, Dick G, Söhne W (1998) GFZ analysis center of IGS–Annual Report. IGS 1996 Annual Report, pp 169–181Google Scholar
- Heflin M, Moore A, Murphy D, Desai S, Bertiger W, Haines B, Kuang D, Sibthorpe A, Sibois A, Ries P, Hemberger D, Dietrich A (2019) Introduction to JPL’s GPS time series GPS time series. https://sideshow.jpl.nasa.gov/post/tables/GNSS_Time_Series.pdf
- Léticée JL, Cornée JJ, Münch P, Fietzke J, Philippon M, Lebrun JF, De Min L, Randrianasolo A (2019) Decreasing uplift rates and Pleistocene marine terraces settlement in the central lesser Antilles fore-arc (La Désirade Island, 16 deg. N). Quat Int 508(March):43–59. https://doi.org/10.1016/j.quaint.2018.10.030. https://doi.org/10.1016/j.quaint.2018.10.030
- Loyer S, Perosanz F, Mercier F, Capdeville H, Marty JC (2012) Zero-difference GPS ambiguity resolution at CNES–CLS IGS Analysis Center. J Geod 86(11):991–1003. https://doi.org/10.1007/s00190-012-0559-2. http://link.springer.com/10.1007/s00190-012-0559-2
- Männel B, Brandt A, Bradke M, Sakic P, Brack A, Nischan T (2020) Status of IGS reprocessing activities at GFZ. In: International Association of Geodesy Symposia. Springer, Berlin. https://doi.org/10.1007/1345_2020_98
- Marty JC, Loyer S, Perosanz F, Mercier F, Bracher G, Legresy B, Portier L, Capdeville H, Fund F, Lemoine JM, Biancale R (2011) GINS: the CNES/GRGS GNSS scientific software. In: 3rd international colloquium scientific and fundamental aspects of the Galileo programme, ESA proceedings WPP326, vol 31, pp 8–10Google Scholar
- Mazzotti S, Déprez A, Henrion E, Masson C, Masson F, Menut JL, Métois M, Nocquet JM, Rolland L, Sakic P, Socquet A, Santamaria-Gomez A, Valty P, Vergnolle M, Vernant P (2020) Comparative analysis of synthetic GNSS time series - bias and precision of velocity estimations. Research report, RESIF. https://hal.archives-ouvertes.fr/hal-02460380
- Sakic P, Mansur G, Kitpracha C, Ballu V (2019) The GeodeZYX toolbox: a versatile Python 3 toolbox for geodetic-oriented purposes. https://doi.org/10.5880/GFZ.1.1.2019.002. http://dataservices.gfz-potsdam.de/panmetaworks/showshort.php?id=escidoc:4754924
- Weil-Accardo J, Feuillet N, Jacques E, Deschamps P, Beauducel F, Cabioch G, Tapponnier P, Saurel JM, Galetzka J (2016) Two hundred thirty years of relative sea level changes due to climate and megathrust tectonics recorded in coral microatolls of Martinique (French West Indies). J Geophys Res Solid Earth 121(4):2873–2903. https://doi.org/10.1002/2015JB012406. http://doi.wiley.com/10.1002/2015JB012406
- Wöppelmann G, Testut L, Créach R (2011) La montée du niveau des océans par marégraphie et géodésie spatiale: contributions françaises à une problématique mondiale. Annales Hydrographiques 6ème Série 8(777):11–14Google Scholar
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.