Global Isostatic Adjustment Research Articles

PSDEFOPAT – TOWARDS AUTOMATIC MODEL BASED PSI POST-PROCESSING

Abstract. The deformation of the surface of the earth is triggered by numerous naturally occurring and artificial processes such as global isostatic adjustment, aseismic and coseismic movement, varying amounts of groundwater or gas stored underground, and soil consolidation due to urbanization. Monitoring these surface deformations is essential to understand the underlying processes and provide authorities and the public with hazard assessments. Remote sensing techniques, such as Persistent Scatterer Interferometry (PSI), have the capability of mapping these deformations. Besides a spatial analysis of the deformation patterns, PSI also provides information on their temporal evolution. Post-processing strategies to analyze the displacement time series have gained interest in recent years. This paper presents our PSI post-processing strategy, which incorporates different deformation models and automatically chooses the best-fitting one based on statistical tests.

Evaluation of physical and chemical proxies used to interpret past glaciations with a focus on the late Paleozoic Ice Age

The late Paleozoic Ice Age (LPIA) was one of Earth's most important Phanerozoic climatic events lasting for over 100 Mys. Despite its importance, its history is controversial with two hypotheses that portray glaciation differently ( Fig. 1 ). Traditional views characterize the LPIA as a continuous glacial event that lasted from the Middle Mississippian until the Late Permian with a massive ice sheet that covered Gondwana throughout this interval. This approach often uses only one or two proxies to define the glaciation. The other emerging hypothesis suggests that numerous ice sheets occurred in Gondwana with individual glacial events lasting up to 10 Mys alternating with glacial minima/non-glacial intervals of similar duration. Both views are still prevalent. Both near- and far-field proxies are used to define the ice age. Near-field proxies include the occurrence/absence of diamictites, glaciotectonic deposits/landforms, striated clasts and clast pavements, outsized clasts (dropstones), rhythmites, cyclic diamictite-bearing successions, glendonites, grooved and striated surfaces, streamline landforms, and U-shaped paleovalleys. Detrital zircons and chemical index of alteration (CIA) studies help to delineate the occurrence, extent, and location of glaciation. Multiple complexities occur with the use of these proxies as different non-glacial processes and driving factors can produce similar features or results. Far-field proxies focus on identifying changes in eustacy. These include the occurrence of cyclic successions composed of alternating nonmarine and marine strata (cyclothems), depth of incised valleys, paleotopographic relief, phosphatic black shales, and changing oxygen isotope ratios. Like the near-field record, far-field proxies are complex indicators with varied nuances that make their application challenging. Here we discuss the limitations and use of these proxies and promote a multiproxy approach to investigating Earth's glacial intervals. We suggest that studies incorporate multiple proxies coupled with detailed environmental, paleoflow, and paleogeographic analyses to better constrain the occurrence, timing, and extent of glaciation and its influence on global systems. This approach will provide a robust view of the LPIA. We also consider the magnitude and nature of sea-level response to changing ice volumes by discussing ice-volume fluctuations, basin subsidence's modification of glacioeustacy, and sea-level's response to global isostatic adjustment (GIA). In considering these features, it becomes apparent that glacioeustacy is more complex than previously envisioned.

Past and present ITRF solutions from geophysical perspectives

Questions about the accuracy of the origin of the different versions of International Terrestrial Reference Frame (ITRF), have been regularly raised. In particular the origin drift between ITRF2000 and ITRF2005 (and subsequent ITRF solutions) is well-known to be problematic. Here, we look forward a sort of geophysical evaluation of ITRF solutions. We investigate GNSS vertical velocities provided by the last four ITRF solutions (ITRF2000 to ITRF2014; Altamimi et al., 2005, 2007, 2011, 2016) that we compare with different Global Isostatic Adjustment (GIA) model predictions. We find that each new ITRF solution appears to be more and more consistent with all GIA predictions, except ITRF2014 whose consistency with the GIA models depends on the date of observation. Indeed, GNSS observations and GIA predictions appear consistent at global scale at a level of ~4 mm/yr using ITRF2000 data, ~2.5–3 mm/yr using ITRF2005 data, and ~2 mm/yr using ITRF2008 data (global weighted root mean squares). For ITRF2014, the consistency between GNSS observations and GIA predictions is extremely high in 2000 (~1.5 mm/yr) but seems then to decrease with time (~2 mm/yr in 2013). This discrepancy is due to the recent ice melting effect that is not accounted for in GIA models, but clearly evidenced by ITRF2014 vertical velocities during the last years of observations, in particular in Greenland.

Horizontal and vertical velocities derived from the IDS contribution to ITRF2014, and comparisons with geophysical models

In the context of the 2014 realization of the International Terrestrial Reference Frame (ITRF2014), the International DORIS Service (IDS) has delivered to the IERS a set of 1140 weekly SINEX files including station coordinates and Earth orientation parameters, covering the time period from 1993.0 to 2015.0. From this set of weekly SINEX files, the IDS Combination Center estimated a cumulative DORIS position and velocity solution to obtain mean horizontal and vertical motion of 160 stations at 71 DORIS sites. The main objective of this study is to validate the velocities of the DORIS sites by comparison with external models or time series. Horizontal velocities are compared with two recent global plate models (GEODVEL 2010 and NNR-MORVEL56). Prior to the comparisons, DORIS horizontal velocities were corrected for Global Isostatic Adjustment (GIA) from the ICE-6G (VM5a) model. For more than half of the sites, the DORIS horizontal velocities differ from the global plate models by less than 2-3 mm/yr. For five of the sites (Arequipa, Dionysos/Gavdos, Manila, Santiago) with horizontal velocity differences wrt these models larger than 10 mm/yr, comparisons with GNSS estimates show the veracity of the DORIS motions. Vertical motions from the DORIS cumulative solution are compared with the vertical velocities derived from the latest GPS cumulative solution over the time span 1995.0-2014.0 from the University of La Rochelle (ULR6) solution at 31 co-located DORIS-GPS sites. These two sets of vertical velocities show a correlation coefficient of 0.83. Vertical differences are larger than 2 mm/yr at 23 percent of the sites. At Thule the disagreement is explained by fine-tuned DORIS discontinuities in line with the mass variations of outlet glaciers. Furthermore, the time evolution of the vertical time series from the DORIS station in Thule show similar trends to the GRACE equivalent water height.

Late Pleistocene–Holocene evolution of the southern Marmara shelf and sub-basins: middle strand of the North Anatolian fault, southern Marmara Sea, Turkey

Although there are many research studies on the northern and southern branches of the North Anatolian fault, cutting through the deep basins of the Sea of Marmara in the north and creating a series of pull-apart basins on the southern mainland, little data is available about the geometrical and kinematical characteristics of the middle strand of the North Anatolian fault. The first detailed geometry of the middle strand of the North Anatolian fault along the southern Marmara shelf, including the Gemlik and Bandirma Bay, will be given in this study, by a combined interpretation of different seismic data sets. The characteristic features of its segments and their importance on the paleogeographic evolution of the southern shelf sub-basins were defined. The longest one of these faults, the Armutlu-Bandirma segment, is a 75-km long dextral strike-slip fault which connects the W–E trending Gencali segment in the east and NE–SW trending Kapidag-Edincik segment in the west. In this context, the Gemlik Bay opened as a pull-apart basin under the control of the middle strand whilst a new fault segment developed during the late Pleistocene, cutting through the eastern rim of the bay. In this region, a delta front forming the paleoshoreline of the Gemlik paleolake was cut and shifted approximately 60 ± 5 m by the new segment. The same offset on this fault was also measured on a natural scarp of acoustic basement to the west and integrated with this paleoshoreline forming the slightly descending topset–foreset reflections of the delta front. Therefore the new segment is believed to be active at least for the last 30,000 years. The annual lateral slip rate representing this period of time will be 2 mm, which is quite consistent with modern GPS measurements. Towards the west, the Bandirma Bay is a rectangular transpressional basin whilst the Erdek Bay is a passive basin under the control of NW–SE trending faults. When the water level of the paleo-Marmara lake dropped down to −90 m, the water levels of the suspended paleolakes of Bandirma and Gemlik on the southern shelf were −50.3 (−3.3 Global Isostatic Adjustment—GIA) and −60.5 (−3.3 GIA) m below the present mean sea level, respectively. As of today a similar example can be seen between the Sea of Marmara and the shallow freshwater lakes of Manyas and Uluabat. Similarly, the paleolakes of Gemlik and Bandirma were affected by the water level fluctuations at different time periods, even though both lakes were isolated from the Sea of Marmara during the glacial periods.

Sea‐level trends and interannual variability in the Caribbean Sea

Sea‐level trends and their forcing have been investigated in the Caribbean Sea using altimetry and tide gauge time series from 19 stations. The basin average sea‐level rise is 1.7 ± 1.3 mm yr−1 for the period 1993–2010. Significant spatial variability of the trends is found. The steric variability above 800 m combined with the global isostatic adjustment explains the observed trends for the altimetry period in most of the basin. Wind forcing changes causes the trends in the southern part of the basin, modulating the sea level through changes in the ocean circulation. The longest time series (102 years) of Cristobal shows a trend of 1.9 ± 0.1 mm yr−1 insignificantly different from the global mean sea‐level rise for the twentieth century. By contrast Cartagena, a world heritage site, has a large trend (5.3 ± 0.3 mm yr−1) significantly affected by local vertical land movements. Stations dominated by the steric contribution have smaller trends (∼1.3 ± 0.2 mm yr−1). Sea‐level trends at tide gauges are not affected by atmospheric pressure changes or by the open ocean steric contribution at most stations. Decadal variability in the sea‐level trends can partly be explained by steric and wind variability. The decadal variability in the trends is not spatially coherent. Interannual sea‐level variability accounts for one third of the total sea‐level variability and can be partly explained by the influence of El Niño‐Southern Oscillation at different time and spatial scales. No correlation with the North Atlantic Oscillation is found.

Recent sea level trends and accelerations: Comparison of tide gauge and satellite results

A comprehensive set of 456 monthly tide gauge records is analyzed for trend and acceleration over the same period that satellite altimetry was analyzed (1993 to 2011). Additionally, a 90 tide gauge record subset is analyzed for which GPS data are available. The selection criterion for the tide gauge data is 85% data completion. All measurements are adjusted for vertical land motion. Results from 456 pairs of tide gauges, adjusted for Global Isostatic Adjustment, and satellite recordings located within 1° root-mean-square latitude and longitude separation differences are compared. The tide gauge trends and accelerations are adjusted for spatial bias using the more globally dense satellite data.The average trends of the 456 and 90 gauge sets (3.26 and 2.68mm/year, respectively) agree reasonably well with the global trend average of the satellite data (3.09mm/year). Average trends for the 456 tide gauges are also in good agreement (within 95% confidence limits) with trends based on satellite data within the 1° satellite proximity criterion (3.26 and 3.31mm/year, respectively). The trends for the 90 gauges with GPS nearby and qualifying satellite locations are 2.68 and 2.74mm/year, respectively. For all datasets analyzed, the accelerations are quite strongly negative but the uncertainty is relatively large. Adjustment of the tide gauge trends for spatial bias modified both trends and accelerations significantly and decreased trend differences between the 456 and 90 gauge datasets. The spatially adjusted tide gauge trends (2.95 and 2.72mm/year, respectively for the 456 and 90 tide gauges sets) are somewhat less than the 1° spatially adjusted satellite data (3.09mm/year). Whether the increased sea level trend of approximately 3mm/year measured by the satellites since the 1990's is a long-term increase from the 20th Century value of approximately 1.7mm/year or part of a cycle will require longer records; however, the negative accelerations support some cyclic character.

Deceleration in the Earth's oblateness

For over three decades, satellite laser ranging (SLR) has recorded the global nature of the long‐wavelength mass change within the Earth system. Analysis of the most recent time series of 30 day SLR‐based estimates of Earth's dynamical oblateness, characterized by the gravitational degree‐2 zonal spherical harmonic J2, indicates that the long‐term variation of J2 appears to be more quadratic than linear in nature. The superposition of a quadratic and an 18.6 year variation leads to the “unknown decadal variation” reported by Cheng and Tapley (2004). Although the primary trend is expected to be linear due to global isostatic adjustment, there is an evident deceleration () in the rate of the decrease in J2 during the last few decades, likely due to changes in the rate of the global mass redistribution from melting of the glaciers and ice sheets as well as mass changes in the atmosphere and ocean.

Analysing the 100 year sea level record of Leixões, Portugal

► The long-term historic sea level data from Leixões (Portugal) has been recovered. ► Estimated mean sea level trend is not in line with global MSL rise predictions. ► Observed sea levels are influenced by prevailing weather systems and local atmospheric pressure. ► Caution is needed using a single tide gauge as representative of regional sea level. A new data set from the tide gauge at Leixões (North–West Portugal) has recently been transferred from its paper format into a digital time series of hourly sea level values. By measuring sea level variations since 1890, this tide gauge station is one of the few in the world with over 100 yr of digitised hourly records and the longest for the South West of Europe and Iberian Peninsula. This paper presents the procedures adopted to recover the Leixões sea level data from paper chart records as well as the data quality control and data editing methodologies. The mean rate of sea level change between 1906 and 2008 is −0.70 ± 0.27 mm yr −1 , which does not agree with the global mean sea level rise of 1–2 mm yr −1 . No evidence for vertical land movement was found and Global Isostatic Adjustment influence on sea level, at this location, can be neglected. It is likely that prevailing weather systems in the North Atlantic, especially in the winter, and local atmospheric pressure, influence sea levels at Leixões. A further contribution is found from tides and surges. The evolution of the port cannot be ignored when trying to understand sea level change.

Minimum time span of TOPEX/Poseidon, Jason-1 and Jason-2 global altimeter data to detect a significant trend and acceleration in sea level change

The present study aims to estimate a minimum time span of the global mean sea level time series (from TOPEX/Poseidon, Jason-1 and Jason-2 satellite altimetry) which is sufficient to detect a statistically meaningful trend in global sea level variation. In addition, the objective of this paper is also to seek a minimum time span required to detect a significant acceleration in sea level change.Detecting a trend is based upon the Cox–Stuart statistical test. In order to make the analysis representative a stepwise procedure is proposed. The input global mean sea level time series is divided into shorter data sets (blocks) of dissimilar lengths. First, a length of block is chosen and the test is performed for the first truncated time series of a fixed length. Subsequently, the block is moved one step forward in each iteration. This produces the dissimilar data sets of the same length. The procedure is repeated for blocks of greater lengths and for each step the Cox–Stuart test is applied. The procedure makes the inference independent of the starting data epoch. Considering the set S={0.002,0.004,…,0.1} of significance levels, the test is applied to compute the probability of trend detection. Detecting an acceleration is based on the same procedure, however, applied to the differenced global mean sea level time series.We infer that the trend in the global mean sea level time series (January 1993–April 2010) corrected for the global isostatic adjustment can be detected for all significance levels from the set S with the probability close to 1 after 110 satellite cycles (approximately 2.99 years). The corresponding estimate based on the data without correcting for the global isostatic adjustment is equal to 114 satellite cycles (approximately 3.09 years). Over the majority of the study period it is impossible to detect any acceleration in sea level rise. However, for the time spans of approximately 400 and 500 satellite cycles an acceleration may be detected with probability of approximately 0.4. These specific time spans correspond to the periods with reduced rates of sea level change.

Precision and accuracy of GPS-derived station displacements

Space geodetic techniques like Global Navigation Satellite Systems (GNSS), Satellite Laser Ranging (SLR) and Very Long Baseline Interferometry (VLBI) provide valuable input for, e.g., studies of Global Isostatic Adjustment (GIA). This paper discusses the current precision and accuracy of GPS-derived vertical and horizontal station displacements. The precision is evaluated by repeatabilities and solutions computed from different subintervals of the data available. However, due to systematic effects, the precision is often much better than the accuracy. The accuracy is evaluated by comparisons of the space geodetic techniques amongst each other and comparisons with geophysical models for atmospherical and hydrological loading. Besides the analysis of time series, co-located GNSS, SLR, and VLBI sites allow for a comparison of velocities estimated in Terrestrial Reference Frame (TRF) solutions of the different techniques.

Signal and noise in Gravity Recovery and Climate Experiment (GRACE) observed surface mass variations

The Gravity Recovery and Climate Experiment (GRACE) product used for this study consists of 43 monthly potential coefficient sets released by the GRACE science team which are used to generate surface mass thickness grids expressed as equivalent water heights (EQWHs). We optimized both the smoothing radius and the level of approximation by empirical orthogonal functions (EOFs) and found that 6.25° and three modes are able to describe more than 73.5% of the variance. The EQWHs obtained by the EOF method describe all known variations in the continental hydrology, present‐day ice sheet melting, and global isostatic adjustment. To assess the quality of the estimated grids, we constructed degree error spectra of EQWHs. We conclude that a significant part of the errors in GRACE can be explained by a scaling factor of 0.85 relative to degree error estimates provided by the GGM02C gravity model but that the present‐day errors in the GRACE data are a factor 2 to 5 larger than forecasted by tide model differences and atmospheric pressure differences. Comparison to a network of 59 International GNSS Service (IGS) stations confined the filter parameter settings to three EOF modes and 5° or 6.25° smoothing radius. Residuals that remain after the EOF method do exhibit S2 aliasing errors and a semiannual continental hydrology signal contained in the Global Land Data Assimilation Systems (GLDAS) model. Further analysis of the residual EOF signal revealed alternating track correlation patterns that are partially explained by the GRACE covariance matrix and the handling of nuisance parameters in the GRACE data processing.

Are sea-level-rise trends along the coasts of the north Indian Ocean consistent with global estimates?

Mean-sea-level data from coastal tide gauges in the north Indian Ocean were used to show that low-frequency variability is consistent among the stations in the basin. Statistically significant trends obtained from records longer than 40 years yielded sea-level-rise estimates between 1.06–1.75 mm yr − 1 , with a regional average of 1.29 mm yr − 1 , when corrected for global isostatic adjustment (GIA) using model data. These estimates are consistent with the 1–2 mm yr − 1 global sea-level-rise estimates reported by the Intergovernmental Panel on Climate Change.

Impacts of eustasy and hydro-isostasy on the evolution and landforms of Pacific atolls

Consideration of eustatic and hydro-isostatic effects on late Quaternary sea levels in the tropical Pacific Ocean indicates that the configuration of modern atolls with emergent annular reef flats is a transient morphology not developed until post-mid-Holocene time. Annular atoll reefs, perched atop carbonate platforms which cap buried volcanic edifices, are underlain by 8–28 m of Holocene limestone disconformably overlying a substratum of last-interglacial or older limestone. Comparable thicknesses (9–23 m) of Holocene sediment are present beneath atoll lagoons that are uniformly <85 m deep. During glacio-eustatic drawdowns in global sea level by 120–125 m, carbonate platforms of modern atoll provinces rose abruptly from the sea as clusters of subaerial limestone plateaus flanked by steep cliffs. Modern analogues are provided by emergent atolls uplifted on the flexural arches of trench forebulges. Slow subsidence coupled with karstic erosion of emergent atolls during the last glaciation lowered the surfaces of last-interglacial reef edifices by the amounts needed to provide accommodation space for Holocene reef growth during the postglacial eustatic rise in sea level. Modern atoll reef caps began to grow after ∼9 ka when rising Holocene sea level overtopped degraded remnants of interglacial reefs, but remained submerged until carbonate buildups approached sea level in mid-Holocene time (6–4 ka). Classic atoll morphology, with circlets of multiple islets dotting annular reefs, formed in combination with late Holocene hydro-isostatic drawdown in tropical Pacific sea level in response to equatorial ocean siphoning, a facet of global isostatic adjustment (GIA) to deglaciation. Early Holocene eustatic rise in sea level and late Holocene hydro-isostatic decline in sea level combined to produce a regionally variable mid-Holocene highstand in tropical Pacific sea level that stood 1.0–2.6 m above modern sea level. Cemented mid-Holocene paleoreef flats now stranded well above sea level serve as resistant foundations for non-migratory ( pinned) islets that were not present along atoll rims until after the local crossover date, when ambient high-tide level first fell below mid-Holocene low-tide level. Existing atoll landforms have a time depth generally <1–2 ka.

Holocene sea levels in oceanic islands: Implications for the rheological structure of the earth's mantle

Relative sea levels in oceanic islands that are situated sufficiently far from glaciated regions were studied in relation to the rheological structure of the earth's mantle. The local hydro-isostatic adjustment associated with the mantle flow from the oceanic side to an inland side depends on the size of the island, the effective upper-mantle viscosity and the elastic plate thickness. The relative sea level has a significant dependence on the upper mantle rheology for islands with a radius larger than 10 km, but it is almost independent of the upper mantle rheology for islands with a radius less than 10 km. In other words, the observed sea-level changes in such small islands follow the global isostatic adjustment which depends strongly on the lower mantle viscosity. In fact, the relative sea levels in small oceanic islands impose important constraints on the lower-mantle viscosity. On the other hand, the effective upper-mantle viscosity and the elastic plate thickness are determined by analyzing the relative sea levels for islands greater than 10 km. The observed sea-level curves in Japan and New Zealand are consistent with the effective viscosity of the upper mantle, equal to 2 to 6 × 10 21 poises. The effective elastic plate thickness for Japan and New Zealand is estimated to be less than 50 km and greater than 100 km, respectively. The thickness of the lithosphere obtained by sea-level curves in Japan and New Zealand is compatible with the results obtained by seismological studies.