Comparison of observed and modeled seasonal crustal vertical displacements derived from multi‐institution GPS and GRACE solutions

Abstract

Eleven GPS crustal vertical displacement (CVD) solutions for 110 IGS08/IGS14 core stations provided by the International Global Navigation Satellite Systems Service Analysis Centers are compared with seven Gravity Recovery and Climate Experiment (GRACE)‐modeled CVD solutions. The results of the internal comparison of the GPS solutions from multiple institutions imply large uncertainty in the GPS postprocessing. There is also evidence that GRACE solutions from both different institutions and different processing approaches (mascon and traditional spherical harmonic coefficients) show similar results, suggesting that GRACE can provide CVD results of good internal consistency. When the uncertainty of the GPS data is accounted for, the GRACE data can explain as much as 50% of the actual signals and more than 80% of the GPS annual signals. Our study strongly indicates that GRACE data have great potential to correct the nontidal loading in GPS time series.

Validation of the EGSIEM GRACE Gravity Fields Using GNSS Coordinate Timeseries and In-Situ Ocean Bottom Pressure Records

“…This phenomenon is also supported by the below-mentioned Tables 2 and 3. These tables indicate that the differences among different GNSS products are larger than the differences among different gravity field solutions, which has also been suggested by Gu et al (2017) [9] who compared multi-institutional GNSS and GRACE products. In other words, the GRACE products are more consistent with one another than the GNSS products.…”

“…GFZ RL05a provides marginally worse cumulative WRMS reductions which might be due to its inferior performance at degree-3, which is also observed by using the other two GNSS time series from EGSIEM-reprocessed (see Figure S1) and JPL (see Figure S5). Nevertheless, the mean cumulative WRMS reductions shown here from all six gravity solutions are better than that from the latest results (maximum mean WRMS reduction up to 15%, see Table S3) published by Gu et al (2017) [9]. One more feature to be gleaned from Figure 1 is that no significant contributions come from degree beyond 30, which has been already demonstrated by Han (2016) [45].…”

“…With the improvement of both GNSS time series and GRACE gravity products, better agreement has been demonstrated by many other studies at both global [6][7][8][9][10] and regional scale [11,12]. For example, the very recent study by Gu et al (2017) [9] has compared vertical displacements of 105 globally distributed stations from GPS and GRACE using multi-institution solutions and demonstrated a mean WRMS reduction up to 15% using the ITSG2016 GRACE products and the GNSS products from the IGS (International GNSS Service) analysis center JPL. In addition, Chanard et al (2018) [10] compared both vertical and horizontal deformation from ITRF 2014 GNSS time series and GRGS RL03 GRACE products and obtained an adjusted mean WRMS reduction up to 14.7% in the vertical component.…”

“…In addition, Table 3 demonstrates again the conclusion that different GNSS products show larger differences than different GRACE products when comparing between GNSS and GRACE products. That is too say, the inconsistencies between GNSS and GRACE are most likely due to errors in the GNSS data like site-specific effects [6,8,9].…”

“…Comparisons of GNSS and GRACE has been undertaken since the first GRACE data releases [5]. With the improvement of both GNSS time series and GRACE gravity products, better agreement has been demonstrated by many other studies at both global [6][7][8][9][10] and regional scale [11,12]. For example, the very recent study by Gu et al (2017) [9] has compared vertical displacements of 105 globally distributed stations from GPS and GRACE using multi-institution solutions and demonstrated a mean WRMS reduction up to 15% using the ITSG2016 GRACE products and the GNSS products from the IGS (International GNSS Service) analysis center JPL.…”