The COST-G Plotter

Stéphane Bourgogne 14 July 2019

On this page you will be able to plot GRACE gravity time-series from the different groups. Choose your data center, your data version, choose the area on which on want to extract data and plot your graph! Read full information at the bottom of the page.

Latest update: May 25, 2023

Information

News
You can now save and share your work with a single click! Click on the "Share" button to generate a link to your work. Thanks to this feature, you will be able to access your work later again, or send it to your colleagues and instantly share your research.

Spherical harmonics
Time series are obtained from the gravity field models downloaded at ICGEM.
Except for CNES, models are filtered with the DDK5 filter (see Kusche et al, 2009).

Geoid heights or equivalent water heights
The unit for geoid heights is the cm of geoid anomalies compared to an average. The unit for EWH is the cm of fresh water anomalies, considering that all gravity anomalies observed by GRACE come from a thin layer of fresh water placed on the surface of the Earth (considered spherical). While this unit is pertinent for all phenomena involving water transports at the surface of the Earth, it is meaningless in the case of processes involving the solid Earth, such as earthquakes in particular. Thus, in the case of solid Earth phenomena, care should be taken to express the GRACE results in units other than EWH (gravity anomalies, for instance). The conversion of spherical harmonics into EWH is performed using the k2’ load Love number coefficients (dependent on the spherical harmonic degree), in order to take into account the elastic response of the Earth to a surface load. As a consequence, the EWH value provided on this plotter is the true thickness of the water anomaly resting on the Earth’s crust.

C20 substitution
Spherical harmonic models are converted into 1°x1° grids after replacement of the C20 by the SLR series provided in GRACE Technical Note 11, and subtraction of the mean over the time span. Coefficients taken into account for the conversion to grid go from degree 2 to degree 90. The time-series are extracted from these grids through a geographical extraction algorithm (see item below).

CNES notice
CNES solutions are not a-posteriori filtered since they are regularized in the inversion process - either by a constraint or by truncated SVD. Coefficients go from degree 2 to degree 80 (RL03) or degree 50 (RL02). The C20 is not replaced in CNES solutions since they already contain SLR data (Lageos 1&2, Starlette, Stella).

Geographical extraction
For single point extraction, a barycentric computation is performed from the values at the 4 surrounding grid points. For rectangles and polygons, an average is performed over grid points situated inside the contour. For basins, an average is performed over grid points from a basin file. In every case, grid values are weighted by cos(latitude), in order to obtain per surface unit values (cm/m²). Complete area information, including list of grid points and weights, is provided in the numerical data box.

Background maps
GRACE gravity maps (trend, annual and semiannual amplitude of equivalent water heights) are obtained through a regression on the global series of harmonic coefficients. DDK5 filtered coefficients are used (except for CNES).

Regressions
Analytic models are adjusted on time series through a least squares process. "Periodic model" includes a linear component, an annual component and a semiannual component. The "Advanced model" has a polynomial component instead of linear, and polynomial-modulated annual and semiannual components. In both case, components can be displayed using "Periodic+" and "Advanced+".

Tips
Click on the legend of the series to activate/deactivate the series on the graph. Highlight an area on the graph to zoom in. Use bias and scale if you wish to visualize different quantities on the same graph (for example equivalent water heights and geoid heights). When requesting regressions, use Periodic+ and Advanced+ to show the components of the adjusted model.