.. _faqs_compare_gnss: How can I compare CGM GNSS data with this product? ================================================== Comparing GNSS vs InSAR in Line-Of-Sight ---------------------------------------- #. To directly compare these measurements in their current state, you will need to "pin" your GNSS velocities or time series data to the same reference station point to which the InSAR data is referenced (to find this station point for your track of interest, see the "Track_*_reference_frame" inside the HDF5 file). For velocities, this requires you to subtract the reference point velocities from your chosen station's velocities. #. Next, we need to project the GNSS data into line-of-sight (LOS). To do this, we need to extract the satellite look direction at the station location from the track's HDF5 file. See the steps in Section 2 to best accomplish this with the resources on your machine. #. Once we have the look direction, :math:`l`, we need to use the dot product to take the 3D GNSS velocities/displacements to 1D LOS velocities/displacements, for example for velocities: .. math:: :label: eqn:vLOS v_\text{LOS} = v_\text{GNSS} \cdot \hat{l} = v_{\text{GNSS}_\text{e}} \cdot \hat{l}_\text{e} + v_{\text{GNSS}_\text{n}} \cdot \hat{l}_\text{n} + v_{\text{GNSS}_\text{u}} \cdot \hat{l}_\text{u} You can similarly project your GNSS uncertainties, using the full covariance matrix, :math:`\Sigma`, available from the ".pos" files, into the LOS direction: .. math:: :label: eqn:varLOS \sigma_\text{LOS}^2 = \hat{l}^\text{T} \Sigma_\text{GNSS} \hat{l} .. math:: :label: eqn:sigLOS \sigma_\text{LOS} = \sqrt{\sigma_\text{e}^2\hat{l}_\text{e}^2 + \sigma_\text{n}^2\hat{l}_\text{n}^2 + \sigma_\text{u}^2\hat{l}_\text{u}^2 + 2\rho_\text{en}\sigma_\text{e}\sigma_\text{n}\hat{l}_\text{e}\hat{l}_\text{n} + 2\rho_\text{eu}\sigma_\text{e}\sigma_\text{u}\hat{l}_\text{e}\hat{l}_\text{u} + 2\rho_\text{nu}\sigma_\text{n}\sigma_\text{u}\hat{l}_\text{n}\hat{l}_\text{u}} #. At this point, you can plot the velocities or displacement time series together for a true comparison. Comparing GNSS vs InSAR in 3D Components ---------------------------------------- This comparison, while possible, requires a 3D decomposition of InSAR time series velocities or displacements, which is currently an active area of CGM research.