Minimize Envisat News

Improvement of Envisat RA-2 reprocessed data v2.1

28 August 2012

An anomaly has been identified in the altimeter instrumental processing of the ENVISAT RA2 GDR v2.1 reprocessed products (see the validation report and in particular section '6.1. Impact of the reprocessing and PTR drift').

The altimeter internal delay measured by means of the altimeter internal calibration (PTR: In Flight Time Delay Calibration Factor) was not determined correctly and this had a subsequent, and significant, impact on the MSL (mean sea level) trend estimates.

Figure 1



A work around has been defined to correct for this anomaly - a specific chain was developed to compute the correct PTR correction for the whole Envisat mission.

Further analysis has allowed us to include several evolutions in the processing which correct for other anomalies that currently impact the operational processing. The main impact is obviously on the MSL trend but some regional impacts are also noticed. This is illustrated on the below figure.

Figure 2



The auxiliary files, which allow updating the altimeter range provided in the ENVISAT v2.1 data set, are available on the ESA diss-nas-fp.eo.esa.int server (in the directory /altimetry_dataset_v2.1/PTR_external_correction_file ). PTR files are delivered in NetCDF format and an example of the product content is provided here.

To apply the correction, users should simply use the delta__ptr_time_delay_Ku correction value whose datation corresponds to that of the GDR. This correction should be removed from all altimeter ranges (Ku-band ocean range,18 Hz Ku-band ocean ranges, …). For example, the sea surface height should be computed as follow:

  • SSH = orbit - (range - Delta_PTR) – all other corrections = orbit - range – all other corrections + Delta_PTR

Users are also advised that:

  • The current PTR cycle file coverage does not correspond strictly to that of the GDR . The PTR file starts with pass 002, pass 001 is included in the PTR file of the previous cycle

  • There is a 3 minutes overlap between cycles. For the corresponding records, users are advised to use the data from the PTR file of the lowest cycle number

  • There are a few data gaps related to the data set used to generate this auxiliary file that does not correspond exactly to the one used for the GDR v2.1 processing

  • Also Side-B data are provided in this release of the PTR correction file
    Note: Using these files, B-Side MSL data are affected by a bias (around 3.5.cm) probably linked to instrumental differences between B and A side and still under investigations

  • The Delta PTR correction for the S band has not been validated yet