also find out what's new for winter sea ice, summer sea ice, sea ice (NRT), polar oceans, polar oceans (NRT), coastal oceans, inland waters,
See the product downloads page for access details.
The following evolutions have been applied in Baseline-D (released May 2025) as compared to Baseline-C (released April 2024). Further details are provided in the Cryo-TEMPO Product Handbook and ATBD documents.
LRM slope correction evolved such that if the POCA location is not within the range window then the measurement is rejected.
Increased the DEM resolution used to relocate LRM measurements over Antarctica from 200m to 100m. Antarctic DEM used REMA-GAPLESS-100 (Dong et al, 2022)
Uncertainty parameter for Antarctica now calculated using additional co-variate of roughness.
Slope and roughness data set used for Antarctica to calculate uncertainty uses a new SVD method (Phillips. J, 2024, Lancs)
Reference DEM parameter now calculated from REMA v2 @ 1km (Antarctica), Arctic DEM v4.1 @ 1km (Greenland)
Dilated ice area mask updated to use new dilation method.
Basin ID parameters updated to include Mouginot (2019) for Greenland.
The netcdf basin id parameters are now:
Basin_id: Zwally (AIS) and Mouginot (GIS)
Basin_id2: Rignot (AIS), Rignot (GIS)
Use of the latest evolution of ESA L1b data set as input (for all products). Only ESA CS2 Baseline-E L1b data is used as input to the CryoTEMPO processing (in CryoTEMPO Baseline-B it was necessary to use a sub-optimal mix of available data from L1b Baseline-D and Baseline-E).
New improved slope correction method used for LRM mode measurements using an adapted Roemer algorithm: ref: https://doi.org/10.1016/j.rse.2006.02.026. This method uses high resolution (100m) Digital Elevation Models of Antarctica (REMA v1.1) and Greenland (ArcticDEM v4.1) to find the point of closest approach to the satellite. Previously this method was considered too slow (compute intensive) for operational production. In Baseline-B a faster (but lower resolution and less accurate) slope model approach was used (as per the method used with ESA Level-2 baseline-E and all previous baseline products).
OCOG retracker threshold reduced to 0.1 (from 0.2 in Baseline-B) to reduce differences to ICESat-2 elevations.
New SARin backscatter bias correction to align SARin and LRM backscatter along track. The switch to using new ESA L1b Baseline-E data as input to the CryoTEMPO processing should correct a problem with backscatter drift that was present in their Baseline-D L1b and consequently in CT Baseline-B products.
The following new global netCDF attributes are provided in Baseline-B land ice products: (example values shown)
The instrument_mode attribute is provided so users can quickly identify files containing measurements from each instrument mode without reading any data variables. Note that each file contains measurements from a single instrument mode.
The ascending_start_record and descending_start_record are provided so that users can identify the direction of the satellite's nadir orbit for any record in the file. This is useful since the latitude and longitude coordinates provided in the file are for the location of the ground measurement (also known as the point of closest approach (POCA)) which is terrain dependent and may zig-zag between ground features.
The src_esa_l1b_file attribute is provided so that users can trace the ESA L1b processing applied to the raw satellite measurements before the Cryo-TEMPO L2 processing has been performed.
A new Maximum Coherence Retracker (as per Aublanc et al, 2021, McMillan 2019) replaces the TCOG retracker for SARin waveforms only. Validation studies have shown that this improves the accuracy of elevation measurements over the ice sheet margins (SARin mask).
Smoothing of waveforms changed from sliding average (width=3) to Savitsky-Golay (width 9).
No algorithm changes, but a change in the SARin retracker causes an ~2dB mean change in the backscatter calculated for SARin measurements as compared to Baseline-A. This is expected as the returned power relates to a new retracking point which will be on average higher up the waveform leading edge (than chosen by the Baseline-A TCOG retracker with a 50% threshold).
The phase difference waveform is now sampled at the location of maximum coherence chosen by the retracker (instead of using a model fit in Baseline-A). No smoothing or interpolation is performed. The phase difference value is then converted into an angle from the boresight vector (given in the L1 data).
In Baseline-A an ambiguity check was performed to detect potential phase wrapping by comparing the measured SARin height with an external DEM. Differences greater than 100m were flagged as ambiguous but no attempt was made to correct the measurements location and elevation. In Baseline-B to determine if the original phase measurement was wrapped, the phase difference is then unwrapped and a second elevation and location is calculated. The location with the smallest difference from the DEM is taken as the POCA and the height measurement is given at that location. This results in the retrieval of a higher number of SARin measurements over the ice sheet margins in Baseline-B.
An additional glaciological drainage basin parameter has been added to the product (netcdf ‘basin2’ parameter). This provides the basin number according to Rignot (2016). Previously only the basins defined by Zwally (2012) were provided in the ‘basin’ parameter. Both these drainage basin definitions are widely used and are specified by the IMBIE project for ice sheet mass balance calculations.
The uncertainty parameter is now calculated from the statistics of differencing closely located elevation measurements from CryoTEMPO and IceSAT-2 ATL-06 v005 per band of surface slope. In Baseline-A the uncertainty was calculated from the statistics of CryoTEMPO crossover differences per band of surface slope. In both cases the statistics are generated from combining differences less than 1 month over a single reference year (2020).