GUNW product suite

[dbekaert]

Background

The GUNW product is the underlying standard product for all ARIA InSAR products. The product is defined to ensure spatial temporal connected network (i.e. Connected for time-series, and spatially connected for stitching over large area's). The underlying formatting of the product is fixed and stored as a CF compliant netcdf file. The product is SAR sensor neutral and has identical layers between different sensors with all meta-data embedded within the product. The product is optimized for storage with 3D cubes and reduced lateral spacing when not needed.

Filename convension and general layers (version 2 product) are documented on product specification page of the ARIA-website:
Sensor-DatasetName-A/D-L/R-tracknumber-mode-YYYYMMDD_YYYYMMDD-HHMMSS-XXYYYN/S_XXYYYN/S-PP/RP/PR/RR-systemTag-versionTag.nc

There are two main product classes that build on the GUNW product (see ARIA website on products flavors):

1. Default standard product (aka ARIA's science product) - DatasetName = GUNW
2. The event standard product (aka ARIA's Cosesimic and volcano product) - DatasetName = GUNW_EVENT; This product was envisioned to cover the complete event area and was pre-stitched in the InSAR processing. TBD what the resolution of this product would be. TBD if this could be archived at the DAAC if we distinct it from GUNW datasetName.

Upgraded for the GUNW product collection.

The proposal is to make a new collection release of the GUNW product. Main changes are as follows:

• The S1-GUNW v2 product has multiple placeholders which will be filled in version 3:
  • Ionosphere as estimated from split spectrum approach and stored at a coarse resolution (e.g. 1 km spacing)
  • Tropospheric (wet and hydrostatic) as estimated using RAiDER package using HRES model and stored as a 3D cube at coarse spacing (e.g. 10km lateral and few hundreds of meter in vertical)
  • Solid Earth Tides computed through Py-solid package and stored as a 3D cube at a coarse spacing (same as existing geometry layers)
• The S1-GUNW product will be reworked to have fixed frames (i.e. no more frame shifting based on S1)
• The S1-GUNW products will have 1-2 full burst overlap (increasing overlap from 2 km to 20-40km for S1)
• Additionally there will be newlayers:
  • Coherence (unfiltered) and at the same posting as unwrapped layer
  • Dense offsets (TBD as this in most cases will not be a usefull layer - too noisy and not sensitive to displacement)
• Would be good to work with CEOS ARD team and do a self-assessment for the InSAR ARD product against the GUNW template.

[dbekaert]

team-discussion:

• We should consider also updating the product filename and then it would simplify enumeration
• HRES versus HRRR, have not done the trade-off.