contribution for a new feature

[ankithva]

Currently, there is no feature to find long-term median bias from varios GR Sweeps (i think so) . I like to contribute for the same (For Indian GR data)

[ghiggi]

Hi @ankithva, I agree with your assessment.

At the moment, the software only provides routines to perform volume matching for individual GR sweeps. Computing a long-term median bias require a custom script/workflow: reading multiple GR sweeps, identifying coincident GPM overpasses, performing the volume matching, and then concatenating the resulting data frames into a single table from which the long-term statistics can be computed.

The main challenge I see in integrating this directly into the software is defining a solution that is sufficiently general to work across different radar networks, I/O formats, and preprocessing choices (e.g., clutter suppression, GR attenuation correction, filtering rules, etc.).

One possible approach could be for you to share the scripts or routines you’ve developed in a separate repository. Alternatively, we could consider creating a dedicated GPM GV repository where you, myself, and other contributors could share and maintain their analysis scripts and workflows.

One of my students worked hard during the past months on developing a quite general workflow to better leverage DPR data for (inter)calibrating X-band and C-band radars (that are significantly more affected by attenuation than S-band radars, for which the volume-matching approach and gates filtering rules were originally designed). He also developed an interactive interface helping to explore the radar offsets variability over time and explore the impact of filtering rules on the obtained bias.

I’d be very interested to hear your thoughts and ideas on what would work best.