Compatibility of NSSL 2-Moment Microphysics (v17) with WRFDA 3DVAR & Hail Detection Sensitivity Issue (NSSL v18)

[naufalkf85]

Dear WRF Community,

I am currently conducting a research study on a hailstorm case using the WRF model. My goal is to simulate hail events as realistically as possible using data assimilation with radar observations via WRFDA 3DVAR.

Based on previous studies in the same region, I used the NSSL 2-moment microphysics scheme (option 17), which is reported to be more sensitive in producing hail variables such as MAXHAILDIAM. However, I understand that option 17 is deprecated in favor of option 18, which combines NSSL microphysics with a more updated infrastructure.

Here is my problem:

• When using NSSL v17 with WRFDA 3DVAR, the da_wrfvar.exe runs successfully.
• However, the resulting wrfout file contains only zero values for all variables, indicating a failure in proper model initialization after DA.
• In contrast, using the same DA setup with other schemes (e.g., NSSL v18, MYJ + Thompson) works and produces physically reasonable outputs.
• "I also tested with NSSL microphysics option 18, which is supported and does not crash, but I found it much less sensitive in generating hail diameter (MAXHAILDIAM) compared to option 17, even when storm and thermodynamic conditions were clearly favorable. Is this expected behavior in NSSL v18? Are there tunable parameters (e.g., thresholds for wet growth or graupel conversion) that could increase hail sensitivity?"

My questions:

1. Is there any known issue with using NSSL 2-moment v17 (microphysics option 17) with WRFDA 3DVAR that prevents successful data assimilation or output initialization?
2. Why might NSSL v18 show significantly reduced sensitivity to hail formation, especially in terms of MAXHAILDIAM, compared to v17? Are there threshold or tuning parameters that differ?
3. Are there recommended workarounds or configurations for users who still require the hail sensitivity of v17 but also wish to use DA with WRFDA?

Any insights, technical clarifications, or references would be very helpful to justify my modeling setup and ensure compatibility between physics schemes and the assimilation framework.

Thank you in advance for your support!

Best regards,
Naufal Kamil Fauzi