Replace ISORROPIA with HETPv1.0

[lizziel]

Name and Institution (Required)

Name: Lizzie Lundgren
Institution: Harvard University

New GEOS-Chem feature or discussion

We plan to replace ISORROPIA in the model with a newer version based on ISORROPIA called HETerogeneous vectorized or Parallel (HETP). HETPv1.0 was developed by Stefan Miller at Environment and Climate Change Canada and is described in a 2024 GMD publication (see here).

We plan to include the HETP software as a separate git submodule within GEOS-Chem Classic and GCHP, alongside HEMCO, GEOS-Chem, and Cloud-J.

Tagging Aerosols WG Co-chairs @beckyalexander, @theloniuspunk, @wporter

[lizziel]

1-month full chemistry GC-Classic benchmark comparing ISORROPIA II and HETP is now posted at https://ftp.as.harvard.edu/gcgrid/geos-chem/validation/isorropia_vs_hetp/1mon_fullchem/. This uses GEOS-Chem branch dev/14.4.0 as base.

There is a significant slow-down in the HETP implementation which I am looking into. I will see if this is the case in GCHP too.

[yantosca]

There is a significant slow-down in the HETP implementation which I am looking into. I will see if this is the case in GCHP too.

NOTE: this was fixed in commit 2cd624f by @lizziel.

[lizziel]

I spot-checked differences in APM and aerosol-only simulations and see small differences comparable to differences in full chemistry.

[lizziel]

This is now merged into 14.4.0.