Insoluble aerosol affecting condensation though kappa by pdziekan · Pull Request #493 · igfuw/libcloudphxx

pdziekan · 2026-06-18T13:59:20Z

no rd_insol in condensation, instead kappa is reduced if theres insoluble aerosol (Petters & Kreidenweis 2007)

This reverts commit db18240, reversing changes made to 06bc0be.

This reverts commit 9dcde41, reversing changes made to 3334310.

…aerosol

Copilot

Pull request overview

This PR updates the Lagrangian microphysics initialization/diagnostics so that insoluble aerosol no longer directly affects condensation via a separate rd_insol term; instead, insoluble material reduces the effective hygroscopicity (kappa) while contributing to the dry radius.

Changes:

Remove rd3_insol from condensation/Köhler calculations and apply insoluble effects by reducing kappa based on insoluble volume fraction.
Introduce/rewire initialization steps to incorporate insoluble aerosol into dry size (add_insol_to_dry) and update kappa initialization accordingly.
Update ice nucleation to use insoluble particle surface area (rd2_insol) and adjust Python bindings + unit tests.

Reviewed changes

Copilot reviewed 33 out of 34 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
tests/python/unit/diag_incloud_time.py	Simplifies dry distro keys (no explicit insoluble usage in this test).
tests/python/unit/api_lgrngn.py	Updates dry-radius diagnostics and adds assertions for effective kappa with insoluble aerosol.
tests/python/unit/api_common.py	Updates `rw3_cr` / `S_cr` calls to new signatures.
src/particles.tpp	Includes new initialization helper for insoluble aerosol handling.
src/particles_diag.ipp	Drops `rd3_insol` from diagnostic transforms calling Köhler helpers.
src/impl/sources_and_relaxation_of_SDs/particles_impl_src_dry_sizes.ipp	Adjusts dry-size init to include insoluble radius and new `init_kappa(kappa, rd_insol)`.
src/impl/sources_and_relaxation_of_SDs/particles_impl_src_dry_distros_simple.ipp	Inserts insoluble contribution into dry size during source creation and updates kappa init signature.
src/impl/sources_and_relaxation_of_SDs/particles_impl_src_dry_distros_matching.ipp	Adds a guard for unsupported `rd_insol` in matching sources and updates kappa init signature.
src/impl/sources_and_relaxation_of_SDs/particles_impl_rlx_dry_distros.ipp	Modernizes iterator loops; clarifies assumptions about produced aerosol and `rd_insol`.
src/impl/particles_impl.ipp	Renames/removes insoluble state from `rd3_insol` to `rd2_insol` (ice-only) and updates init method signatures.
src/impl/initialization/particles_impl_reserve_hskpng_npart.ipp	Reserves `rd2_insol` only when ice is enabled; removes `rd3_insol` reserve.
src/impl/initialization/particles_impl_init_wet.ipp	Removes `rd3_insol` from wet-radius equilibrium calculation path.
src/impl/initialization/particles_impl_init_T_freeze.ipp	Updates freezing init to use `rd2_insol` (surface area proxy).
src/impl/initialization/particles_impl_init_SD_with_sizes.ipp	Uses `rd_sol + rd_insol` in dry size init; updates kappa + insol init sequencing.
src/impl/initialization/particles_impl_init_SD_with_distros.ipp	Adds insoluble contribution to dry size and updates kappa initialization sequencing.
src/impl/initialization/particles_impl_init_sanity_check.ipp	Disallows insoluble aerosol initialization when chemistry is enabled.
src/impl/initialization/particles_impl_init_kappa.ipp	Computes effective kappa after insoluble addition (volume-fraction reduction).
src/impl/initialization/particles_impl_init_insol_dry_sizes.ipp	Stores insoluble size as `rd2_insol` (radius squared).
src/impl/initialization/particles_impl_init_dry_const_multi.ipp	Whitespace-only tweak in signature line.
src/impl/initialization/particles_impl_add_insol_to_dry.ipp	New helper for adding insoluble aerosol contribution into dry size.
src/impl/ice/particles_impl_ice_nucl_melt.ipp	Switches ice nucleation inputs from `rd3_insol` to `rd2_insol`.
src/impl/ice/particles_impl_ice_dep.ipp	Removes unused tuple inputs (rd3/kpa) from deposition helper zip iterator.
src/impl/housekeeping/particles_impl_hskpng_rc2.ipp	Removes `rd3_insol` from rc2 approximation inputs.
src/impl/diagnose_SD_attributes/particles_impl_update_incloud_time.ipp	Removes `rd3_insol` from incloud-time diagnostic transform inputs.
src/impl/diagnose_SD_attributes/particles_impl_fill_outbuf.ipp	Renames exposed attribute from `rd3_insol` to `rd2_insol` and gates it behind `ice_switch`.
src/impl/condensation/perparticle/perparticle_nomixing_adaptive_sstp_cond.ipp	Removes `rd3_insol` from condensation substepping tuple payloads.
src/impl/condensation/perparticle/perparticle_advance_rw2.ipp	Removes `rd3_insol` from advance tuple payloads.
src/impl/condensation/percell/particles_impl_cond.ipp	Removes `rd3_insol` from per-cell condensation tuple payloads.
src/impl/condensation/common/particles_impl_cond_common.ipp	Removes `rd3_insol` from activity-of-water usage and tuple parsing/diagnostics.
include/libcloudph++/lgrngn/opts_init.hpp	Switches `rlx_dry_distros` to a shared alias type.
include/libcloudph++/lgrngn/distro_t.hpp	Adds `rlx_dry_distros_t` alias definition.
include/libcloudph++/common/kappa_koehler.hpp	Removes insoluble term from Köhler helpers (activity/equilibrium/critical values).
include/libcloudph++/common/ice_nucleation.hpp	Updates ice nucleation API to use insoluble surface area (`rd2_insol`).
bindings/python/common.hpp	Updates Python binding signatures for `rw3_cr` and `S_cr`.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

      thrust::transform(
-        thrust::make_zip_iterator(thrust::make_tuple(rd3_insol.begin() + n_part_old, u01.begin())),
-        thrust::make_zip_iterator(thrust::make_tuple(rd3_insol.begin() + n_part_old, u01.begin())) + n_part_to_init,
+        thrust::make_zip_iterator(thrust::make_tuple(rd2_insol.begin() + n_part_old, u01.begin() + n_part_old)),
+        thrust::make_zip_iterator(thrust::make_tuple(rd2_insol.end(),   u01.end())),
        T_freeze.begin() + n_part_old,
        T_freeze_CDF_inv_functor<real_t>(opts_init.inp_type)
      );


+
+      // add the insoluble aerosol
+      const real_t rd3_insol = rd_insol * rd_insol * rd_insol;
+      thrust::transform(
+        rd3.begin()+n_part_old,
+        rd3.end(),
+        thrust::make_constant_iterator<real_t>(rd3_insol),
+        rd3.begin()+n_part_old,
+        // arg::_1 + arg::_2
+        thrust::plus<real_t>()
+      );


+
+      debug::print(kpa);
+


      auto p_sdd = sdd.cbegin();
+      assert(p_sdd->first.rd_insol == 0); // rd insol in matching source not implemented yet 



-assert (n == 20 ).all()
-assert isclose(k, n * kappa2, rtol=1e-20)
+for rd_sol, kappa, n_stp in [(1e-6, kappa1, 30), (15.e-6, kappa1, 10), (1.2e-6, kappa2, 20), (12.e-6, kappa2, 15)]:
+    print("testig rd_sol: ", rd_sol, " kappa: ", kappa, " n_stp: ", n_stp)


AgnieszkaMakulska · 2026-06-18T15:44:12Z

+        rd3.begin() + n_part_old, rd3.end(),
+        kpa.begin() + n_part_old,
+        (arg::_1 - rd3_insol) / arg::_1 * kappa
+        // assuming that soluble and insoluble densities are the same (hence using volume fraction mixing, not mass fraction as should (?) be done)


I think it's ok following Petters and Kreidenweis:
"ε is defined as the volume fraction of a model salt in a dry particle consisting of the model salt (κ_m) and an insoluble
core. Two-component system of a model salt and an insoluble species (κ=0) gives κ=ε × κ_m ."

AgnieszkaMakulska · 2026-06-18T16:39:39Z

+        thrust::make_zip_iterator(thrust::make_tuple(rd2_insol.end(),   u01.end())),
        T_freeze.begin() + n_part_old,
        T_freeze_CDF_inv_functor<real_t>(opts_init.inp_type)
      );


that problem with u01 for T_freeze was fixed in my PR, so we need to keep changes from my PR here

…wet and init_dry

…into rd_insol_in_kappa

Copilot

Pull request overview

Copilot reviewed 57 out of 58 changed files in this pull request and generated 7 comments.

      {
        assert(kappa > 0);
-        if ((rw3 - rd3 - rd3_insol) / si::cubic_meters  <= std::nextafter(real_t(0),real_t(1)))
-          return real_t(0.);
-        return (rw3 - rd3 - rd3_insol) / (rw3 - rd3 * (real_t(1) - kappa) - rd3_insol);
+        return (rw3 - rd3) / (rw3 - rd3 * (real_t(1) - kappa));
      }


+        BOOST_GPU_ENABLED
+        real_t operator()(real_t rd3_val) const
+        {
+          return pow(insol_fraction * rd3_val, real_t(2) / real_t(3));
+        }


        const real_t &kappa(dsi->first.kappa);
-        const real_t &rd_insol(dsi->first.rd_insol);
+        const real_t &soluble_fraction(dsi->first.soluble_fraction);
        const auto &size_number_map(dsi->second);


+      // init other properties of SDs
      init_kappa(
-        p_sdd->first.kappa
+        p_sdd->first.kappa,
+        p_sdd->first.soluble_fraction
      );


    // if any SDs with dry radius similar to the one to be added are present,
    // we increase their multiplicity instead of adding new SDs
    // TODO: make it work for sdd.size()>1
+    // -------- TODO: match not only sizes of old particles, but also kappas, rd_insol, chemical composition... --------


      // analyze distribution to get rd_min and max needed for bin sizes
      // TODO: this could be done once at the beginning of the simulation
+      // TODO: take rd_insol into account here?
      init_dist_analysis_sd_conc(


-assert (n == 20 ).all()
-assert isclose(k, n * kappa2, rtol=1e-20)
+for rd, kappa, n_stp in [(1e-6, kappa1, 30), (15.e-6, kappa1, 10), (1.2e-6, kappa2, 20), (12.e-6, kappa2, 15)]:
+    print("testig rd: ", rd, " kappa: ", kappa, " n_stp: ", n_stp)


pdziekan added 8 commits June 17, 2026 14:22

Revert "Merge pull request igfuw#491 from AgnieszkaMakulska/ice"

e206202

This reverts commit db18240, reversing changes made to 06bc0be.

Revert "Merge pull request igfuw#488 from AgnieszkaMakulska/ice"

7ca9dc8

This reverts commit 9dcde41, reversing changes made to 3334310.

add rd insol to rd in init; init kappa taking into account insoluble …

cf8b6f5

…aerosol

rd_insol in sources

9de1ab2

rd_insol in relaxation

ccf2aeb

dont pass rd3 and kappa to deposition

a2742ba

sanity check: chemistry doesnt work with rd_insol

c2e8d97

api_lgrngn test: account for rd and kappa changes

f462191

Copilot AI review requested due to automatic review settings June 18, 2026 13:59

Copilot started reviewing on behalf of pdziekan June 18, 2026 13:59 View session

Copilot AI reviewed Jun 18, 2026

View reviewed changes

AgnieszkaMakulska reviewed Jun 18, 2026

View reviewed changes

pdziekan added 7 commits June 19, 2026 11:14

replace rd_insol with soluble_fraction

5bc373d

rename init_insol_dry_sizes to init_insol to be consistent with init_…

e49bb0c

…wet and init_dry

bring back correct u01 itgeration in T_freeze init

3c0537b

Merge branch 'master' into rd_insol_in_kappa

27d9bed

init_insol: use functor instead of lambda

ea88691

Merge branch 'rd_insol_in_kappa' of github.com:pdziekan/libcloudphxx …

375f024

…into rd_insol_in_kappa

cond test: 100% soluble

d07cb04

AgnieszkaMakulska added the condensation label Jun 24, 2026

pdziekan requested a review from Copilot June 24, 2026 13:38

Copilot started reviewing on behalf of pdziekan June 24, 2026 13:38 View session

Copilot AI reviewed Jun 24, 2026

View reviewed changes

This was referenced Jun 24, 2026

Fix a_w() to handle kappa==0 safely #495

Draft

Fail fast on invalid soluble_fraction in src_dry_sizes #496

Draft

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Insoluble aerosol affecting condensation though kappa#493

Insoluble aerosol affecting condensation though kappa#493
pdziekan wants to merge 15 commits into
igfuw:masterfrom
pdziekan:rd_insol_in_kappa

pdziekan commented Jun 18, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

AgnieszkaMakulska Jun 18, 2026

Uh oh!

AgnieszkaMakulska Jun 18, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

		auto p_sdd = sdd.cbegin();
		assert(p_sdd->first.rd_insol == 0); // rd insol in matching source not implemented yet

Uh oh!

Conversation

pdziekan commented Jun 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

AgnieszkaMakulska Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

AgnieszkaMakulska Jun 18, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

pdziekan commented Jun 18, 2026 •

edited

Loading