Add minimal modal run to model dry deposition by zdaq12 · Pull Request #211 · compdyn/partmc

zdaq12 · 2025-07-14T10:53:31Z

No description provided.

…e modal run; handle input and output for processing

… default parameters

…th arrays

slayoo · 2025-09-16T09:12:21Z

closing and reopening to trigger codecov analysis

codecov · 2025-09-16T09:12:41Z

Codecov Report

❌ Patch coverage is 18.91026% with 253 lines in your changes missing coverage. Please review.
✅ Project coverage is 75.50%. Comparing base (e6356e7) to head (3e3b5ec).

Files with missing lines	Patch %	Lines
src/scenario.F90	34.17%	104 Missing ⚠️
src/output.F90	0.00%	44 Missing ⚠️
src/partmc.F90	0.00%	37 Missing ⚠️
src/run_modal.F90	0.00%	37 Missing ⚠️
src/aero_dist.F90	0.00%	30 Missing ⚠️
src/spec_file.F90	83.33%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #211      +/-   ##
==========================================
- Coverage   77.45%   75.50%   -1.95%     
==========================================
  Files          55       56       +1     
  Lines        9420     9713     +293     
==========================================
+ Hits         7296     7334      +38     
- Misses       2124     2379     +255

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jcurtis2

Looks pretty good in my first pass.

I think we need to add some protections to the modal model. For example, I can attempt to run with emissions/dilution (it won't complain) but they won't do anything. However these are in the spec file (they're zero for the purpose of this scenario but it leads one to think that maybe they could be changed) and they are read in (due to other PartMC subroutines). I think maybe we just add some asserts in partmc_modal to warn/error out on unsupported behavior.
Similarly, I could also attempt to turn coagulation on. This isn't actually read anywhere. I guess what we really need is an input_format_modal similar to how we have them for input_format_particle, sectional to say what is supported.
We do need to figure out what to do with the scenario. Honestly I'm wondering if we should consider switching this scenario to being a test case (this is what we did with the immersion freezing). This scenario already fits in pretty well with what is found in test/loss. Making it a test would also make codecov happy. Alternatively, I think if it wants to reach scenario levels by making it interesting, we could consider running it for the particle-resolved case and/or sectional case. Just something to consider.

jcurtis2 · 2026-04-23T20:29:26Z

+
+#ifdef PMC_USE_QUADPACK
+     ! Do numerical integration when QUADPACK is available
+     print *, "***USING QUADPACK FOR DRY DEPOSITION***"


I don't think we need this - will print frequently.

Suggested change

print *, "***USING QUADPACK FOR DRY DEPOSITION***"

jcurtis2 · 2026-04-23T20:36:16Z

+    if (ier > 0) then
+     call die_msg(837465, "QUADPACK integration failed (error code: " &
+                         // trim(integer_to_string(ier)))
+    endif


Suggest that we switch to assert_msg. Test for ier = zero rather than ier > 0 (unless there is a good reason for testing >0 which would include negatives (if quadpack allows it?)). And replace number with random 9 digit number. Suggestion would look somewhat like the following (admittedly I didn't test it yet)

Suggested change

if (ier > 0) then

call die_msg(837465, "QUADPACK integration failed (error code: " &

// trim(integer_to_string(ier)))

endif

call assert_msg(909106718, ier == 0, &

"QUADPACK integration failed, error code: " &

// trim(integer_to_string(ier)))

jcurtis2 · 2026-04-23T20:42:00Z

+       if (new_N < 1d0) then
+          aero_dist%mode(i_mode)%num_conc = 0d0
+          aero_dist%mode(i_mode)%char_radius = 0d0
+          cycle
+       end if


Is this some sort of cutoff for small number concentrations? I think a comment would help.

And is it necessary to set the char_radius to 0 or is num_conc = 0 sufficient? I can envision this could cause some issue (maybe currently only in post-processing) if char_radius = 0 is ever used in a lognormal equation.

jcurtis2 · 2026-04-23T20:42:31Z

+    else if (scenario%loss_function_type == SCENARIO_LOSS_FUNCTION_CONSTANT) then
+       return
+    else if (scenario%loss_function_type == SCENARIO_LOSS_FUNCTION_DRYDEP) then
+      ! loss


Suggested change

! loss

jcurtis2 · 2026-04-23T20:44:17Z

+    !> Density for each mode.
+    real(kind=dp), intent(in) :: density


Comment says each mode but density is a scalar.

jcurtis2 · 2026-04-23T20:45:48Z

+    !> Dry deposition parameters.
+    type(drydep_params_t), intent(inout) :: drydep_params
+
+    !> \page input_format_chamber Input File Format: Dry Deposition Parameters


Copy/pasted \page from elsewhere (chamber.F90). Should be renamed.

jcurtis2 · 2026-04-23T20:46:53Z

+      do i_mode = 1,aero_dist_n_mode(aero_dist)
+       aero_mode = aero_dist%mode(i_mode)
+       N = aero_mode%num_conc
+
+       if (N == 0d0) then
+          cycle
+       end if
+
+       d_pg = aero_mode%char_radius * 2.0d0
+       ln_sigma_g = aero_mode%log10_std_dev_radius / log10(exp(1.0d0))
+
+       ! Integrated deposition rate for the 0-th moment (num. conc.)
+       m_0_rate = -1.0d0 * scenario_integrated_loss_rate_drydep(scenario, aero_mode, &
+                   0.0d0, density, env_state)
+       new_N = N * exp(m_0_rate * del_t)
+
+       if (new_N < 1d0) then
+          aero_dist%mode(i_mode)%num_conc = 0d0
+          aero_dist%mode(i_mode)%char_radius = 0d0
+          cycle
+       end if
+
+       aero_dist%mode(i_mode)%num_conc = new_N
+
+       ! Integrated deposition rate for the 3-rd moment (proportional to volume conc.)
+       m_3_rate = -1.0d0 * scenario_integrated_loss_rate_drydep(scenario, aero_mode, 3.0d0, &
+                                                                 density, env_state)
+       M = N * d_pg**3.0d0 * exp((3.0d0**2.0d0)/2.0d0 * (ln_sigma_g**2.0d0))
+       new_M = M * exp(m_3_rate * del_t)
+
+       ! New geometric mean diameter
+       new_d_pg = (new_M / new_N * exp(-(3.0d0**2.0d0)/2.0d0 * (ln_sigma_g**2.0d0)))**(1.0d0/3.0d0)
+       aero_dist%mode(i_mode)%char_radius = new_d_pg / 2.0d0
+      end do


Fix indentation on the do loop.

jcurtis2 · 2026-04-23T20:48:22Z

@@ -0,0 +1,33 @@
+run_type modal                       # modal run
+output_prefix data/modal_dpg_00001000000000000000_sig_2_5_emerson_grass


Suggest that you change the directory to out (consistent with other scenarios and also specified in the dockerignore)

jcurtis2 · 2026-04-23T20:49:19Z

+
+# The data should have already been generated by ./1_run_dry_dep_modal.sh
+
+../../build/drydep_modal_process


Complains that it needs an output prefix.

jcurtis2 · 2026-04-23T20:50:26Z

+#     dens (kg/m^3)   ions in soln (1)    molec wght (kg/mole)   kappa (1)
+SO4            1800                0                   96d-3      0.65


Need to tack on the two new freezing parameters. Should be something like this:

Suggested change

# dens (kg/m^3) ions in soln (1) molec wght (kg/mole) kappa (1)

SO4 1800 0 96d-3 0.65

# dens (kg/m^3) ions in soln (1) molec wght (kg/mole) kappa (1) abifm_m (1) abifm_c (1)

SO4 1800 0 96d-3 0.65 0. 0.

jcurtis2 · 2026-04-23T21:22:10Z

+    integer :: total_size, i, N
+


Looks like these variables tagged along from elsewhere and aren't needed here (or actually even where you probably got it from). But clean them up here only so this PR stays focused on drydep changes.

Suggested change

integer :: total_size, i, N

zdaq12 added 23 commits June 7, 2025 10:59

adding scaffolding for modal run

0f786d4

add modal run to driver

9e60333

add run_modal.F90 to CMakeLists

3d70db2

add input and output routines for modal run

8ac4cd2

add input and output routines for modal parameters

6526f22

fix use statements and typos

f3dd453

fix typos

2872bdc

add subroutines for computing integrated deposition velocities for th…

fbe3daf

…e modal run; handle input and output for processing

add support for QUADPACK; comment out dry dep scenario for now

b27627f

remove debugging code for modal run

e530f6f

clean up implementation of numerical integration with QUADPACK

562c066

formatting

374d571

remove unused flag for emissions/dilution

537b2c3

fix conditional for inputting dry deposition parameterization

6da5eeb

allow for reading drydep_params from text file otherwise fall back on…

bdc2de6

… default parameters

remove debugging and commented code

ca4142b

update MPI routines to handle drydep parameters

a855f69

update processing script for modal runs and add exectuable

093b334

update aero_dist netcdf input and output for modal runs to operate wi…

cc7ff7e

…th arrays

remove duplicated line

6f186e9

update dockerignore

cd9d90e

dry_dep --> drydep

ed72197

add input files and scripts for modal drydep runs

d2191d5

slayoo closed this Sep 16, 2025

slayoo reopened this Sep 16, 2025

zdaq12 added 3 commits September 20, 2025 08:54

update doc for spec_file_read_drydep_params

02c7e11

fix aero_dist naming convention

1c3610c

drydep_params.dat comments update

eeb1c47

zdaq12 added 2 commits October 8, 2025 14:44

text file clean up

e7e23ea

newlines at end of files

bd9f931

zdaq12 marked this pull request as ready for review October 8, 2025 20:03

jcurtis2 self-requested a review December 4, 2025 04:47

zdaq12 and others added 2 commits April 23, 2026 14:44

minor refactor

d3444d6

Merge branch 'master' into final_modal

3e3b5ec

jcurtis2 reviewed Apr 23, 2026

View reviewed changes

		!> Density for each mode.
		real(kind=dp), intent(in) :: density

		@@ -0,0 +1,33 @@
		run_type modal # modal run
		output_prefix data/modal_dpg_00001000000000000000_sig_2_5_emerson_grass


		# The data should have already been generated by ./1_run_dry_dep_modal.sh

		../../build/drydep_modal_process

		# dens (kg/m^3) ions in soln (1) molec wght (kg/mole) kappa (1)
		SO4 1800 0 96d-3 0.65

Conversation

zdaq12 commented Jul 14, 2025

Uh oh!

slayoo commented Sep 16, 2025

Uh oh!

codecov Bot commented Sep 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

jcurtis2 left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jcurtis2 Apr 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

codecov Bot commented Sep 16, 2025 •

edited

Loading

jcurtis2 left a comment •

edited

Loading

jcurtis2 Apr 23, 2026 •

edited

Loading