Update icepack with negative silicate bug fix

[eclare108213]

Icepack BGC bug fix for silicate limitation, which was mistakenly wrapped in an if statement for iron tracers that caused negative values.

Fixes #7127

[non-BFB] only with active BGC, BFB otherwise.

[eclare108213]

This bug fix was thoroughly tested using Icepack's test suites, which confirmed that it is BFB for pure physics runs but not for BGC runs. See CICE-Consortium/Icepack#519

It was then merged into E3SM-Project/Icepack and tested against E3SM master using the E3SM-Polar_Developer.sh script in 3-month D case:

 --- Average Differences for Month 03, Year 0001:

    D3.nseti0.update_icepack_20250402.icepack-fix-negative-silicate.E3SM-Project.chrysalis k000 - 
    D3.nseti0.update_icepack_20250402.master.E3SM-Project.chrysalis k000:

    iceAreaCell 	BFB
    iceVolumeCell 	BFB
    icePressure 	BFB
    uVelocityGeo 	BFB
    vVelocityGeo 	BFB

    totalIceExtent 	BFB
    totalIceVolume 	BFB
    totalSnowVolume 	BFB
    totalKineticEnergy 	BFB
    averageAlbedo 	BFB

@njeffery has also tested it in a fully-coupled biogeochemistry simulation. Simulation page and comparisons with a control run are here:
https://acme-climate.atlassian.net/wiki/spaces/HESF/pages/5121212417/20250314.v3.LR.CBGC.FrazilEcosys.noSpinup.pm-cpu

[proteanplanet]

@eclare108213 Please can you share the namelist settings for the above D-case test?

[eclare108213]

They were entirely the defaults:

(base) [ac.eclare@chrlogin1 PolarGroup]$ less nseti0.nlk 

[mpassi]


nseti0.nlk (END)

[proteanplanet]

Approve on the basis that this change does not affect any physics-only simulations on the sea ice model in E3SM.

[njeffery]

Approved. Tested within a 20-year fully coupled biogeochemistry simulation.

[jonbob]

Passes:

• SMS_D_Ld1.ne30pg2_r05_IcoswISC30E3r5.WCYCL1850.chrysalis_intel.allactive-wcprod
• SMS_Ld2.ne30pg2_r05_IcoswISC30E3r5.BGCEXP_CNTL_CNPECACNT_1850.chrysalis_intel.elm-bgcexp

merged to next

[jonbob]

merged to master