Add per-region dust emission multiplier

[davidfillmore]

Summary

Commit f92d1a3 adds a 7-element namelist array dst_rgn_scale that multiplies the horizontal saltation flux in dst/dstmbl.F per named dust source region. Gridpoints outside all regions use scale 1.0.

Changes

• src/control.com — declare dst_rgn_scale(7) in namelist /nlist/
• src/main.F — initialize defaults to 1.0
• dst/dstmbl.F — region-box logic, applies the multiplier to flx_mss_hrz_slt_ttl alongside lnd_frc_mbl and flx_mss_fdg_fct

Region boundaries (hardcoded lat/lon boxes)

Idx	Region	Lat	Lon
1	Sahara/Sahel	10–30N	20W–30E
2	Arabia	15–30N	30–55E
3	Central Asia	30–50N	55–90E
4	Gobi/Taklamakan	35–50N	90–115E
5	Australia	35–15S	115–145E
6	SW N. America	25–40N	115W–100W
7	Patagonia	55–35S	75W–60W

Motivation

The older global soil-moisture rescale (vwc_scale / vwc_offset) cannot simultaneously match multiple source regions against assimilated AOD — hyperarid Sahara (CORe SOILW ≈ 0.02 → max emission) needs strong suppression while moderate Gobi/Australia are under-emitting. A per-region multiplier decouples them. The legacy global parameters remain wired in for backwards compatibility but are superseded.

Known limitations of the current source

• Region boxes are hardcoded in dst/dstmbl.F. Adding a region or relocating a box requires a source edit, not a namelist change.
• The SW N. America box as defined contains little baseline emission, so its multiplier has no leverage — the box needs relocation to actual source zones before it can be tuned.
• The region test runs once per gridpoint per call; for very high-resolution grids the box-membership checks could be precomputed into an integer region index field.
• No diagnostic output of which region (if any) each gridpoint falls in — would help verify boundary placement.

Follow-ups (separate issues)

• Tuning campaign to fit dst_rgn_scale per calendar month against assimilated AOD climatologies — see AOD_TUNE.md.
• Sea salt scaling (sslt_scale) — added in 3914c2e, separate concern.