Use historical aerosol data for 1849 and 2023

[penguian]

Closes #116 again. See also #117 for the previous pull request. The difference is that this change uses years 1849 and 2023 from the datasets, instead of duplicating 1850 and 2022.

[penguian]

Tested using u-dq819.trunk@342770. Run output is in gadi:/scratch/tm70/pcl851/cylc-run/u-dq819.trunk/run86. Ancillary output is in gadi:/g/data/tm70/pcl851/CMIP7/esm1p6_ancil/2026-01-06.119:

/g/data/tm70/pcl851/CMIP7/esm1p6_ancil/2026.01.06.119
└── modern
    ├── historical
    │   └── atmosphere
    │       └── aerosol
    │           └── global.N96
    │               └── 2026.01.06
    │                   ├── BC_1849_2023_cmip7.anc
    │                   ├── Bio_1849_2023_cmip7.anc
    │                   ├── OCFF_1849_2023_cmip7.anc
    │                   └── scycl_1849_2023_cmip7.anc
    └── pre-industrial
        └── atmosphere
            └── aerosol
                └── global.N96
                    └── 2026.01.06
                        ├── BC_1850_cmip7.anc
                        ├── Bio_1850_cmip7.anc
                        ├── OCFF_1850_cmip7.anc
                        └── scycl_1850_cmip7.anc

[penguian]

The current input path for CEDS anthropogenic emissions is
gadi:/g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/PNNL-JGCRI/CEDS-CMIP-2025-04-18/atmos/mon/{species}_em_anthro/gn/v20250421/{species}-em-anthro_input4MIPs_emissions_CMIP_CEDS-CMIP-2025-04-18_gn_{date_range}.nc.

The current paths for historical DRES biomass burning emissions are
gadi: /g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/DRES/DRES-CMIP-BB4CMIP7-2-1/atmos/mon/{species}/gn/v20250612/{species}_input4MIPs_emissions_CMIP_DRES-CMIP-BB4CMIP7-2-1_gn_{date_range}.nc
for each species, and
gadi:/g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/DRES/DRES-CMIP-BB4CMIP7-2-1/atmos/mon/{species}percentage{source}/gn/v20250612/{species}percentage{source}_input4MIPs_emissions_CMIP_DRES-CMIP-BB4CMIP7-2-1_gn_{date_range}.nc
for percentages from sources.

[blimlim]

Not sure how informative the global totals for the aerosols are, but the following show the global total emissions for each variable at the 1849-1850 and 2022-2023 boundaries. For most of the variables, the totals are quite similar with the new and old versions:

[blimlim]

Thanks @penguian, the changes look good to me!