Reconstructions, Code, and Analysis for "Coupled Atmosphere-Ocean Reconstruction of the Last Millennium Using Online Data Assimilation"

This contains the code, results, and analysis from the publication "Coupled Atmosphere-Ocean Reconstruction of the Last Millennium Using Online Data Assimilation".

History

v1.0 - initial upload

v1.1 - Fix scaling issue with the MPI, offline and piControl reconstruction spatial fields from a bug in the conversion to condensed netcdfs.  Add spatial ensemble variance for all reconstruction experiements.

analysis_notebooks.tar.gz

Notebooks containing code used to produce the paper figures.

reconstruction_output.tar.gz

Reduced reconstruction output from all experiments. 

Each experiment includes the full ensemble output (50 MC iterations x 100 ensemble members) of climate indices, the grand ensemble average of the spatial fields, and spatial ensemble variance averaged over the MC iterations.

Indices include:

• PDO, Nino 3, 3.4, and 4, NPI, SOI [ as two gridpoints representing darwin & tahiti locations], and the SAM [as zonal averages of 45S and 65S]
• global averages of SST, TAS, TOA outgoing longwave, TOA outgoing shortwave, 0-700m OHC
• regional averages (incl. Europe and N. Atlantic) and hemispheric averages of TAS

Spatial Fields Include:

• surface precipitation (pr_ sfc)
• sea level pressure (psl_sfc)
• TOA outgoing longwave radiation (rlut_toa)
• TOA outgoing shortwave radiation (rsut_toa)
• surface air temperature (tas_sfc)
• 500 hPa geopotential height (zg_500hPa)
• sea surface salinity (sos_sfc)
• sea surface temperature (tos_sfc)
• dynamic ocean height (zos_sfc)
• 0-700 m ocean heat content (ohc_0-700m)

code_and_input_data.tar.gz

LMROnline and pyLIM code repositories, configuration files for the main manuscript reconstructions, and an input data directory (LMRdata) sufficient for reproducing the reconstruction results.

WARNING

We provide all spatial field output from the reconstruction experiment, but only the TAS, TOS, and OHC fields have been validated against instrumental data.  We strongly suggest users take caution with the other provided fields, as there is no guarantee of skillful reconstruction.  Please validate, explore, and compare the CCSM4 and MPI reconstructions for these fields before deciding whether they are useful to you.