The PalMod project, funded by the German Federal Ministry of Education and Research (BMBF) under grant number 01LP1919B, aimed to quantify the contribution of fundamental processes to the climate dynamics and variability of the last glacial cycle using comprehensive Earth system models (ESMs). The sub‑project PalModII‑WP2.1‑SP3, led by Dr. Tatiana Ilyina at the Max‑Planck‑Institute for Meteorology, ran from 1 September 2019 to 30 April 2023, with a cost‑neutral extension until 30 April 2023. The focus was on transient simulations of ocean biogeochemistry and its interaction with the atmosphere during the last deglaciation, a task that had not previously been undertaken at full ESM scale.

Technically, the team implemented and extended the Hamburg Ocean Carbon Cycle Model (HAMOCC) within the MPI‑ESM to enable interactive carbon cycle simulations. Two distinct circulation states for the Last Glacial Maximum (LGM, 21 ka) were calibrated, and two marine particle sinking schemes were compared. The combination of a weak, flat AMOC and the prognostic “M4AGO” sinking scheme maximised deep‑ocean carbon storage and improved agreement with glacial Atlantic observations. In the first transient deglacial run with prognostic CO₂, the model showed that CO₂ outgassing is primarily driven by surface‑water warming. A comparison of runs with different HAMOCC tuning revealed that current observational validation approaches, typical of CMIP6, are insufficient, highlighting critical aspects for future model tuning.

The model‑data comparison of the oceanic ¹³C Suess effect in the present ocean uncovered two major methodological uncertainties in observational data products, underscoring the need to use models not only to reproduce data but also to interpret uncertainties. In the LGM CO₂ simulations, all PalMod models consistently produced a CO₂ drop to about 190 ppmv when LGM alkalinity was increased by 5–10 %, a result that will be further examined in Phase III transient and future simulations.

The project delivered three main outputs: D2.1‑1 (transient deglacial simulation without interactive carbon cycle), D2.1‑4 (with interactive carbon cycle), and D2.1‑2 (transient last‑glacial simulation without interactive carbon cycle). The omission of a glacial inception run was a deliberate decision due to limited resources and a shift in scientific focus toward the last glaciation and future climate.

Collaboration was extensive. The sub‑project worked closely with PalMod partners in Working Group 1 and Work Packages 2.2 and 2.3 at MPI‑M, coordinating model development, tuning, and the calculation of deglacial weathering fluxes. The post‑doctoral researcher Dr. Bo Liu, funded for 36 months, led the technical work and contributed to the parallel computation of biogeochemical tracers, accelerating transient runs for Phase III. The team also engaged with the broader palaeo‑community, contributing to the first PMIP carbon multimodel comparison and co‑organising the upcoming PMIP carbon activity scheduled for April 2024. Funding was primarily allocated to personnel costs for the post‑doc, with additional support for conference travel and workshops. The project’s outcomes lay the groundwork for Phase III, where the focus will remain on the climate‑carbon cycle coupling of the last glaciation and future scenarios, building on the technical advances and collaborative framework established in this phase.