Dear all,
I am using CESM2.1.5 with BHIST_BPRP at f09_g17 for paired simulations starting in 1985. I have customized land cover/land use datasets that are identical in 1985 and diverge afterward. However, I have no customized data before 1985.
I am considering two spinup approaches:
1. Fully coupled pre-integration: Starting from the official 1940 BHIST_BPRP restart files, I will simulate 1984 while holding the land cover and land use at my customized 1985 baseline. During this time, the other historical forcings will advance normally. This approach would use normal CLM mode without AD. Both experiments would then start from the resulting 1985 coupled state.
2. Land-only spinup: Use the same 1985 land cover/land use baseline in an I1850Clm50BgcCropCru case. Run it for 150 years in AD mode, followed by 350 years in normal mode. Then, use the resulting CLM restart, together with suitable initial and restart files for the other components, to initialize BHIST_BPRP. In this case, the 1985 data refers to land cover and land use, not meteorological forcing.
What are the main differences and trade-offs between these approaches for BHIST_BPRP? Is it preferable to inherit the historical restart and perform the coupled pre-integration rather than re-equilibrate the land carbon pools offline?
Thank you for any guidance.
I am using CESM2.1.5 with BHIST_BPRP at f09_g17 for paired simulations starting in 1985. I have customized land cover/land use datasets that are identical in 1985 and diverge afterward. However, I have no customized data before 1985.
I am considering two spinup approaches:
1. Fully coupled pre-integration: Starting from the official 1940 BHIST_BPRP restart files, I will simulate 1984 while holding the land cover and land use at my customized 1985 baseline. During this time, the other historical forcings will advance normally. This approach would use normal CLM mode without AD. Both experiments would then start from the resulting 1985 coupled state.
2. Land-only spinup: Use the same 1985 land cover/land use baseline in an I1850Clm50BgcCropCru case. Run it for 150 years in AD mode, followed by 350 years in normal mode. Then, use the resulting CLM restart, together with suitable initial and restart files for the other components, to initialize BHIST_BPRP. In this case, the 1985 data refers to land cover and land use, not meteorological forcing.
What are the main differences and trade-offs between these approaches for BHIST_BPRP? Is it preferable to inherit the historical restart and perform the coupled pre-integration rather than re-equilibrate the land carbon pools offline?
Thank you for any guidance.
