DECK and Water Cycle Experiments Water cycle science

  • Slides: 7
Download presentation
DECK and Water Cycle Experiments

DECK and Water Cycle Experiments

Water cycle science questions • What forcings dominate water cycle changes in the past

Water cycle science questions • What forcings dominate water cycle changes in the past and how may this change in the future? • Does increasing model resolution improve simulations of water cycle processes and how? • How well/much do variable resolution simulations reproduce the global high resolution simulations of water cycle processes? • What are the pros and cons of different strategies for initializing coupled simulations and how do they affect water cycle simulations?

DECK experiments (low resolution - LR) • LR PI control: 500 years of simulation

DECK experiments (low resolution - LR) • LR PI control: 500 years of simulation with PI forcing (1 member) • LR historical: branch off after 300 years of PI control for another 165 simulations with historical forcings out to 2015 (3 ensemble members) • LR 4 x. CO 2: branch off after 300 years of PI control to abruptly increase CO 2 to 4 x. CO 2 for 150 years to 2000 (1 member) • LR 1%/yr CO 2 increase: branch off after 300 years of PI control to increase CO 2 by 1%/yr for 150 years to 2000 (1 member)

Water cycle: LR simulations • • LR PI control-A: simulation initialized from CORE-II MPAS

Water cycle: LR simulations • • LR PI control-A: simulation initialized from CORE-II MPAS low resolution ocean/ice run with 1850 climatological forcing for 100 years (this is different from LR PI control in the DECK experiments) LR transient with all forcings-A: initialize using LR PI control-A on year 1950 with all forcings for 100 years (1950 – 2050) (3 ensemble members) - first 20 years used as adjustment LR transient with GHG-only forcing: initialize using LR PI control-A on year 1950 with GHG-only forcing for 100 years (1950 -2050) (3 ensemble members) - first 20 years used as adjustment LR transient with GHG+aerosols only forcing: initialize using LR control-A on year 1950 with GHG+aerosols only forcing for 100 years (1950 -2050) (3 ensemble members) first 20 years used as adjustment LR transient with all forcings-B: initialize using CORE-II MPAS low resolution ocean/ice run with historical atmospheric forcing for 100 years (1950 -2050, first 20 years used as spinup) LR transient with all forcings-C: extend LR historical (from DECK) for 2015 -2050 Total = 100 + 100 x 3 + 100 + 35 = 1135 years

Water cycle: High resolution (HR) simulations • HR PI control-A: simulation initialized from CORE-II

Water cycle: High resolution (HR) simulations • HR PI control-A: simulation initialized from CORE-II MPAS high resolution ocean/ice run with 1850 climatological forcing for 100 years • HR transient with all forcings-A: initialize using HR PI control-A on year 1950 with all forcings for 100 years (1950 -2050) (1 member) - first 20 years used as adjustment • HR transient with all forcings-B: initialize using CORE-II MPAS high resolution ocean/ice run with historical atmospheric forcing for 100 years (1950 -2050, first 20 years used as spinup) • Total = 100 + 100 = 300 years • Extend to 3 ensemble members in 2018

Water cycle: Variable resolution (VR) simulations • VR PI control-A: simulation initialized from CORE-II

Water cycle: Variable resolution (VR) simulations • VR PI control-A: simulation initialized from CORE-II MPAS variable resolution ocean/ice run with 1850 climatological forcing for 100 years • VR transient with all forcings-A: initialize using VR PI control-A on year 1950 with all forcings for 100 years (1950 -2050) (1 member) - first 20 years used as adjustment • Total = (100 + 100) x 2 regions = 400 years; the two regions with refinement are over the US and Asia • Extend to 3 or more members in 2018

Analysis • Science question (1) will be answered only using low resolution simulations by

Analysis • Science question (1) will be answered only using low resolution simulations by comparing the various LR transient simulations with different forcings initialized from LR PI control-A for 1970 -2050. • Science question (2) will be answered by comparing HR and LR transient with all forcings-A and all forcings-B for 1970 -2010 with observations. • Science question (3) will be answered by comparing simulations with all forcings-A for VR in the low resolution region with global LR, and VR in the refined region with global HR for 1970 -2050. • Science question (4) will be answered by comparing (a) LR PI control and LR PI control-A; (b) LR transient all forcings-A, all forcings-B, and all forcings-C; (c) HR transient all forcings-A and HR transient all forcings-B for 1970 -2050.