Geoscientific Model Development (Mar 2021)

PMIP4 experiments using MIROC-ES2L Earth system model

  • R. Ohgaito,
  • A. Yamamoto,
  • T. Hajima,
  • R. O'ishi,
  • M. Abe,
  • H. Tatebe,
  • A. Abe-Ouchi,
  • A. Abe-Ouchi,
  • A. Abe-Ouchi,
  • M. Kawamiya

DOI
https://doi.org/10.5194/gmd-14-1195-2021
Journal volume & issue
Vol. 14
pp. 1195 – 1217

Abstract

Read online

Following the protocol of the fourth phase of the Paleoclimate Modelling Intercomparison Project (PMIP4), we performed numerical experiments targeting distinctive past time periods using the Model for Interdisciplinary Research on Climate, Earth System version 2 for Long-term simulations (MIROC-ES2L), which is an Earth system model. Setup and basic performance of the experiments are presented. The Last Glacial Maximum was one of the most extreme climate states during the Quaternary and conducting numerical modeling experiments of this period has long been a challenge for the paleoclimate community. We conducted a Last Glacial Maximum experiment with a long spin-up of nearly 9000 years. Globally, there was reasonable agreement between the anomalies relative to the present day derived from model climatology and those derived from proxy data archives, while some regional discrepancies remained. By changing orbital and greenhouse gas forcings, we conducted experiments for two interglacial periods: 6000 and 127 000 years before present. Model anomalies relative to the present day were qualitatively consistent with variations in solar forcing. However, anomalies in the model were smaller than those derived from proxy data archives, suggesting that processes that play a role in past interglacial climates remain lacking in this state-of-the-art model. We conducted transient simulations from 850 to 1850 CE and from 1850 to 2014 CE. Cooling in the model indicated a clear response to huge volcanic eruptions, consistent with paleo-proxy data. The contrast between cooling during the Little Ice Age and warming during the 20th to 21st centuries was represented well at the multidecadal timescale.