Environmental Research Letters (Jan 2020)

Annual and seasonal mean tropical and subtropical precipitation bias in CMIP5 and CMIP6 models

  • J-L F Li,
  • Kuan-Man Xu,
  • Mark Richardson,
  • Wei-Liang Lee,
  • J H Jiang,
  • Jia-Yuh Yu,
  • Yi-Hui Wang,
  • Eric Fetzer,
  • Li-Chiao Wang,
  • Graeme Stephens,
  • Hsin-Chien Liang

DOI
https://doi.org/10.1088/1748-9326/abc7dd
Journal volume & issue
Vol. 15, no. 12
p. 124068

Abstract

Read online

Annual and seasonal mean tropical and subtropical precipitation biases in present-day climate simulations by phase 6 of the Coupled Model Intercomparison Project (CMIP6) models are evaluated. We analyze 23 CMIP6 models, which are grouped into those without (NOS) and with (SON) falling ice (snow) radiative effects (FIREs). The SON group is further divided according to treatments of ice-cloud radiative properties with separate (SON2) and combined cloud ice and snow contents (SON1). SON2 mitigates precipitation bias drastically relative to NOS, but SON1 overestimates precipitation in the subtropics, which deteriorates its performance relative to SON2 and NOS. Large offsets between SON2 and SON1 and between SON and NOS contribute to the slight improvement of CMIP6 relative to CMIP5, largely due to the contribution of SON2. The different treatments of ice-cloud radiative properties between SON2 and SON1 seem to be potentially linked to differences in precipitation biases. Results suggest that further in-depth investigation of cloud ice–radiation–circulation interactions is required.

Keywords