Rising temperatures over the Greater Mekong Subregion in CMIP6: Present-day biases and constraint future projections

Abstract

This study evaluates the historical simulation of surface air temperature (SAT) over the Greater Mekong Subregion (GMS) during the period 1981–2014 with 31 models from Coupled Model Intercomparison Project phase 6 (CMIP6). The CMIP6 multi-model ensemble shows a weaker SAT growth trend than the reanalysis data in all seasons. The SAT increase is underestimated mostly in winter, by around 0.2 °C/decade. The simulated SAT climatology shows positive bias over the Gaoligong Mountain in western Yunnan, the Rakhine Mountain in Myanmar, and the Annamite Range in southern Vietnam, and negative bias elsewhere. We make a best-model ensemble to improve the representability of the GMS SAT. The range of SAT interannual variability is underestimated over most parts of GMS in boreal spring, autumn, and winter. In summer, the SAT range bias shows a dipole pattern with negative bias over the northwestern GMS and positive bias over the southeastern GMS. Dynamical diagnosis suggests that the bias in vertical velocity perturbation, manifesting as weakened regional Hadley circulation, is a major contributor to the SAT range bias in summer via vertical advection of mean potential temperature. The bias-corrected projection of the summer SAT is presented for three shared socio-economic pathways (SSPs). The SAT rising relative to the reference period 1995–2014 tends to be more spatially homogeneous under higher emission scenarios. The range of SAT interannual variability would increase over the northern GMS but decrease over the southern GMS under three SSPs scenarios. This SAT dipole variation is associated with intensified regional Hadley circulation over the GMS. The results also reiterate the urgent need for net zero emission by 2050 to maintain a sustainable environment.