资源环境科技发展态势分析平台

Radiative forcing from volcanic aerosol impacts surface temperatures; however, the background climate state also affects the response. A key question thus concerns whether constraining forcing estimates is more important than constraining initial conditions for accurate simulation and attribution of posteruption climate anomalies. Here we test whether different realistic volcanic forcing magnitudes for the 1815 Tambora eruption yield distinguishable ensemble surface temperature responses. We perform a cluster analysis on a superensemble of climate simulations including three 30-member ensembles using the same set of initial conditions but different volcanic forcings based on uncertainty estimates. Results clarify how forcing uncertainties can overwhelm initial-condition spread in boreal summer due to strong direct radiative impact, while the effect of initial conditions predominate in winter, when dynamics contribute to large ensemble spread. In our setup, current uncertainties affecting reconstruction-simulation comparisons prevent conclusions about the magnitude of the Tambora eruption and its relation to the "year without summer."

Plain Language Summary Strong volcanic eruptions are a major natural forcing of climate, and there is increasing awareness of their importance for understanding and prediction of climate. Much of our current understanding is based on past eruptions for which there is no direct observation and that are therefore subject to substantial uncertainties regarding the eruption characteristics, the associated volcanic forcing, and the climatic response. With these premises, our study tackles the question of whether, for a past or analogously future eruption, it is more important to constrain the uncertainty in the volcanic forcing it generates rather than the initial climate conditions upon which it occurs, in order to obtain robust information about the associated climate responses. We use the 1815 eruption of Mount Tambora as a test case. We demonstrate that the climatic effects of initial condition uncertainties can overwhelm those of forcing uncertainties especially in winter and at the regional scale. We show how the European "year without a summer" seen in climate reconstructions is compatible with very different combinations of forcing magnitude and initial conditions.