15 July 2016

Volcanic contribution to decadal changes in tropospheric temperature

Nature Geoscience
Science

Objective

  • Improve scientific understanding of the volcanic contribution to the post-1998 “warming hiatus”

Research

  • To determine whether signals of early 21st century volcanic activity are identifiable in satellite observations
  • To understand the causes of differences between modeled and observed tropospheric warming trends
Impact
  • Early 21st century volcanic activity had a cooling influence on surface and tropospheric temperature
  • This cooling influence is statistically identifiable in satellite observa(ons of correlations between stratospheric aerosol optical depth (SAOD) and tropospheric temperature (and between SAOD and SW fluxes at the top of atmosphere)

  • Omitting the cooling effect of early 21st century volcanic activity from CMIP-5 simulations of historical climate change contributes to a model-average warm bias in tropospheric temperature trends

  • Results do not support claims of a factor of three error in model climate sensitivity estimates

Summary

Despite continued growth in atmospheric levels of greenhouse gases, global-mean surface and tropospheric temperatures show minimal increases since 1998. Possible explanations for this “warming hiatus” include internal climate variability, external cooling influences, and observational errors. One contributory factor to the relatively muted surface warming – early 21st century volcanic forcing – has been examined in several modelling studies. Here, we present the first analysis of the impact of recent volcanic forcing on tropospheric temperature, and the first observational assessment of the significance of early 21st century volcanic signals. We find statistically significant signals in the correlations between stratospheric aerosol optical depth and satellite-based estimates of tropospheric temperature/short-wave fluxes at the top of the atmosphere. We show that climate model simulations without early 21st century volcanic activity overestimate the tropospheric warming observed since 1998. In two simulations with more realistic volcanic forcing following the 1991 Pinatubo eruption, differences between modeled and observed tropospheric temperature trends over 1998 to 2012 are decreased by up to 15%, with large uncertainties in the size of the effect. Reducing these uncertainties will require better observational understanding of eruption-specific differences in volcanic aerosol properties, and improved representation of these differences in model simulations.

Contact
Benjamin Santer
Lawrence Livermore National Laboratory (LLNL)