Background
The authors write that "explosive volcanic eruptions inject into the lower stratosphere millions of tons of chemically and micro-physically active gases and solid aerosol particles," and noting that these substances "are recognized as an important climate forcing," they go on to say that "efforts are underway to accommodate their impact in the assessment of future climates in recent IPPC reports."

What was done
The objective of this enterprise, according to Li and Sharma, has historically been sought by "assessing the response of hydrometeorological variables, especially atmospheric water vapor, to volcanic eruptions by using climate data from GCMs, reanalysis products, and observations." And they say that the focus of their investigation was "to assess (1) whether atmospheric water vapor as well as other hydrological variables (pressure, specific humidity, latent heat and precipitation) are affected by large-scale volcanic eruptions and (2) whether model simulations of this effect on atmospheric water vapor exhibit similarity to what is noted in observations and reanalysis data."

What was learned
As a result of their efforts in this regard, the two Australian researchers reported discovering that (1) "the percentage of global area with volcanic impacts on water vapor was overestimated based on a CMIP3 multi-model average and a single CMIP5 model," that (2) "the spatial pattern of the water vapor variability following the Pinatubo eruption was not well captured by GCMs," that (3) "the amplitude of the water vapor change from the model simulations is much smaller than that from observed and reanalysis data," that (4) "the observed temporal pattern of the water vapor variability was also not as distinct as depicted in GCMs," that (5) "the model-simulated strong negative correlation between atmospheric water vapor residual and aerosol optical depth in the tropics was also not found to be representative of the observed and reanalysis data," and that (6) "the effect of volcanic forcing may have been represented in an overly simplified manner in the CMIP3 multi-model mean ... resulting in water vapor simulations that exhibit different distributional attributes compared to the observed record."

What it means
In concluding their report, Li and Sharma state that "despite the remarkable improvements of GCMs over the past years, it is still a challenge to simulate volcanic impacts for all GCMs, which has also been confirmed by Driscoll et al. (2012), who examined 12 CMIP5 GCMs and found that they all overestimated the cooling in the tropical troposphere following the nine explosive eruptions in the nineteenth and twentieth centuries."

Reference
Driscoll, S., Bozzo, A., Gray, L.J., Robock, A. and Stenchikov, G. 2012. Coupled model intercomparison project 5 (CMIP5) simulations of climate following volcanic eruptions. Journal of Geophysical Research 117: 1.01029/2012JD017607.