One of the major questions surrounding all volcanic activity is: what controls the transitions between different styles of activity? This applies both on longer timescales of years to decades and even centuries but also on shorter ones of seconds to minutes and hours. The latter consideration is my particular area of interest. In a video of activity leading up to the main lava fountain (see video half way down the page), we can see that smaller strombolian like explosions seem to increase in frequency and vigor leading up to the lava fountain (indeed patterns have been noticed in strombolian activity at Etna - a shameless plug of some of my own work). Could this pattern be analysed to predict the occurrence of a lava fountain before it occurs? How could we do this? The answer lies in monitoring gas emissions - a difficult task at night and during more ash rich eruptions!
The transition between strombolian type activity and Hawaiian occurs when magma rise speed is high enough such that it prevents the coalescence of bubbles (which are rising at a slower rate than the magma) which drive strombolian style activity when bubbles grow to a large enough size(see this paper by Parfitt and Wilson).