Karymsky eruption and balloon launches in 2003, 2004, 2005.
Power Point presentation


We have done the first-ever-made attempts to sample ash clouds of erupting volcanoes.
Sampling device was carried into eruptive clouds of Karymsky volcano by tethered balloons filled with helium.


To know grain-size distribution of pyroclasts in eruptive cloud.


One of important problems of volcanology is REAL grain size distribution of pyroclastic material in eruptive cloud. Without knowing that it is impossible to calculate precisely erupted volumes of magma, to understand processes of magma fragmentation, to predict effects of eruptions on climate, and to estimate volcanic danger to aircraft. Currently the problem is solved only approximately, mainly by means of various calculations of the original grain size basing on grain size analyses of ash-fall deposits collected on the ground. The calculations are necessary because ash-fall deposits result from complex gravitational settling of the original population of pyroclasts from turbulent eruption cloud through layer of air. As a result larger particles fall out in more proximal areas to the source than do smaller ones, while finest particles can stay suspended for years in atmosphere, and thus their percentage can not be calculated precisely. Percentage of the finest fraction is the key question. Rough estimations show that mass of fines can be at least equal to, or up to ten times exceed the mass of larger particles, which form ash-fall deposits. Volume of fine-grained pyroclastic material ejected by one large eruption can reach hundreds cubic kilometers.


Obviously, the only reliable solution of the above problem, is DIRECT sampling of eruptive cloud. Although sounds unrealistic, for small-to-moderate-scale persistent eruptions that can be done easily using rather simple equipment. There are no principle obstacles to sample also clouds of big eruptions. To determine the original grain-size distribution it is necessary to catch and isolate some volume of the eruptive cloud and after complete settling of pyroclasts, make standard analysis of the sample (sieve, pipette, laser particle counter, SEM). Such as the most interesting question is content of finest fraction, volume of the sample should not be very large - fines will be statistically represented even in several liters - several tens liters of eruptive cloud, depending on concentration. To carry sampler into the cloud, the unmanned aircraft can be used. Seems strange, but to our knowledge, until now no attempts to do that were made.


In summer 2003 we have made the fist ever made attempt to sample eruptive cloud. Karymsky was selected as relatively safe volcano producing regular small-to-moderate scale explosions of vulcanian type. Cone of the volcano is 1540 m. high, with relative elevation about 700 m. In the time of the experiment, explosions occurred every 5 - 45 min, ejecting incandescent bombs up to 0.5 km high, and convective ash cloud rose up to 1,5 km above the crater (3 km a.s.l.). To carry the sampler into eruptive cloud we used a tethered balloon filled with helium. The tether was rolled in and rolled out with a hand winch. Two types of balloons were used: Zeppelin-shaped plastic aerostat and standard latex weather balloon. Each balloon contained about 6 cub.m. of helium. Two basic schemes of sampling were tried: launch from location downwind from the crater (simple lifting of the balloon in drifting eruptive cloud) and launch from upwind location (when wind tilted the balloon toward the crater). In the first scheme, sampling device was working in already cooled and strongly diluted part of the eruptive cloud. Conditions inside the cloud were not destructive for the balloon, thus sampler was directly attached to it. The low-temperature sampler was a 30-liters plastic box like a big mouse trap, which had electronic triggering device based on a dust sensor. In the second scheme the eruptive cloud was thought to be still hot and containing high concentration of large strongly abrasive particles. Thus the balloon in this configuration ought to fly higher than the eruptive cloud, and sampler was hanging down from the balloon on a long (500m) metal line. The 10-liters high-temperature sampler was also like a mouse trap, but smaller and made of metal. Triggering mechanism represented a plastic filament which kept the lid of the sampler in opened condition. When the sampler entered hot cloud the filament melted allowing the lid to be closed.


Totally 4 launches were made with the maximum achieved altitude about 2,5 km above the launching site and 1,7 km above the crater (3300 a.s.l.). Although the sampling itself was not achieved, the experiment have demonstrated principle validity of the designed samplers as well as sampling procedure. Main reason of sampling failure was that we used rather cheap types of balloons, which were not sturdy enough to withstand harsh weather conditions on the volcano. Apart from budget shortages, cheap balloons were used because risk to loose an expensive balloon due to tether breaking seemed rather high. Our experience have shown that main danger to balloons is their breaking in strong wind. Even with cheap balloons we were very close to our goal. The experiment allowed us to receive valuable experience in operating balloons above erupting volcano. Also we understood how to improve the equipment for the future success. The experiment has been continued at 2004 and 2005.