M. Ramsey, Department of Geology and Planetary Science, University of Pittsburgh, Pittsburgh, PA

D. Schneider, U.S. Geological Survey Alaska Science Center, Alaska Volcano Observatory, Anchorage, AK

R. Wessels, U.S. Geological Survey Alaska Science Center, Alaska Volcano Observatory, Anchorage, AK

M. Clark, Department of Geology and Planetary Science, University of Pittsburgh, Pittsburgh, PA

A helicopter-mounted forward-looking infrared radiometer (FLIR) was deployed by the Cascades Volcano Observatory (CVO) during the initial phases of renewed volcanic activity at Mount St. Helens (MSH). The instrument began collecting data on October 1, 2004, and was flown twice daily (weather permitting) during the first month of the eruption. Thermal infrared (TIR) data collected during the first seven flight-days imaged the 1980 crater floor, phreatic explosion pits, and ash cooling on the surface. However, it was not until October 11, 2004, that the FLIR data confirmed a lava body was present at the surface. After the initial extrusion of new dome rock, subsequent surveys from the helicopter-mounted and a hand-held FLIR were used to model its rise rate and volume. During the pre-extrusive phase of the eruption, the FLIR camera commonly imaged the same regions of the older lava dome and portions of the uplifted crater floor. Qualitative changes in the heat flux were noted at the time, but the rapid escalation in activity and long days of field work precluded a more quantitative approach.

In an effort to understand the early phases of the eruption process, FLIR data from times prior to October 11 have been re-analyzed. Data from these days were ideal because of good weather conditions, relative stability of ground surface features, and a minimal amount of plume within the crater itself. Two distinct geomorphic features identifiable in each day's data were clipped from the full resolution images. The data were first corrected for distance, atmospheric pathlength energy, and surface emissivity and then converted to radiant flux. Assuming the radiative heat output at these locations is related to the conductive heat loss from the magma body below, heat balance models can be employed to estimate the depth of the magma. Assumptions that must be made for such an approach include: 1) the average pre-eruptive temperature for dacitic magma; 2) the contribution of convective heat loss; and 3) the variability of the thermal conductivity and heat transfer coefficient terms with depth. Preliminary model results of the daily FLIR surveys predict an average magma rise rate similar to that observed following dome eruption onto the surface of the crater floor. However, this modeling does not take into account multi-dimensional or convective heat loss. Therefore, work is ongoing to further constrain the magma body depth by comparing the results from the thermal modeling with other geophysical and petrologic studies of the conduit system. A study such as this confirms the utility and importance of high temporal resolution FLIR surveys at restless volcanoes prior to lava extrusion. It also reinforces the need for development of thermal modeling processing routines to cope with the rapid data flow during the precursory stage of an eruption. For example, with an adequate amount of time and resources in the first week of FLIR surveys, such a study could have been performed to better constrain when potential lava would have reached the surface at Mount St. Helens.

------------------
Submitted: American Geophysical Union Fall Meeting
Date: December 5 - 9, 2004