Eruption History and Pyroclastic Flow Modeling at Azufral Volcano, Colombia.

摘要

In this work, the eruptive history and morphology of Azufral volcano, Colombia is explored and analyzed to provide a more complete picture of the past eruptions, as well as infer what eruption styles may occur in the future (Chapter 1). Through the use of principal component analysis, domes can be correlated to the pyroclastic deposits, enabling the identification of a full eruptive sequence. Geomorphology of the dome complex is reinterpreted to gain a better grasp on growth sequence and identify a previously unrecognized dome. This reinterpretation, along with geochemical analysis, allows for reclassification of a deposit, originally thought to be a juvenile block-and-ash flow, as a volcanic debris avalanche. The findings overall demonstrate that eruptive activity at Azufral volcano is largely explosive, experiencing long periods of quiescence, punctuated by short periods of pyroclastic activity.;In the course of the application of a geophysical mass flow model (TITAN2D) to Azufral, problems with the digital elevation model were observed motivating Chapter 2. The integer valued high-resolution 3 m DEM contained what is now termed "terracing", caused by the inability of this data precision to approximate the surface. The appearance of this phenomenon is large areas of non-real zero surface slope, causing decreased runout distances of simulated flows. Tests were performed using the model in conjunction with a synthetic DEM (i.e., a numerically defined surface), and natural data (an unprocessed floating point precision 3 m DEM). Experiments show that terracing in integer precision DEMs artificially decrease runout of simulated flows, with the greatest effect in high-resolution DEMs. Comparison of model runs with natural DEM data show that precision effects are secondary to resolution effects, at least for the 3 m DEM used here.;Probabilistic PDC hazard maps accounting for model input uncertainty are also created for Azufral volcano, and presented in Chapter 3. Three PHMs are presented here following a new methodology that limits the range of basal friction angle and volume values input into the model, to better approximate the volume-dependent basal friction angle apparent in natural PDCs. This method is, however, more computationally expensive than the original single PHM method. In the process of the creation of these PHMs, it was necessary to use different data sets (3 m DEM and 30 m SRTM). Results from Chapter 2 can be extrapolated to PHMs, alluding to the need for testing and comparison of PHMs constructed with different DEMs of the same area. The PHMs are finally compared to the maximum extents of the previously mapped deposits at Azufral. It is found that the larger deposits of Cortadera and Espino extend past the bounds of the Scenario 2 PHM but lie within the Scenario 3 PHM.;Major results of this work can be summarized as follows: 1) The eruptive history of Azufral has been slightly better defined with implications toward eruption style and time scales of eruptions. 2) Use of high-resolution digital elevation models with geophysical mass flow models, like TITAN2D, require the user to understand the quality of the inputs. High-resolution integer-valued DEMs and DEMs with poor surface approximation should be used very cautiously. 3) A presentation of a new PHM methodology that better approximates apparent basal friction angles of natural PDCs.