Uncertainty in two-channel infrared remote sensing retrievals of a well-characterised volcanic ash cloud

This work provides a sensitivity study of a two-channel passive infrared remote sensing retrieval of effective radius and optical depth using the Spinning Enhanced Visible and Infrared Imager with channels centred at 10.8 and 12.0 μm and a look-up table approach to calculate mass column loading. The...

Full description

Bibliographic Details
Published in:Bulletin of Volcanology
Main Authors: Western, Luke, Watson, Matthew, Francis, Peter
Format: Article in Journal/Newspaper
Language:English
Published: 2015
Subjects:
Online Access:https://hdl.handle.net/1983/39c1604c-9e4d-41cf-8932-f0596c47c1ac
https://research-information.bris.ac.uk/en/publications/39c1604c-9e4d-41cf-8932-f0596c47c1ac
https://doi.org/10.1007/s00445-015-0950-y
Description
Summary:This work provides a sensitivity study of a two-channel passive infrared remote sensing retrieval of effective radius and optical depth using the Spinning Enhanced Visible and Infrared Imager with channels centred at 10.8 and 12.0 μm and a look-up table approach to calculate mass column loading. The retrieval is applied to images of two ash clouds from the 2010 Eyjafjallajökull eruption on 6 and 13 May 2010. The 2010 eruption of Eyjafjallajökull is well characterised, especially in terms of the airborne volcanic ash, which allows the relative uncertainties to be investigated within the realms of observation and reasonable approximation. The parameters investigated are as follows: refractive index, surface temperature, cloud top temperature, ash bulk density and, in particular, the uncertainties related to the spread of the ash particle size distribution—in terms of the geometric standard deviation of a lognormal particle size distribution. The lack of constraint on particle size distribution is shown to cause the largest uncertainty in retrieved mass column loading for the 6 May and 13 May ash cloud. A review of measured in situ size distributions of airborne particles is presented with justification for the choice of a lognormal size distribution for the 2010 Eyjafjallajökull eruption.