Determination of mass loadings and plume heights of volcanic ash clouds from satellite data

A retrieval scheme for determining some microphysical properties of volcanic ash clouds from thermal infrared satellite data is described. The mathematical details are provided here, while the broader implications of this research are outlined in a paper published by the Quarterly Journal (Prata and Grant, 2001). The report also provides details of the methods used to determine ash cloud heights from their shadows cast on land or ocean. The optical properties of a model andesite ash cloud are tabulated. These may be used in conjunction with data from advanced satellite sensors to infer ash cloud microphysics. Radiative transfer (RT) calculations were also performed to investigate the relationship between top-of-the-atmosphere brightness temperatures (or radiances) in the split-window channels and water vapour amount for some current and planned satellite radiometers (e.g. AVHRR, ATSR-2, GMS-5 and GLI). The input data to the RT model come from the NCEP re-analysis and provide information on a spatial scale of 2.5 x 2.5 degrees and temporal resolution of 6 hours. Strong correlations were found between the difference in brightness temperatures (referred to as T4-T5) in the split-window channels and precipitable water, and between brightness temperature in one channel and precipitable water. Thus it follows that the relationship between the brightness temperature in one channel and the brightness temperature difference is a strong function of the precipitable water. Based on this, an empirical relationship is developed to describe the behaviour of T4T5 as a function of T4 for global conditions. The relationship is then used to provide a water vapour correction for a modified volcanic ash detection scheme. The modified scheme retrieves the fraction of ash present in a pixel and is based on a simple model of the T4-T5 signature arising from ash and the empirical water vapour correction formula. Image-based estimates of the clear-sky, ash-free T4-T5, the cloud-top temperature and the clear-sky surface temperature are required. Results using the modified scheme are shown for four test case eruptions.