Investigating the use of the Saharan dust index as a tool for the detection of volcanic ash in SEVIRI imagery

摘要: Abstract Despite the similar spectral signatures of ash and desert dust, relatively little has been done to explore application dust detection techniques problem volcanic detection. The Saharan index (SDI) is routinely implemented for monitoring at some centres could be utilised with computational expense, thereby providing a product that forecasters already have familiarity complement suite existing tools. We illustrate one way in which purpose by applying it three scenes containing from 2010 Eyjafjallajokull eruption, Iceland 2011 eruption Puyehue, Chile. It was also applied an image acquired over Etna January 2011, where plume clearly visible but unlikely contain any ash. These examples demonstrate potential SDI as tool under different environmental atmospheric conditions. In addition presenting valuable qualitative aid monitoring, this work includes quantitative assessment skill using manually constructed expert mask. optimum implementation technique likely dependent on both conditions properties imaged (which often unknown real-time situation). Here we take advantage access ‘truth’ rarely available situation calculate mask based threshold specific scene, then used SDI. compared masks calculated simplistic more traditional split window method, again exploiting our set most appropriate each probabilistic method without reference provides useful insights into cloud-/ash-contamination pixel. Since sensitivity methods tailored thresholds not tested (such tailoring possible real situation), study presents maximum anticipated context although are lower situation. results comparable those other methods, true score 80.02% night-time scene (compared 88.81% 46.63% respectively) 90.06% day-time 97.61% 56.96%). For Puyehue image, resulted 74.85%, while approach achieved 99.62%. imply SDI, operationally detection, techniques.