Limits...
Probabilistic detection of volcanic ash using a Bayesian approach.

Mackie S, Watson M - J Geophys Res Atmos (2014)

Bottom Line: Such techniques are generally specific to data from particular sensors, and most approaches result in a binary classification of pixels into "ash" and "ash free" classes with no indication of the classification certainty for individual pixels.The technique has already been successfully applied to other detection problems in remote sensing, and this work shows that it will be a useful and effective tool for ash detection.Presentation of a probabilistic volcanic ash detection schemeMethod for calculation of probability density function for ash observationsDemonstration of a remote sensing technique for monitoring volcanic ash hazards.

View Article: PubMed Central - PubMed

Affiliation: School of Earth Sciences, University of Bristol Bristol, UK.

ABSTRACT

: Airborne volcanic ash can pose a hazard to aviation, agriculture, and both human and animal health. It is therefore important that ash clouds are monitored both day and night, even when they travel far from their source. Infrared satellite data provide perhaps the only means of doing this, and since the hugely expensive ash crisis that followed the 2010 Eyjafjalljökull eruption, much research has been carried out into techniques for discriminating ash in such data and for deriving key properties. Such techniques are generally specific to data from particular sensors, and most approaches result in a binary classification of pixels into "ash" and "ash free" classes with no indication of the classification certainty for individual pixels. Furthermore, almost all operational methods rely on expert-set thresholds to determine what constitutes "ash" and can therefore be criticized for being subjective and dependent on expertise that may not remain with an institution. Very few existing methods exploit available contemporaneous atmospheric data to inform the detection, despite the sensitivity of most techniques to atmospheric parameters. The Bayesian method proposed here does exploit such data and gives a probabilistic, physically based classification. We provide an example of the method's implementation for a scene containing both land and sea observations, and a large area of desert dust (often misidentified as ash by other methods). The technique has already been successfully applied to other detection problems in remote sensing, and this work shows that it will be a useful and effective tool for ash detection.

Key points: Presentation of a probabilistic volcanic ash detection schemeMethod for calculation of probability density function for ash observationsDemonstration of a remote sensing technique for monitoring volcanic ash hazards.

No MeSH data available.


Related in: MedlinePlus

Example of the PDF for cloud for one ST-TCWV group (TCWV > 10 kg/m2, 280 K < ST < 300 K), for daytime observations made over sea.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4379904&req=5

fig03: Example of the PDF for cloud for one ST-TCWV group (TCWV > 10 kg/m2, 280 K < ST < 300 K), for daytime observations made over sea.

Mentions: The inferred cloud observations for each profile are weighted according to the relative likelihood assumed for each of the simulated clouds, given the atmospheric profile. The distribution of cloud observations for each individual profile is convolved with uncertainties using equation 5, as for the clear-sky case, creating a PDF for cloud specific to the atmospheric profile. The weightings are discussed in section 3.2.1, and the uncertainties are discussed in section 3.2.2. Since the profiles are assumed to be representative of all spatial and temporal variations in the noncloud parameters to which the observations are most sensitive, the profile-specific PDFs could all be averaged to create a PDF for cloud that is independent of the atmospheric data. This would be equivalent to setting P(y/ccloud) = P(y/x,ccloud) in equation 1. Some atmospheric conditions make certain types of cloud more likely than others, and it is preferable that our definition of a cloud class reflects a distribution of cloud properties that is realistic for the atmosphere above the pixel rather than being representative of all atmospheric conditions. To achieve this, profiles in the data set are grouped according to ST and TCWV and the profile-specific PDFs corresponding to each group are averaged to create group-specific PDFs for cloud, which are used as LUTs for the detection. For a given pixel, the appropriate LUT from which to read P(y/x,ccloud) is selected according to the ST and TCWV attributed to the pixel. An example is shown in Figure 3. Previous studies have noted that these two parameters affect the IR spectral signature for ash and can therefore affect the performance of more traditional detection methods, necessitating various correction techniques, e.g., Yu et al. [2002], and so using these to characterize an atmosphere for the purposes of ash detection is appropriate. Frequency distributions for ST and TCWV within the profiles data set were used to decide appropriate groupings; see Figure 4 and Table  2.


Probabilistic detection of volcanic ash using a Bayesian approach.

Mackie S, Watson M - J Geophys Res Atmos (2014)

Example of the PDF for cloud for one ST-TCWV group (TCWV > 10 kg/m2, 280 K < ST < 300 K), for daytime observations made over sea.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4379904&req=5

fig03: Example of the PDF for cloud for one ST-TCWV group (TCWV > 10 kg/m2, 280 K < ST < 300 K), for daytime observations made over sea.
Mentions: The inferred cloud observations for each profile are weighted according to the relative likelihood assumed for each of the simulated clouds, given the atmospheric profile. The distribution of cloud observations for each individual profile is convolved with uncertainties using equation 5, as for the clear-sky case, creating a PDF for cloud specific to the atmospheric profile. The weightings are discussed in section 3.2.1, and the uncertainties are discussed in section 3.2.2. Since the profiles are assumed to be representative of all spatial and temporal variations in the noncloud parameters to which the observations are most sensitive, the profile-specific PDFs could all be averaged to create a PDF for cloud that is independent of the atmospheric data. This would be equivalent to setting P(y/ccloud) = P(y/x,ccloud) in equation 1. Some atmospheric conditions make certain types of cloud more likely than others, and it is preferable that our definition of a cloud class reflects a distribution of cloud properties that is realistic for the atmosphere above the pixel rather than being representative of all atmospheric conditions. To achieve this, profiles in the data set are grouped according to ST and TCWV and the profile-specific PDFs corresponding to each group are averaged to create group-specific PDFs for cloud, which are used as LUTs for the detection. For a given pixel, the appropriate LUT from which to read P(y/x,ccloud) is selected according to the ST and TCWV attributed to the pixel. An example is shown in Figure 3. Previous studies have noted that these two parameters affect the IR spectral signature for ash and can therefore affect the performance of more traditional detection methods, necessitating various correction techniques, e.g., Yu et al. [2002], and so using these to characterize an atmosphere for the purposes of ash detection is appropriate. Frequency distributions for ST and TCWV within the profiles data set were used to decide appropriate groupings; see Figure 4 and Table  2.

Bottom Line: Such techniques are generally specific to data from particular sensors, and most approaches result in a binary classification of pixels into "ash" and "ash free" classes with no indication of the classification certainty for individual pixels.The technique has already been successfully applied to other detection problems in remote sensing, and this work shows that it will be a useful and effective tool for ash detection.Presentation of a probabilistic volcanic ash detection schemeMethod for calculation of probability density function for ash observationsDemonstration of a remote sensing technique for monitoring volcanic ash hazards.

View Article: PubMed Central - PubMed

Affiliation: School of Earth Sciences, University of Bristol Bristol, UK.

ABSTRACT

: Airborne volcanic ash can pose a hazard to aviation, agriculture, and both human and animal health. It is therefore important that ash clouds are monitored both day and night, even when they travel far from their source. Infrared satellite data provide perhaps the only means of doing this, and since the hugely expensive ash crisis that followed the 2010 Eyjafjalljökull eruption, much research has been carried out into techniques for discriminating ash in such data and for deriving key properties. Such techniques are generally specific to data from particular sensors, and most approaches result in a binary classification of pixels into "ash" and "ash free" classes with no indication of the classification certainty for individual pixels. Furthermore, almost all operational methods rely on expert-set thresholds to determine what constitutes "ash" and can therefore be criticized for being subjective and dependent on expertise that may not remain with an institution. Very few existing methods exploit available contemporaneous atmospheric data to inform the detection, despite the sensitivity of most techniques to atmospheric parameters. The Bayesian method proposed here does exploit such data and gives a probabilistic, physically based classification. We provide an example of the method's implementation for a scene containing both land and sea observations, and a large area of desert dust (often misidentified as ash by other methods). The technique has already been successfully applied to other detection problems in remote sensing, and this work shows that it will be a useful and effective tool for ash detection.

Key points: Presentation of a probabilistic volcanic ash detection schemeMethod for calculation of probability density function for ash observationsDemonstration of a remote sensing technique for monitoring volcanic ash hazards.

No MeSH data available.


Related in: MedlinePlus