Automatic volcanic ash detection from MODIS observations using a back-propagation neural network
Due to the climate effects and aviation threats of volcanic eruptions, it is important to accurately locate ash in the atmosphere. This study aims to explore the accuracy and reliability of training a neural network to identify cases of ash using observations from the Moderate Resolution Imaging Spe...
Published in: | Atmospheric Measurement Techniques |
---|---|
Main Authors: | , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Copernicus Publications
2015
|
Subjects: | |
Online Access: | https://doi.org/10.5194/amt-8-5089-2015 https://doaj.org/article/45ee9ebef3dd470d9879012914def10e |
id |
ftdoajarticles:oai:doaj.org/article:45ee9ebef3dd470d9879012914def10e |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:45ee9ebef3dd470d9879012914def10e 2023-05-15T16:52:01+02:00 Automatic volcanic ash detection from MODIS observations using a back-propagation neural network T. M. Gray R. Bennartz 2015-12-01T00:00:00Z https://doi.org/10.5194/amt-8-5089-2015 https://doaj.org/article/45ee9ebef3dd470d9879012914def10e EN eng Copernicus Publications http://www.atmos-meas-tech.net/8/5089/2015/amt-8-5089-2015.pdf https://doaj.org/toc/1867-1381 https://doaj.org/toc/1867-8548 1867-1381 1867-8548 doi:10.5194/amt-8-5089-2015 https://doaj.org/article/45ee9ebef3dd470d9879012914def10e Atmospheric Measurement Techniques, Vol 8, Iss 12, Pp 5089-5097 (2015) Environmental engineering TA170-171 Earthwork. Foundations TA715-787 article 2015 ftdoajarticles https://doi.org/10.5194/amt-8-5089-2015 2022-12-31T14:43:36Z Due to the climate effects and aviation threats of volcanic eruptions, it is important to accurately locate ash in the atmosphere. This study aims to explore the accuracy and reliability of training a neural network to identify cases of ash using observations from the Moderate Resolution Imaging Spectroradiometer (MODIS). Satellite images were obtained for the following eruptions: Kasatochi, Aleutian Islands, 2008; Okmok, Aleutian Islands, 2008; Grímsvötn, northeastern Iceland, 2011; Chaitén, southern Chile, 2008; Puyehue-Cordón Caulle, central Chile, 2011; Sangeang Api, Indonesia, 2014; and Kelut, Indonesia, 2014. The Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model was used to obtain ash concentrations for the same archived eruptions. Two back-propagation neural networks were then trained using brightness temperature differences as inputs obtained via the following band combinations: 12–11, 11–8.6, 11–7.3, and 11 μm. Using the ash concentrations determined via HYSPLIT, flags were created to differentiate between ash (1) and no ash (0) and SO 2 -rich ash (1) and no SO 2 -rich ash (0) and used as output. When neural network output was compared to the test data set, 93 % of pixels containing ash were correctly identified and 7 % were missed. Nearly 100 % of pixels containing SO 2 -rich ash were correctly identified. The optimal thresholds, determined using Heidke skill scores, for ash retrieval and SO 2 -rich ash retrieval were 0.48 and 0.47, respectively. The networks show significantly less accuracy in the presence of high water vapor, liquid water, ice, or dust concentrations. Significant errors are also observed at the edge of the MODIS swath. Article in Journal/Newspaper Iceland Aleutian Islands Directory of Open Access Journals: DOAJ Articles Atmospheric Measurement Techniques 8 12 5089 5097 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
Environmental engineering TA170-171 Earthwork. Foundations TA715-787 |
spellingShingle |
Environmental engineering TA170-171 Earthwork. Foundations TA715-787 T. M. Gray R. Bennartz Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
topic_facet |
Environmental engineering TA170-171 Earthwork. Foundations TA715-787 |
description |
Due to the climate effects and aviation threats of volcanic eruptions, it is important to accurately locate ash in the atmosphere. This study aims to explore the accuracy and reliability of training a neural network to identify cases of ash using observations from the Moderate Resolution Imaging Spectroradiometer (MODIS). Satellite images were obtained for the following eruptions: Kasatochi, Aleutian Islands, 2008; Okmok, Aleutian Islands, 2008; Grímsvötn, northeastern Iceland, 2011; Chaitén, southern Chile, 2008; Puyehue-Cordón Caulle, central Chile, 2011; Sangeang Api, Indonesia, 2014; and Kelut, Indonesia, 2014. The Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model was used to obtain ash concentrations for the same archived eruptions. Two back-propagation neural networks were then trained using brightness temperature differences as inputs obtained via the following band combinations: 12–11, 11–8.6, 11–7.3, and 11 μm. Using the ash concentrations determined via HYSPLIT, flags were created to differentiate between ash (1) and no ash (0) and SO 2 -rich ash (1) and no SO 2 -rich ash (0) and used as output. When neural network output was compared to the test data set, 93 % of pixels containing ash were correctly identified and 7 % were missed. Nearly 100 % of pixels containing SO 2 -rich ash were correctly identified. The optimal thresholds, determined using Heidke skill scores, for ash retrieval and SO 2 -rich ash retrieval were 0.48 and 0.47, respectively. The networks show significantly less accuracy in the presence of high water vapor, liquid water, ice, or dust concentrations. Significant errors are also observed at the edge of the MODIS swath. |
format |
Article in Journal/Newspaper |
author |
T. M. Gray R. Bennartz |
author_facet |
T. M. Gray R. Bennartz |
author_sort |
T. M. Gray |
title |
Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
title_short |
Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
title_full |
Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
title_fullStr |
Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
title_full_unstemmed |
Automatic volcanic ash detection from MODIS observations using a back-propagation neural network |
title_sort |
automatic volcanic ash detection from modis observations using a back-propagation neural network |
publisher |
Copernicus Publications |
publishDate |
2015 |
url |
https://doi.org/10.5194/amt-8-5089-2015 https://doaj.org/article/45ee9ebef3dd470d9879012914def10e |
genre |
Iceland Aleutian Islands |
genre_facet |
Iceland Aleutian Islands |
op_source |
Atmospheric Measurement Techniques, Vol 8, Iss 12, Pp 5089-5097 (2015) |
op_relation |
http://www.atmos-meas-tech.net/8/5089/2015/amt-8-5089-2015.pdf https://doaj.org/toc/1867-1381 https://doaj.org/toc/1867-8548 1867-1381 1867-8548 doi:10.5194/amt-8-5089-2015 https://doaj.org/article/45ee9ebef3dd470d9879012914def10e |
op_doi |
https://doi.org/10.5194/amt-8-5089-2015 |
container_title |
Atmospheric Measurement Techniques |
container_volume |
8 |
container_issue |
12 |
container_start_page |
5089 |
op_container_end_page |
5097 |
_version_ |
1766042157388398592 |