Predicting wind-driven spatial deposition through simulated color images using deep autoencoders
Abstract For centuries, scientists have observed nature to understand the laws that govern the physical world. The traditional process of turning observations into physical understanding is slow. Imperfect models are constructed and tested to explain relationships in data. Powerful new algorithms ca...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-28590-4 |
| _version_ | 1811175891452035072 |
|---|---|
| author | M. Giselle Fernández-Godino Donald D. Lucas Qingkai Kong |
| author_facet | M. Giselle Fernández-Godino Donald D. Lucas Qingkai Kong |
| author_sort | M. Giselle Fernández-Godino |
| collection | DOAJ |
| description | Abstract For centuries, scientists have observed nature to understand the laws that govern the physical world. The traditional process of turning observations into physical understanding is slow. Imperfect models are constructed and tested to explain relationships in data. Powerful new algorithms can enable computers to learn physics by observing images and videos. Inspired by this idea, instead of training machine learning models using physical quantities, we used images, that is, pixel information. For this work, and as a proof of concept, the physics of interest are wind-driven spatial patterns. These phenomena include features in Aeolian dunes and volcanic ash deposition, wildfire smoke, and air pollution plumes. We use computer model simulations of spatial deposition patterns to approximate images from a hypothetical imaging device whose outputs are red, green, and blue (RGB) color images with channel values ranging from 0 to 255. In this paper, we explore deep convolutional neural network-based autoencoders to exploit relationships in wind-driven spatial patterns, which commonly occur in geosciences, and reduce their dimensionality. Reducing the data dimension size with an encoder enables training deep, fully connected neural network models linking geographic and meteorological scalar input quantities to the encoded space. Once this is achieved, full spatial patterns are reconstructed using the decoder. We demonstrate this approach on images of spatial deposition from a pollution source, where the encoder compresses the dimensionality to 0.02% of the original size, and the full predictive model performance on test data achieves a normalized root mean squared error of 8%, a figure of merit in space of 94% and a precision-recall area under the curve of 0.93. |
| first_indexed | 2024-04-10T19:44:22Z |
| format | Article |
| id | doaj.art-a3352c71d14044c188d68314a5cbc95e |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-10T19:44:22Z |
| publishDate | 2023-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-a3352c71d14044c188d68314a5cbc95e2023-01-29T12:09:46ZengNature PortfolioScientific Reports2045-23222023-01-0113111210.1038/s41598-023-28590-4Predicting wind-driven spatial deposition through simulated color images using deep autoencodersM. Giselle Fernández-Godino0Donald D. Lucas1Qingkai Kong2Lawrence Livermore National LaboratoryLawrence Livermore National LaboratoryLawrence Livermore National LaboratoryAbstract For centuries, scientists have observed nature to understand the laws that govern the physical world. The traditional process of turning observations into physical understanding is slow. Imperfect models are constructed and tested to explain relationships in data. Powerful new algorithms can enable computers to learn physics by observing images and videos. Inspired by this idea, instead of training machine learning models using physical quantities, we used images, that is, pixel information. For this work, and as a proof of concept, the physics of interest are wind-driven spatial patterns. These phenomena include features in Aeolian dunes and volcanic ash deposition, wildfire smoke, and air pollution plumes. We use computer model simulations of spatial deposition patterns to approximate images from a hypothetical imaging device whose outputs are red, green, and blue (RGB) color images with channel values ranging from 0 to 255. In this paper, we explore deep convolutional neural network-based autoencoders to exploit relationships in wind-driven spatial patterns, which commonly occur in geosciences, and reduce their dimensionality. Reducing the data dimension size with an encoder enables training deep, fully connected neural network models linking geographic and meteorological scalar input quantities to the encoded space. Once this is achieved, full spatial patterns are reconstructed using the decoder. We demonstrate this approach on images of spatial deposition from a pollution source, where the encoder compresses the dimensionality to 0.02% of the original size, and the full predictive model performance on test data achieves a normalized root mean squared error of 8%, a figure of merit in space of 94% and a precision-recall area under the curve of 0.93.https://doi.org/10.1038/s41598-023-28590-4 |
| spellingShingle | M. Giselle Fernández-Godino Donald D. Lucas Qingkai Kong Predicting wind-driven spatial deposition through simulated color images using deep autoencoders Scientific Reports |
| title | Predicting wind-driven spatial deposition through simulated color images using deep autoencoders |
| title_full | Predicting wind-driven spatial deposition through simulated color images using deep autoencoders |
| title_fullStr | Predicting wind-driven spatial deposition through simulated color images using deep autoencoders |
| title_full_unstemmed | Predicting wind-driven spatial deposition through simulated color images using deep autoencoders |
| title_short | Predicting wind-driven spatial deposition through simulated color images using deep autoencoders |
| title_sort | predicting wind driven spatial deposition through simulated color images using deep autoencoders |
| url | https://doi.org/10.1038/s41598-023-28590-4 |
| work_keys_str_mv | AT mgisellefernandezgodino predictingwinddrivenspatialdepositionthroughsimulatedcolorimagesusingdeepautoencoders AT donalddlucas predictingwinddrivenspatialdepositionthroughsimulatedcolorimagesusingdeepautoencoders AT qingkaikong predictingwinddrivenspatialdepositionthroughsimulatedcolorimagesusingdeepautoencoders |