Digital Repository

Deep Learning for Super-Resolution of Meteorological Data

Show simple item record

dc.contributor.advisor Kumar, Bipin en_US
dc.contributor.author ATEY, KAUSTUBH en_US
dc.date.accessioned 2022-05-13T13:55:47Z
dc.date.available 2022-05-13T13:55:47Z
dc.date.issued 2022-05
dc.identifier.citation 65 en_US
dc.identifier.uri http://dr.iiserpune.ac.in:8080/xmlui/handle/123456789/6947
dc.description.abstract For any country, understanding and interpreting precipitation dynamics is of great importance. In India, rainfall patterns profoundly influence daily livelihoods and economic development. Thus, it is essential to get cognizance of local rainfall for better policy making. Often, downscaling methods are used to produce high-resolution projections from low-resolution GCM outputs or observation data. This study applies a deep generative model called SRGAN to statistically downscale precipitation data from IMD over the Indian region. Our analysis shows that SRGAN performs comparatively better than other deep learning methods used for downscaling. We used SRGAN to downscale the precipitation data from 0.25° to 0.125° and 0.0625° resolutions and found that the downscaling results closely matched station observations. We also introduce a custom trained VGG based feature extractor that can act as a backbone for other DL models using meteorological data. Our study establishes that SRGAN can be used as a reliable statistical downscaling model. SRGAN yields result faster than the dynamical RCMs, allowing for more practical real-time applications. en_US
dc.language.iso en en_US
dc.subject Deep Learning en_US
dc.subject Super-Resolution en_US
dc.subject Generative Adversarial Networks en_US
dc.subject Meteorological Data en_US
dc.title Deep Learning for Super-Resolution of Meteorological Data en_US
dc.type Thesis en_US
dc.type.degree BS-MS en_US
dc.contributor.department Interdisciplinary en_US
dc.contributor.registration 20171172 en_US


Files in this item

This item appears in the following Collection(s)

  • MS THESES [1705]
    Thesis submitted to IISER Pune in partial fulfilment of the requirements for the BS-MS Dual Degree Programme/MSc. Programme/MS-Exit Programme

Show simple item record

Search Repository


Advanced Search

Browse

My Account