Date of Award

8-28-2020

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Systems Engineering

First Advisor

Ibrahim Nisanci

Abstract

Atmospheric studies have been carried out for decades to forecast weather and climate trends such as tropical storms, etc. Of recent concern is the level of greenhouse gases in the atmosphere, which has a relative influence on the weather and climate trends. Over the years, large weather satellites like the GOES series have been used to monitor these GHGs. However, an in-depth understanding of the movement of these gases requires consistent monitoring, which has given rise to the need for miniaturized satellites. Since the miniaturized satellites are novel, they require a ground-truth like the GOES-16 satellite to validate these observations. However, the remotely sensed imagery from the GOES-16 is hampered by the cloud, which disrupts accurate GHG observations. The Cloud-Net (FCN cloud detection) model is used for the cloud pixel detection in the downloaded GOES-16 imagery and, creates a cloud mask to be used for further analysis of water vapor imagery.

Included in

Engineering Commons

Share

COinS