Enhancing Glacial Lake Monitoring Through CNN-Based UNet Models in a Streamlit Application

Glacial lake monitoring is important to learn about climate change and to address challenges such as GLOFs in the future. This paper presents an improved UNet model based on CNN applied to a Streamlit application for glacial lake monitoring. The system takes input of both image and csv datasets, divides glacial lakes, computes the rates of growth and generates risk along with alerts in real time. The results are stored in the cloud, in Google Firebase shall be added, and the report is generated automatically. Through the application, the users are allowed to upload images or ask questions with parameters’ help, improving the utilization of glacial lake studies.