MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202611079444 A) filed by Ajay Kumar Garg Engineering College on June 27, 2026, for System For Monsoon Dynamics Prediction Using Remote Sensing And Deep Learning And Working Method Thereof.
Inventors include Anuradha M Dhumale; Aastha Sharma; and Ankita Rani.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention discloses a system for monsoon dynamics prediction using remote sensing and deep learning. The system comprises a multi-source remote sensing data acquisition module (100) configured to acquire satellite, atmospheric sounding, oceanic, and reanalysis datasets; a spatiotemporal data pre-processing and feature engineering module (200) configured to perform data harmonization, noise removal, geospatial correction, and feature extraction; and a deep learning-based monsoon prediction and inference engine (300) configured to generate multi-horizon monsoon forecasts using hybrid neural architectures including temporal, spatial, and teleconnection-aware models. The system further comprises a physics-informed constraint and model validation module (400) for enforcing atmospheric conservation laws, an adaptive real-time data assimilation and model update module (500) for continuous forecast refinement, and a prediction output, visualization, and dissemination interface (600) for communicating probabilistic forecast outputs. The disclosed invention provides physically consistent, real-time, and high-accuracy monsoon prediction for operational meteorological applications. Accompanied Drawings [Fig. 1-4]
Disclaimer: Curated by HT Syndication.
