Using physical parameters in a deep neural network for regional VTEC prediction

Key Points

• VTEC prediction improved with deep neural networks using physical parameters, highlighting effective modeling.
• Key findings show a notable accuracy boost of 20% in predictions when utilizing specific physical parameters.
• Analysis employs deep neural networks designed specifically for regional VTEC prediction with an emphasis on physical inputs.
• These findings suggest potential for enhanced forecasting methods in meteorological applications, emphasizing model adaptability.