Executive Summary : | Air quality level is closely associated with our day-to-day life due to its serious negative impact on human health. Therefore, air pollution prediction is a significant step in air pollution management for citizen's protection of any country. The information on spatio-temporal air pollution concentration is vital in the growing urban conditions. In recent years, much research has been conducted to perform air quality modeling using statistical methods, machine learning, and deep learning techniques. Prime research has focused on estimation of temporal variations of air pollutions at different scales. However, the spatial variation of air pollution at a high temporal resolution/granularity is an evolving problem in the recent years. Apart from the conventional problems associated with air pollution, it has a great impact on the rainfall and temperature, especially in the pre-monsoon season (March, April, and May) over India, particularly the eastern coastal states. Pre-monsoon season is unique for heat waves and localized short-spell rainfall occurrence in many parts of India. It is expected that in the near future, the eastern parts of India (e.g., Bhubaneswar, Visakhapatnam, Amaravathi, etc) will be experiencing more urbanization. This would really gearup of as per the state government policy of Andhra Pradesh to bifurcate many towns and make more districts. Increased urbanization and population growth results in high conditions of air pollution, which can have strong influence on weather and climate extremes such as rainfall and temperature. Atmospheric temperature and rainfall extremes have shown to increase (Bingheng et al. 2008). Under SAFAR, IITM with its collaborating agencies has been providing pollution details over different parts of the country like Pune, Mumbai, Delhi and Ahmedabad. The CPCB warns Andhra Pradesh, as five of its cities namely Visakhapatnam, Rajahmundry, Vijayawada, Tirumala and Kurnool crossed the levels of Air Quality Index (AQI). This project is aimed to explore the air pollution at high spatio-temporal resolutions and its subsequent impact on meteorological parameters such as rainfall and temperature for the various cities in Andhra Pradesh. Further, to the best of our knowledge, no safe route navigation systems have been available in Andhra Pradesh for public transport, which will be one of the objectives of this proposal. This proposal can provide a scope to compare the deep learning technique with the available 3D atmospheric model outputs. The analysis will be carried out for pre-monsoon season. |