Multi-view Based Smog Data Analysis Methods

With the rapid economic development,air pollution is becoming a severe environmental problem in developing countries such as China and India.To better understand the problems of air pollution,the first of all is to monitor air quality.PM2.5 is tiny particulate matter less than 2.5 micrometers in size,they can penetrate deep into the lungs,and measuring them is considered a more accurate reflection of air quality than other methods.But this ground-based measurements are sparsely and unevenly distributed in space.Satellite observations offer valuable global information about PM2.5 concentrations,but have limited accuracy and completeness.Aiming at the problems above and focusing on multi-view method,in contrast to satellite domain-driven methods for PM2.5 retrieval,our approach is satellite data-driven.Challenges and our proposed solutions discussed here in context of global scale PM2.5 estimation include:First PM2.5 regression from Aerosol Optical Depth(AOD)data;Second training such a multi-view model for robust performance across multiple satellite measures and the model for incomplete data avoids direct imputation of the missing element;Finally smog data is automatic acquisition by using Python technology and integration of air quality monitoring system.Results on synthetic and real-world aerosol data comprising many satellite-borne sensors indicate the benefits of the proposed approach.