IdeaSmart Surveillance of Air Quality

Company Idea and Objectives: It is very far-reaching to effectively predict air quality index (AQI). Predicting air pollution has been a major area of research in several different fields, ranging from environmental science to computer science and statistics. Deterministic models in environmental science attempt to understand the behavior of air pollutants at the molecular level, simulating diffusion and diffusion patterns according to the size and type of molecules. In computer science and statistics, linear machine learning models have been used in more data-driven approaches, notably using multiple linear regression and autoregressive moving averages. Air pollution caused by the presence in the atmosphere of substances that are harmful to the health of humans and other living things or that cause damage to the climate or materials. Research has identified many different types of air pollutants, such as gases (including ammonia, carbon monoxide, sulfur dioxide, nitrous oxide, methane, carbon dioxide, and chlorofluorocarbons), particulates (organic and inorganic), and biomolecules. In statistics, linear regression is a type of regression analysis that models the relationship between one or more independent variables and a dependent variable using the least squares function known as the linear regression equation. This function is a linear combination of one or more model parameters called regression coefficients (all independent variables are powers of one). The case of only one independent variable is called simple regression, and the case of more than one independent variable is called multiple regression. Expected Impact of Your Idea on Sustainable Development The project will serve for the sustainable development goal SDG 13 Climate Action Air pollution is caused by the presence in the atmosphere of substances that are harmful to the health of humans and other living things or that cause damage to the climate or materials. Air pollution is an important risk factor for many pollution related diseases, including respiratory tract infection, heart disease, chronic obstructive pulmonary disease, stroke and lung cancer. More and more evidence shows that exposure to air pollution may be related to lower IQ scores, increased risk of mental diseases such as cognitive impairment, depression, and harmful perinatal health. The impact of poor air quality on human health is far-reaching, but it mainly affects the human respiratory system and cardiovascular system. An air quality index (AQI) is used by government agencies to communicate to the public how polluted the air currently is or how polluted it is forecast to become. Different countries have their own air quality indices, corresponding to different national air quality standards. Some of these are the Air Quality Health Index (Canada), the Air Pollution Index (Malaysia), and the Pollutant Standards Index (Singapore). The United States: The United States Environmental Protection Agency (EPA) has developed an Air Quality Index that is used to report air quality. This AQI is divided into six categories indicating increasing levels of health concern. An AQI value over 300 represents hazardous air quality and below 50 the air quality is good. European: In November 2017, the European Environment Agency announced the European Air Quality Index (EAQI) and started encouraging its use on websites and for other ways of informing the public about air quality. My profile: I am Zixin An from Sierra Canyon High School CA, USA . I have been exploring the ways to monitor the impacts of different air pollutants on air quality indicators. Pollutants emitted into the atmosphere by human activity include: • Particulate matter (PM2.5 and PM10) > Particulate matter is a mix of solids and liquids, including carbon, complex organic chemicals, sulphates, nitrates, mineral dust, and water suspended in the air. The most damaging particles are the smaller particles, known as PM10 and PM2.5. • Nitrogen Oxides (NO, NO2, NOx) > Nitrogen oxides are a group of seven gases and compounds composed of nitrogen and oxygen, sometimes collectively known as NOx gases. The two most common and hazardous oxides of nitrogen are nitric oxide(NO) and nitrogen dioxide(NO2) • Sulphur Dioxide(SO2) > Sulfur dioxide, or SO2 is a colourless gas with a strong odor, similar to a just-struck match. It is formed when fuel containing sulfur, such as coal and oil, is burned, creating air pollution. • Carbon Monoxide(CO) > Carbon monoxide is a colourless, highly poisonous gas. Under pressure, it becomes a liquid. It is produced by burning gasoline, natural gas, charcoal, wood, and other fuels. • Benzene, Toluene and Xylene (BTX) > Benzene, toluene, xylene, and formaldehyde are well-known indoor air pollutants, especially after house decoration. They are also common pollutants in the working places of the plastic industry, chemical industry, and leather industry • Ammonia( NH3) > Ammonia pollution is pollution by the chemical ammonia (NH3) – a compound of nitrogen and hydrogen which is a by-product of agriculture and industry. • Ozone(O3) > Ground-level ozone is a colourless and highly irritating gas that forms just above the earth's surface. It is called a "secondary" pollutant because it is produced when two primary pollutants react in sunlight and stagnant air. These two primary pollutants are nitrogen oxides (NOx) and volatile organic compounds (VOCs). III. Plans for Sustainability and Implementation Predicting air pollution has been a major area of research in several different fields, ranging from environmental science to computer science and statistics. Deterministic models in environmental science attempt to understand the behavior of air pollutants at the molecular level, simulating diffusion and diffusion patterns according to the size and type of molecules. In computer science and statistics, linear machine learning models have been used in more data-driven approaches, notably using multiple linear regression and autoregressive moving averages. Nevertheless, the limitation of linearity hampers the accuracy of predictions because many pollutants behave nonlinearly. Therefore support vector regression is proposed. However, these methods mainly predict the pollution at the current time step, not the future air pollution. For predicting future air pollution, using data over longer time spans from previous years to look for pollution patterns will improve accuracy. This project will use deep learning and machine learning methods to first collect air quality data sets, clean the data according to the main indicators of air pollution, select relevant features as the input of the model, and then use the AQI standard as the real output of the model. Then build an air quality prediction model, and choose the model that makes the air quality prediction better by comparing the models. A.Datasets In the currently public datasets, the year span is relatively large, and the air quality datasets of Delhi and Beijing in India are the ones with relatively complete information. The Delhi dataset contains hourly and daily air quality data and AQI at various sites across multiple cities in India. The data is organized into 16 columns, with two additional columns added by manual calculations, and each row contains details of pollutant levels recorded on a daily basis for a different city from 2015 to 2020. B.Data Preprocess Using the pandas data package to view the data, it is found that the total dimension of India's air data is: 29531*16. The time span is from 2015 to 2020, listed as air pollutant data captured by sensors at various city sites, we can find that there are many null values in the data, which will greatly interfere with subsequent modeling, so the first step is to clean the data.

What is my expertise ? My solid foundation at statistic computing and logic modelling will allow me to execute accurate and data-based inference and decision-making when designing the idea My adept skills at linear machine learning models, particularly multiple linear regression and autoregressive moving averages will help me build feasible relation formulas to predict different air conditions My excellent perception and techniques at data gathering and analysis will have my research and conclusion complete How can I support on other organizations ? The research result of our paper will help the environmental protection authorities and organizations in monitoring the air quality My accurate and data-driven research tools will help the climate administrations and weather forecast institutions to supervise air pollutant index My research data will serve as important references to facilitate the professional investigation offices to conduct related surveys