Introduction
Air pollution is a rising concern worldwide, and it isn’t easy to estimate because it varies by location, depending on historical air quality data, weather, terrain, and anthropogenic factors. Short-term, continuous monitoring of air quality data is required at local and regional scales, dependent on population density, to issue health alerts when levels of hazardous air pollutants surpass specific limits.
The Air Quality Index (AQI) method measures the concentrations of specific pollutants about national criteria hourly and daily. AQIs are widely used and successfully monitor short-term (hours/days forward) air quality on simple scales. The use of APIs allows for effective hazard warnings in poor air quality.
Why is historical air quality data required?
When the surrounding air includes gases, dust, fumes, or scents in such quantities to are detrimental to humans and animals or damage plants and materials, air pollution occurs.
One of the worst killers of our day is air pollution. Polluted air killed 6.4 million people globally in 2015, with 2.8 million deaths caused by poor indoor air quality and 4.2 million fatalities caused by ambient (outside) air pollution. 2,3 According to data from that year, air pollution caused:
- 19% of all stroke cases
- 24% of deaths from ischemic heart disease
- 21% of deaths from stroke
- 23% of deaths from lung cancer
Furthermore, air pollution appears to be a major risk factor for children’s neurodevelopmental disorders5 and adult neurodegenerative diseases. Forecasting technology is becoming increasingly crucial with the economic, environmental, and human costs of air pollution.
The advantages of historical air quality data and forecasting
Individual, community, national, and global air pollution forecasting are valuable investments. Accurate forecasting allows people to plan, reducing negative health impacts and costs.
People will be more aware of variations in the air quality, the impact of pollutants on health, concentrations likely to cause bad effects, and pollution-control measures if they are aware of these issues. Furthermore, because individuals demand air quality information, there is a higher possibility of inspiring improvements in individual behavior and governmental policy7. 8
Such understanding does have the potential to result in healthier people and a cleaner environment. Governments also use early forecasting to develop processes for reducing the severity of localized pollution levels.
Forecasting air quality with accuracy
There are several elements to consider when determining air quality, many of which are highly unpredictable. Beijing’s authorities, for example, have been known to shut down coal plants and companies and prohibit a section of the city’s millions of registered automobiles from operating. Local weather conditions and pollution emissions are highly linked to air pollution levels. Long-range pollution transfer, such as through high winds, is an important influencing element that must be considered when projecting local AQI readings.
Predicting air quality thus necessitates not just the problems of weather forecasting but also data and knowledge of:
- Local pollutant levels and emissions
- Pollutant concentration and emissions from remote sites
- Pollutant movements and probable transformations
- Because of the many variables involved in estimating air quality, air pollution forecasting is internal and external.
Techniques for predicting air quality
There are numerous forecast models of this type, which are more complex than weather forecasting techniques. These models are computer simulations of how contaminants in the air disperse.
Forecasting the weather
A good weather forecast is the first step toward an effective air quality forecast. The three-primary meteorological (weather) forecasting types are climatology, statistical approaches, and three-dimensional (3-D) models are the three-primary meteorological (weather) forecasting types.
Climatology
The idea behind climatology is that history is a good predictor of the future. This technique is one-dimensional because it is dependent on the association between specific meteorological conditions and pollution levels. This method is frequently extended to include weather and pollution pattern matching. This method has several limitations, and it is best used as a supplement to other forecasting methods.
Statistical methods
Statistical approaches can quantify the link between air quality with weather patterns.
- CART (classification and regression tree) is designed to categorize data into dissimilar categories. Variables that correlate to ambient pollution levels are identified using the software. The data is utilized to forecast concentrations based on weather circumstances and connected pollutant concentrations.
- Regression analysis determines how variables are related. Associations between pollution levels and meteorological data variables are discovered by analyzing historical data sets. As a result, an equation that may be used to predict future pollution levels has been created.
- Adaptive learning & pattern recognition techniques are used in artificial neural networks. Computer-based algorithms are created to mimic the human brain’s pattern recognition abilities. Due to the multi-approach, this is debatably the best tool for forecasting pollution. The preceding statistical methods have the disadvantage of assuming stability by organizational procedures that affect air quality. As a result, any significant changes in emissions or climate (brief or long term) will significantly reduce the accuracy of these methods. There are more advanced solutions that attempt to adjust for these flaws. Three-dimensional models are what they’re called.
Three-dimensional (3-D) models
Three-dimensional simulations Represent all of the main mechanisms that affect ambient air pollution levels mathematically. Three-dimensional models use numerous submodels to simulate the emission, transportation, and modification of air pollution, including:
- Emissions model: This model simulates the geographical extent of pollutants from natural and human sources.
- Meteorological model: Using a 3-D meteorological model plus emissions data, builds a trajectory model to anticipate ambient pollution levels.
- Chemical model: Determines the pollutant’s consequence by examining the change of primary (emitted) pollution into secondary pollution.
Pollution forecasting systems are fast improving, and their accuracy will continue to improve. Historical air quality data predictions that are accurate and easily accessible assist increase public awareness and allow vulnerable groups to plan and provide authorities with data for public health alerts. This is an intriguing new field with a bright future for academics and scientists.
References
https://www.iqair.com/us/blog/air-quality/can-air-pollution-be-predicted
https://www.hindawi.com/journals/complexity/2020/8049504/
https://www.frontiersin.org/articles/10.3389/fdata.2022.822573/full
https://www.iqair.com/us/blog/air-quality/can-air-pollution-be-predicted
https://journalofbigdata.springeropen.com/articles/10.1186/s40537-021-00548-1