A Hyderabad resident may soon receive advance warning about pollution spikes before stepping outside. A study published in 'Scientific Reports' has developed a data-driven model that forecasts urban air pollution levels using five years of air quality and weather data, potentially serving as an early warning system for cities grappling with deteriorating air quality.
Study Details and Methodology
The study, 'Deciphering Seasonal Dynamics and Time Series Forecasting of Urban Air Quality: A Case Study in Hyderabad, Vijayawada and Visakhapatnam', was published in May. Conducted by Shreyas D, Nishith B, N Neelima, TV Smitha, and Vivek Venugopal from Amrita Vishwa Vidyapeetham, along with Tolga Ozer from Afyon Kocatepe University in Turkiye, the research compared two machine-learning approaches to forecast the Air Quality Index (AQI).
The direct forecasting method employed a deep feedforward neural network with residual blocks to predict the final AQI in one step. The indirect method first predicted individual pollutants—including PM2.5, PM10, and NO2—using models such as Random Forest, then calculated the AQI from those estimates.
Data Sources and Key Findings
Researchers used five years of continuous data from January 1, 2020, to December 31, 2024, obtained from Central Pollution Control Board regulatory-grade monitoring stations. The study covered Hyderabad in Telangana and Vijayawada and Visakhapatnam in Andhra Pradesh. For Hyderabad, data was recorded at the Bollaram Industrial Area station.
The dataset included pollutants such as PM2.5, PM10, NO, NO2, NOx, CO, SO2, ozone, NH3, benzene, toluene, and xylene, along with meteorological variables like temperature, humidity, visibility, wind speed, and wind direction. Continuous data from 2025 tested the models' real-world accuracy.
Both modelling approaches achieved R2 values exceeding 0.96 for all three cities, meaning the models explained over 96% of the variation in air quality data. For Hyderabad, the direct FNN model achieved an R2 of 0.97 and a Root Mean Square Error of 11.25. In one validation sample, the model predicted an AQI of 118.45 for January 4, 2025, against the actual recorded value of 125.50—an error of just 5.6%.
Seasonal Patterns and City Comparisons
The study found that particulate matter, especially PM2.5 and PM10, was the primary driver of AQI across the three cities. A clear seasonal pattern emerged: pollution levels peaked in winter, particularly December and January, and in post-monsoon months, while air quality improved during the monsoon (July to September) due to atmospheric cleansing.
Among the three cities, Visakhapatnam recorded the highest average AQI at about 112, followed by Hyderabad at 94 and Vijayawada at 78. Visakhapatnam's average pollution was about 19% higher than Hyderabad's and 43% higher than Vijayawada's. For Hyderabad, winter pollution spikes were attributed to vehicular emissions, road dust, and open burning, worsened by meteorological conditions trapping pollutants. As a sprawling industrial and technology hub, the city's air quality is affected by a combination of emissions and local weather.
Potential Applications
Such forecasting systems could be used for public health early warnings, air-quality advisories, and regulatory responses. The models could help officials activate measures like the graded response action plan, manage traffic or industrial activity before pollution levels worsen, and use long-term data for sustainable urban planning. The indirect forecasting method could also help policymakers identify the specific pollutant driving a spike, enabling targeted action against sources such as diesel vehicles or industrial plants.
