Abstract:Urban air pollution poses a critical issue, affecting health and economies. Cities worldwide strive to enhance air quality through long‐term monitoring and modeling efforts. However, the immediate need for actionable data has led to an increased reliance on short‐term statistical approaches. This study demonstrates the ARIMA search grid method's capacity to predict Budapest's air quality, revealing diverse accuracy among different stations. These insights are vital for assessing forecast reliability and informing policies to address urban air pollution effectively. Air pollution poses a persistent challenge for urban management departments and policymakers due to its significant health and economic impacts. Various cities worldwide have implemented diverse strategies and initiatives to enhance air quality monitoring and modeling standards. However, the outcomes of these efforts often manifest over the long term, leading to a preference for short‐term statistical methods. The autoregressive integrated moving average (ARIMA) search grid modeling approach has gained widespread use for forecasting air quality. This paper presents a comprehensive time series analysis conducted to predict air quality in urban areas of Budapest, Hungary, with a focus on nitrogen dioxide (NO2) and particulate matter (PM10), using air quality data spanning from 2018 to 2022 for four monitoring categories: Urban traffic, industrial background, urban background, and suburban background. The study employs the ARIMA search grid method to forecast concentrations of these pollutants at multiple air quality monitoring stations based on Akaike information criteria (AIC) and the Bayesian information criteria (BIC) criteria along with the results of augmented Dickey–Fuller (ADF) test. The results demonstrate varying levels of forecast accuracy across different stations, indicating the model's effectiveness in short‐term predicting of air quality. These findings are essential for assessing the reliability of air quality forecasts in Budapest and can inform decisions regarding air quality management and the development of strategies to address air pollution and particulate matter concerns in the region.

Data-driven Real-time Short-term Prediction of Air Quality: Comparison of ES, ARIMA, and LSTM

A comparative analysis of LSTM and ARIMA for enhanced real-time air pollutant levels forecasting using sensor fusion with ground station data

Evaluation of Time Series Forecasting Models for Estimation of PM2.5 Levels in Air

Time Series Analysis Based Air Pollution Level Prediction Using Long Short Term Memory Recurrent Neural Network with Attention Mechanism

Online learning of ARIMA for time series prediction.(2016)

Deep learning based multimodal urban air quality prediction and traffic analytics

Air quality index prediction using seasonal autoregressive integrated moving average transductive long short‐term memory

Short-term air quality prediction based on EMD-transformer-BiLSTM

A Deep Learning Approach for Forecasting Air Pollution in South Korea Using LSTM

A novel Encoder-Decoder model based on read-first LSTM for air pollutant prediction

Air Quality Forecasting with Hybrid LSTM and Extended Stationary Wavelet Transform

Short‐term predictions of PM10 and NO2 concentrations in urban environments based on ARIMA search grid modeling

The Prediction of Influenza-like Illness and Respiratory Disease Using LSTM and ARIMA

Time Series Forecasting for Air Pollution in Seoul

Air-Quality Prediction Based on the EMD–IPSO–LSTM Combination Model

Machine Learning Algorithms for High-Resolution Prediction of Spatiotemporal Distribution of Air Pollution from Meteorological and Soil Parameters

A Long Short-Term Memory Approach to Predicting Air Quality Based on Social Media Data

Advancing air quality forecasting in Abu Dhabi, UAE using time series models

A Deep Spatial-Temporal Ensemble Model for Air Quality Prediction

REAL TIME AIR QUALITY PREDICTION AND ANOMALY DETECTION

A new method for prediction of air pollution based on intelligent computation