• Zaid Ahmad
• Amarendra Singh
• Saumya Singh
• Prabhat Kumar Patel

The aim of this research is to compare the accuracy of predictive models using Aerosol Optical Depth (AOD) for the cities of Delhi and Lucknow. We applied both Long Short-Term Memory (LSTM) and Auto Regressive Integrated Moving Average (ARIMA) models for AOD prediction for January 2023 using past 10 years data (i.e., 01-01-2013 to 01-01-2023) in both locations in order to accurately analyse their performances. It is observed that ARIMA model (in both cases) performs better than LSTM model and have better accuracy in predicting. The ARIMA models exhibited lower Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) values (i.e., Lucknow - MSE: 0.01843, RMSE: 0.1357; Delhi - MSE: 0.03672, RMSE: 0.1916) along with lower Mean Absolute Percentage Error (MAPE) values (i.e., Lucknow - 18.9975%; Delhi - 20.80004%). Furthermore, the ARIMA models showed higher R-Square values (i.e., R2: Delhi - 0.5994, Lucknow - 0.7335;) compared to the LSTM models (i.e., R2: Lucknow - 0.5452, Delhi - 0.5590). The study gives valuable insights into the spatial variability of AOD and underscores the significance of employing ARIMA models for time series prediction in regions characterized by distinct AOD patterns. To evaluate air quality as well as mitigate environmental effects, it is essential to comprehend these aerosol patterns.

Aerosol optical depth; LSTM; ARIMA; Lucknow; Delhi.

"Comparative Analysis Of LSTM And ARIMA Models For Predicting Aerosol Optical Depth In Lucknow And Delhi Cities", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 9, page no.f629-f640, September-2023, Available :http://www.jetir.org/papers/JETIR2309576.pdf

"Comparative Analysis Of LSTM And ARIMA Models For Predicting Aerosol Optical Depth In Lucknow And Delhi Cities", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 9, page no. ppf629-f640, September-2023, Available at : http://www.jetir.org/papers/JETIR2309576.pdf