Utilizing an Artificial Neural Network with Limited Meteorological Data in AL-Fallujah City to Estimate AOD550 with MODIS AODs

Ahmed Yousif Ismael *, Abdel-Radi Abdel-Rahman Abdel-Qader and Abdul Rahim Bashir Hamed Bashir

College of education Al-Hasahisa, University of Gezira, Sudan.
 
Research Article
International Journal of Science and Research Archive, 2024, 12(02), 2198–2204.
Article DOI: 10.30574/ijsra.2024.12.2.1477
Publication history: 
Received on 07 July 2024; revised on 17 August 2024; accepted on 19 August 2024
 
Abstract: 
The amount of solar radiation that is absorbed by particles in the atmosphere is measured by the Aerosol Optical Depth (AOD550). Determining the precise role of aerosols requires an understanding of the fluctuations in worldwide AOD550. A neural network (ANN) model was created in AL-Fallujah, Iraq, with the purpose of training and estimating daily (AOD550). The statistical parameters Root Mean Square Error (RMSE), which are dependent on the correlation coefficient (R), and standard division (SD) were established to assess the produced ANN-models. For both the hidden and output layers, the activation functions (Segmiont, Gaussean, Hyperbolic Tangent, Hyperbolic Secant), hidden layers (1, 2, 3, and 4), and changes ranging from 10000 to 60000 with a 10000 interval were used. The created artificial neural network models were examined based on statistical criteria that were computed. ANN model (25) is found to be the best among all the models that were studied. Their corresponding (R) and (RMSE) values for the estimate phases were 0.939, 0.962, and 0.055, 4.3. These results show the generalization ability and full efficacy of the ANN model in pollution assessments.
 
Keywords: 
AOD550; Aerosols; Particulate Matter; ANN; AL-Fallujah; Iraq
 
Full text article in PDF: 
 
Copyright information: 

Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0