Modelling and Analysis of Aerosol and Cloud-Precipitable Water Inter-Hemispheric Interactions of Aerosol-Satellite Data Using Ground Observation

Abstract

This study investigates the seasonal and spatiotemporal variations of aerosol optical depth (AOD) retrieval algorithm with cloud fraction (CF) and precipitable water (PW) from Ozone Monitoring Instrument (OMI/Aura) as satellite-based and Aeronet Robotic Network (AERONET) as ground-based stations for the period from 2008 to 2019 and optimized from 2020 to 2025 using neural network model. The exact interaction of AOD-CF-PW remains a good area of scientific interest but has remained poorly represented. However, the idea of the statistical approach of data interpolation and extrapolation to extract AOD-CF-PW data gaps using inter-hemispheric approach of the Northern hemisphere (Ilorin 8.484 N, 4.675 E and Saada 31.626 N, 8.156 W) and Southern hemisphere (ICIPE-Mbita 0.432 S, 34.206 E and Skukuza 24.992 S, 31.587 E) of the African continent. An estimation of relative bias error values of both AODs-CF-PW data was validated by correlating the results and performing the standard deviation (SD) analysis. To conclude the interaction error in AOD-CF-PW values, root mean square error (RMSE), relative bias (RD), absolute bias (AD) and mean absolute error (MAE) were performed. Furthermore, to provide knowledge of major aerosols contributors over both hemispheres, an estimation of jet winds to see the zonal and meridional impact through 7 days kinematic back trajectories at various initial pressures was performed. Winter season of Skukuza presents correlation results of OMI-PW (r = 0.125), CF (r = − 0.041) and AERONET-derived-PW (r = − 0.075) and CF (r = − 0.010) with corresponding analysis of (%EE = 22.33), (RMSE = 0.017), (RB = 0.287), (MAE = 0.002), (SD = 0.285), (AD = 0.243). Interestingly, the Skukuza of the southern hemisphere during the month of winter shows the least value of (RMSE = 0.017), indicating good agreement of OMI and AERONET AODs. This result indicated that OMI satellite remote sensing does not give an accountable result of AOD-CF-PW interactions but depends more on geographic terrain.