A synergy of linear model and wavelet analysis towards space-time characterization of aerosol optical depth (AOD) during pre-monsoon season (2007-2016) over Indian sub-continent

The pre-monsoon aerosol concentration plays a significant role in modifying precipitation amount over the In-dian sub-continent. A large variety of aerosol from different sources produces a complex radiative and climate response through the interaction with the hydrometeorological parameters. In this study, we analyzed the spacetime dynamics of aerosol optical depth (AOD) in relation to the meteorological and surface parameters over Indian sub-continent during pre-monsoon season from 2007 to 2016. The level-3 daily aerosol products from Moderate Resolution Imaging Spectroradiometer (MODIS) and Multi-angle Imaging Spectroradiometer (MISR) were used. The IMD gridded rainfall and temperature, ECMWF derived RH and wind velocity, and GLADS derived soil moisture data were also used at daily time scale. For the space-time model, the Mann-Kendall trend test and a pixel-based multiple linear regression were used, while, wavelet transformation was used on these daily observations for analyzing periodicities of AOD. The time-series average shows moderate to high AOD (0.4-0.8), including a consistent positive anomaly in the Indo-Gangetic basin (IGP) in north India. A significant inter-annual variation is also observed both in MODIS and MISR datasets. The trend statistics shows an increasing trend of aerosol concentration in the eastern and southern India. The linear regression shows a variable response of AOD with changing magnitude of meteorological factors. However, a substantial spatial coverage of significantly decreased AOD is observed with increasing soil moisture content (beta <-0.04). The wavelet analysis manifests the abundances of 32-128 days of cycle with a periodic interjection of 8-32 days of cycle suggesting occurrence of fine and coarse mode aerosols events, respectively. The coherency of time series AOD and other covariates shows varying leading and lagging dynamics in these two principal periodicities. The findings, however, evidenced a notable difference in the space-time patterns of AOD in MODIS and MISR datasets. The analyzed AOD cycles are coincided with the Madden-Julian-Oscillation (MJO) that recurs every 30-60 days interval. The findings also support the theoretical proposition of elevated heat pump theory (EHP) driven by fine mode aerosols for occurring pre-monsoon and monsoon precipitation over the Indian sub-continent. The analyzed periodicity of AOD can provide useful insights in studying the short/long term variability of precipitation over polluted environments during the pre-monsoon season.