Long term (2007–2013) observations of columnar aerosol optical properties and retrieved size distributions over Anantapur, India using multi wavelength solar radiometer

Abstract This paper presents the long – term observational studies on aerosol optical properties measured at Sri Krishnadevaraya University (SKU) campus (14° 62′ N, 77° 65′ E, 331m asl), Anantapur, in southern India during 2007–2013 using a ground based Multi – Wavelength solar Radiometer (MWR). Seasonal mean values of Aerosol Optical Depth (AOD) for the whole study period were observed to be 0.34 ± 0.03, 0.45 ± 0.04, 0.24 ± 0.04, and 0.31 ± 0.03 during the winter, summer, monsoon and post – monsoon, respectively. Annual mean values of Angstrom exponent (α) (turbidity coefficient (β)) varied from 0.68 ± 0.25 (0.18 ± 0.03) to 1.1 ± 0.12 (0.32 ± 0.12) during 2007–2013. However, high values of α in the range of 0.8–1.1 were observed during the winter, while low values in the range 0.3–0.7 were noticed during the monsoon. The frequency distribution of AODs during winter in the range of 0.2–0.4 is about 78%, while summer these were shifted from 0.3 to 0.6 around ∼70%, which indicated the dust strongly affects this region. The accumulated frequencies of Angstrom exponent (α) less than 1.0 and greater than 1.0 were about 69% and 31%, respectively, occurred in the summer months indicates the dominance of coarse particles. Columnar size distributions, retrieved from the spectral optical depths, in general, show a bimodal log normal distribution in the optically active size range. The seasonal mean effective radius (R eff ) was found to be high in monsoon (∼0.61 μm) and low in winter (∼0.38 μm). The highest mean mass loading values are lies between 475 ± 36 to 769 ± 49 mg m −2 during the summer, whereas the lowest value in the range 204 ± 19 to 278 ± 23 mg m −2 during the monsoon. The difference between α and curvature effect have been studied as a function of AOD on seasonal basis. Finally, to understand the contribution of long range transported aerosols, we have investigated this analysis making use of back trajectories obtained from the HYSPLIT model for different seasons.