Long-Term Variability of Radiative Flux

Since 1950s, significant decadal changes in Rs were observed worldwide. Continuous observations of Rs at different locations around the world revealed reduction in surface solar radiation with time (Ohmura et al. 1989; Russak 1990;

Gilgen et al. 1998; Liepert 2002; Stanhill and Cohen 2001; Abakumova et al. 1996; Wild et al. 2005). This decreasing trend in surface solar radiation is known as global/solar dimming (Stanhill and Cohen 2001). After 1980s, a partial recovery has been observed at some stations around the globe (Wild et al. 2005; Pinker et al. 2005; Wild et al. 2007; Ruckstuhl et al. 2008; Philipona et al. 2009). Thereby, the term so-called brightening was coined. The updates beyond 2000 provide a less coherent picture as compared to the preceding dimming and brightening (Wild et al. 2009; Wild et al. 2012). After 2000, continuation of brightening was observed in Europe, USA, and some parts of Asia, leveling off at sites in Japan and Antarctica, and renewed dimming in China. These changes are mainly attributed to changes in aerosols, clouds, and their interactions (Wild et al. 2009). Anthropogenic aerosols are found to be important contributors to these changes (Streets et al. 2006). High-quality surface radiation measurements over seven different stations in USA for the period 1996-2011 showed increasing trend in shortwave radiation. This brightening was attributed to a decrease in cloud coverage; aerosols had only a minor effect (Augustin and Datton 2013). On the other hand, Tibetan Plateau, the cleanest region in the world, experienced a transition from brightening to dimming around the end of 1970s (Tang et al. 2011). Over the last three decades, solar radiation as well as total cloud cover over the Tibetan Plateau showed a decreasing trend. Also the extinction due to aerosol loading is one order lower in magnitude than the observed dimming. It is concluded that dimming over the Tibetan Plateau is mainly due to increase in the amount of water vapor and deep cloud cover but not due to aerosol loading (Yang et al. 2012).

Over India, aerosols and clouds are found to be the two important factors that lead to large variability in the solar radiation reaching at the surface. Aerosol concentrations over the Indian subcontinent show large seasonal variability and consistent increase. Long-term aerosol optical depth (AOD) variability observed using a network of multi-wavelength radiometers showed an increase of 2-9 % year-1 (Moorthy et al. 1993,2001). Based on satellite data, AOD trends are increasing over all the major cities in India, with higher optical depths in the northern part of India as compared to southern part of India (Sarkar et al. 2006; Gautam et al. 2007). From surface observations, atmospheric turbidity is also found to be increasing significantly at different stations in India (Soni and Kannan 2003; Soni et al. 2012). There are many studies showing region specific short-term variations of radiative fluxes and aerosol radiative forcing. Simultaneous measurements of aerosol properties and radiative fluxes over an urban station, Pune, during the dry seasons of 2001 and 2002 showed a radiative forcing of -33, 0, and 33 W m-2 at the surface, top of the atmosphere, and the atmosphere, respectively (Pandithurai et al. 2004). Radiative forcing efficiency of-97 W m-2 was shown over Bengaluru (Babu et al. 2002). Over New Delhi, high aerosol radiative forcing during pre-monsoon is attributed to dust aerosols (Singh et al. 2005; Pandithurai et al. 2008), while during winter radiative forcing is influenced by haze and fog conditions (Ganguly et al. 2006). A recent study over Ranchi for the period February 2011-January 2012 showed an average reduction of 9 % in Rs due to aerosols (Latha et al. 2014). India is one of the few regions around the world, which showed decreasing trend in Rs (Ramanathan et al. 2005; Padmakumari et al. 2007;

Soni et al. 2012) under all-sky conditions, where aerosols and clouds together contributed to the annual trend. Over the Indian subcontinent, studies on long-term surface energy budget are lacking, due to lack of continuous long-term observational data.

Padmakumari et al. (2007) studied the long-term variability in Rs for the period 1981-2004 over 12 different stations in the Indian subcontinent. They found a decreasing trend in Rs over all the stations ranging from 0.17 to 1.44 W m-2 year-1, and the average dimming is observed to be * 0.86 W m-2 year-1. The stations in the north showed strong decline as compared to the stations in the southern parts of India. Decadal variability showed a strong decline during the decade 1991-2000 as compared to 1981-1990. Soni et al. (2012) also studied the long-term variability in Rs as well as sunshine duration over 12 stations for the period 1971-2005. They observed decreasing trend in Rs over all the stations varying from 0.3 to 9.0 W m-2 per decade and sunshine duration also showed decreasing trend in correspondence with Rs.

Clouds display a wide variability on annual as well as on seasonal scales. Based on reanalysis data, earlier studies have shown increasing trend in cloud optical depths over the Indian subcontinent (Rajeevan et al. 2000; Badrinath et al. 2010). Cloud type information also helps to understand the variability in surface radiation budget. It was found that low and deep clouds produce more cooling effect in shortwave band, while warming effect in long-wave band, resulting in net cooling effect at the surface (Balachandran and Rajeevan 2007). They also observed that high clouds contribute more to shortwave radiative forcing over oceans, while both middle and high clouds over land.

To examine the possible reasons of decreasing trend in Rs, Padmakumari and Goswami (2010) examined the trends in Rs for clear-sky and cloudy-sky conditions. They studied the long-term Rs data from 1981 to 2006 over widely distributed 12 different stations in India. The study brought out new insights into the origin of solar dimming over India. The rate of dimming is found to be twice (12 W m-2 decade-1) as large during cloudy conditions as compared to the clear-sky conditions (6 W m-2 decade 1)(Fig. 2). Under cloudy conditions the changes in Rs is expected as clouds

Fig. 2 Annual mean surface reaching radiative flux, averaged over 12 stations which are spatially distributed over the Indian subcontinent, under all-sky, clear-sky and cloudy-sky conditions. Straight lines represent linear trend lines. All the trends are statistically significant at the 99 % confidence level (Padmakumari and Goswami 2010)

are the strongest modulators of Rs. However,it is not very clear why it should show a strong decreasing trend. Figure 2 depicts that as compared to aerosols, clouds play a larger role in the observed solar dimming. To understand this fact, seasonal variation (MJJASO, NDJFMA) of Rs under clear and cloudy conditions is studied (Fig. 3).In both the seasons the clear-sky dimming is attributed to direct effect of aerosols (Fig. 3a). Under cloudy conditions, strong decreasing trend is observed during both the seasons (Fig. 3b). To understand the cloud effect on Rs, seasonal mean outgoing long-wave radiation (OLR) over the Indian region is considered (Fig. 3c). As OLR is linked to cloud top temperatures, it is used as a proxy to identify the intensity of cloud development. During the MJJASO, the decreasing trend of OLR represents increasing cloud amount, while during NDJFMA the increasing trend represents increasing shallower clouds. Also, during MJJASO a significant decreasing trend of OLR (<220 W/m2) represents that deep clouds are increasing and the area covered by deep clouds is also found increasing (Fig. 3d) as compared to all other types of clouds. Thus, under cloudy conditions, solar dimming during the summer season may be due to increasing convective clouds covering a larger area, while during the winter season, it may be due to the aerosol indirect effect.

Seasonal variation of surface reaching radiative flux

Fig. 3 Seasonal variation of surface reaching radiative flux (Rs) and OLR over India for two seasons MJJASO (May-October) and NDJFMA (November-April). a Rs under clear-sky conditions, b Rs under cloudy conditions, and c total OLR. d Seasonal variation of total number of grids covered by deep clouds (OLR < 220 W/m2) over India (Padmakumari and Goswami 2010)

 
Source
< Prev   CONTENTS   Source   Next >