1 Introduction

Solar radiation is the fundamental energy driving the processes of evapotranspiration, photosynthesis, heating of the soil, and energy storage in vegetation. The surface radiation balance is one of the most important components to climate system. Changes in the radiation balance are closely related to the sensible and latent fluxes, plant growth, vegetation variation, wind circulation, and melting of snow. Knowledge about the radiation balance and its relation to environmental factors is therefore important for understanding the current climate system, as well as for predicting climate changes in the future ( Zhang et al., 2010; Shi and Liang, 2013).

Surface albedo is defined as the ratio of the reflected to the incoming shortwave (0.3~3 μm) solar radiation striking a surface. As one of the key land-surface parameters, albedo directly controls radiative energy distribution and hence determines the amount of energy available for heating the lower atmosphere; it affects evapotranspiration, photosynthesis, respiration, and potentially precipitation, etc. (Zhang et al., 2013; Meng et al., 2014). Also, surface albedo is a fundamental forcing parameter in general circulation models, climate models, numerical weather models, hydrology models, etc.; surface albedo plays an important role in controlling the Earth's radiation energy budget and determining earth surface temperature and evapotranspiration (Zheng et al., 2015). Small changes in surface albedo, even lower than the detection limits of existing satellite-derived products, can result in global temperature changes equivalent to that attributable to the anthropogenic increase of any enhanced greenhouse gas (Charlson et al., 2005). Regional modeling studies have documented that the positive forcing resulting from decreased surface albedo in boreal forests can be sufficiently large to offset the negative forcing expected from increased carbon sequestration by these forests (Betts, 2000). Therefore, accurate description of albedo is of great importance in land–atmosphere interaction and regional climate-modelling investigations.

The Tibetan Plateau (TP), known as the "world's third pole," has attracted great attention because of its sensitivity to global climate changes (Qiu, 2008; Yang et al., 2011; Yao et al., 2012). The surface radiation balance of the TP is important for characterization of the spatiotemporal variation in atmosphere–surface radiation and energy interactions, for analysis of variations in atmospheric conditions and land cover, and for providing evidence of the impacts of and responses to climate change over the TP (Hong and Kim, 2008; Shi and Liang, 2013). Hence, investigating the characteristics of the surface radiation balance in this area is important for the improvement of the land-surface process parameterization schemes, study on global energy and the water cycle, and global changes in the terrestrial ecosystems. In this paper, we report the results from one year of measurements of radiation fluxes and surface albedo over an alpine meadow and an alpine wetland in the eastern TP. The objectives of this work are to (1) quantify the radiation balance over two alpine underlying surfaces with the same climatic background; and (2) discuss the changes of the surface albedo with solar elevation angle, soil-water content, and weather conditions.

2 Materials and methods

The experiment was conducted at two sites at the Zoige Plateau Wetland Ecosystem Research Station (Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences). The alpine meadow site (33.92°N, 102.10°E, 3,440 m a.s.l.) and an alpine wetland site (33.93°N, 102.87°E, 3,434 m a.s.l.) were located in the eastern Tibetan Plateau (Figure 1) and characterized by a subfrigid humid climate. The annual mean temperature and rainfall was 1.7 °C and 600 mm at the alpine meadow site, and 1.1 °C and 650 mm at the alpine wetland site. In this study, we analyzed data collected in 2014. At the alpine meadow site, the observation sensors for this study included a four-way radiometer (CNR1; Kipp and Zonen, Netherlands) at a height of 2.0 m, and a volumetric soil-water content probe (CS616; Campbell Scientific, USA) at depth of 0.05 m. The signals from those sensors were averaged half-hourly by a data logger CR23XTD (Campbell Scientific, USA). The measurement variables and heights/depths at the alpine wetland site were same as at the alpine meadow site; but the four-way radiometer was NRO1 (Hukseflux, Netherlands), and the data logger was CR1000 (Campbell Scientific, USA).

The radiation balance in an ecosystem can be expressed as follows (Zhang et al., 2010):

${R_n} = (1 - \alpha ){R_s} - {L_n}$

(1)

where R _n is net radiation, R _s is incident solar radiation, α is ground surface albedo, and L _n is the net longwave radiation (effective terrestrial radiation), which is the difference between upward longwave radiation emitted from the surface (L _u ) and the downward longwave radiation from the atmosphere (L _d ).

To investigate the effect of cloudiness on the radiation balance, we used a clearness indexK _t to describe the changes in cloudiness (Gu et al., 1999):

${K_t} = \frac{{{R_s}}}{{{S_e}}}$

(2)

${S_e} = {S_{sc}}\left[ {1 + 0.33\cos \left(\frac{{360{t_d}}}{{365}}\right)} \right]\sin {\theta _h}$

(3)

$\sin {\theta _h} = \sin \varphi \cdot \sin \delta + \cos \varphi \cdot \cos \delta \cdot \cos \omega $

(4)

where S _e (W/m²) is the extraterrestrial irradiance at a plane parallel to the Earth's surface, S _sc is the solar constant (W/m²), t _d is the day of year, θ _h is the solar elevation angle, φ is degree of latitude, δ is the declination of the sun, and ω is the time angle. According to the previous work of Kudish and Ianetz (1996), K _t >0.65, 0.35<K _t <0.65, andK _t <0.35 represent the clear sky, cloudy sky and overcast sky, respectively.

3 Results and discussion 3.1 Seasonal variation of radiation fluxes

Figure 2a shows the annual variation of incoming solar radiation. The upper-boundary value represented clear-sky data, with the highest and lowest values corresponding to the beginning of June and the beginning of January, respectively. For both sites, the daily R _s on clear days changed in the scope of about 13 to 35 MJ/(m²·d), with the average of 21.12 MJ/(m²·d) for the alpine meadow and 19.69 MJ/(m²·d) for the alpine wetland. The annualR _s was 6,447 and 6,012 MJ/(m²·a) for the alpine meadow and the alpine wetland, respectively. Solar elevation and weather conditions are the most important factors that affectR _s (Zhang et al., 2010). In this paper, the difference of solar elevation between our two sites can be neglected due to their close proximity. Although our two sites belong to same climatic background, there was still a definite difference in weather conditions. The clearness index at the alpine meadow site was about 7% higher than that at the alpine wetland site; this value approximated the difference of annual R _s between the two sites. Thus, for two sites with close geographical proximity, the difference of R _s mainly resulted from the difference of cloudiness at the sites. The annual R _s values at our two sites were comparable to the results measured at Haibei in the northeastern TP (Zhang et al., 2010) but much lower than the value observed at Naqu in the northern TP (Ma et al., 2005). Compared with the values measured at the same latitude in Japan (Li et al., 2005), maximum daily R _s and annual R _s were much higher, this pattern being consistent with the result that solar radiation reaching the TP is about 1.2~2 times that of the same latitudes at low altitudes (Ye and Gao, 1979).

The downward longwave radiation (L _d ) and upward longwave radiation (L _u ) showed a similar temporal variation pattern (Figures 2b and 2c), with the maximum in the beginning of August and the end of December. The daily L _u was significantly higher than the daily L _d , with averages of 29.2 and 22.8 MJ/(m²·d) at the alpine meadow site and 29 and 22.5 MJ/(m²·d) at the alpine wetland site. In comparison withL _u and L _d , the annual variation of daily net longwave radiation L _n was relatively small (Figure 2d), with a range of about 0.5~13 MJ/(m²·d).L _d on cloudy days was higher than on clear days; inversely, L _n on cloudy days was much lower than on clear days, due to the effect of clouds. The annual L _n was 2,193 and 2,370 MJ/(m²·a) for the alpine meadow and the alpine wetland, respectively. The largeL _n over the alpine wetland site means that the differences between land-surface and air temperature at the alpine wetland site were also higher than at the alpine meadow site because the long-wave radiation was mainly determined by the temperature. The annualL _n values at our two sites were also comparable to the results measured at Haibei in the northeastern TP (Zhang et al., 2010) but much lower than the value observed at Naqu in the northern TP (Ma et al., 2005). Moreover, the energy loss on the TP from L _n was generally much higher than the reported global mean (1,511~1,734 MJ/(m²·a)) (Gupta et al., 1999; Wild et al., 2013), which is partly because of the low L _d on the TP (Zhang et al., 2010). The high L _n on the TP indicated that the TP always lost more energy through the exchange of longwave radiation compared with results on the global scale.

Higher net radiation (R _n ) was recorded in late July, and lower values were in late December (Figure 2e). The daily R _n was almost always positive throughout the year, with an annual R _n of 2,648 and 2,544 MJ/(m²·a) for the alpine meadow site and the alpine wetland site, respectively, close to the values (2,532~2,772 MJ/(m²·a)) measured over other alpine grasslands on the TP (Gu et al., 2005; Zhang et al., 2010; Ma et al., 2015; Shang et al., 2015; Wang et al., 2016).

3.2 Seasonal variation of radiation balance

Daily albedo α (the ratio of daily reflected to incident solar radiation) of the two surfaces had similar seasonal variation (Figure 3a). In winter and early spring, under snow-free conditions, α for both sites exhibited small seasonal variations; and there was no significant difference between the alpine meadow and the alpine wetland surfaces. In April, α for both sites gradually decreased as the frozen soil started to thaw. As the vegetation emerged, α increased toward a mid-summer peak and then declined in autumn as vegetation began to senesce. If we omit the snow-cover data (daily α>0.26), during the period when the open water of the wetland surface was frozen (December to March), the dailyα at the wetland surface was slightly higher than at the alpine meadow site, due to the high albedo of the ice surface (Lei et al., 2011). However, during the period with the open water of the wetland surface was unfrozen (April to December), the daily α at the wetland surface was significantly lower than at the alpine meadow site. In this study, the annual average albedo was 0.22 and 0.18 for the alpine meadow surface and the alpine wetland surface, respectively, comparable to values measured at Haibei in a seasonally frozen region of the TP (Zhang et al., 2010) but much lower than the values observed at Tanggula, Xidatan, and Wudaoliang in the permafrost region of TP (Yao et al., 2009; Xiao et al., 2010). Furthermore, the surface albedos on the TP were higher than the average albedo of the Earth and the average obtained for the Northern Hemisphere (Table 1).

Normalized effective radiation (L _n /R _s ) at our two sites showed a similar pattern of albedo (Figure 3b). When we omit the snow cover data (daily α>0.26), dailyL _n /R _s over the alpine meadow surface (the alpine wetland surface) showed the highest value of about 0.85 (0.87) in December, the lowest value of about 0.15 (0.13) in August. The annual average values were 0.38 and 0.42 for the alpine meadow surface and the alpine wetland surface, respectively. In other words, the alpine meadow and the alpine wetland emitted about 38% and 42%, respectively, of annual incident solar radiation back into atmosphere in the form of net longwave radiation. The reasons for the higher ratio ofL _n /R _s over the alpine wetland surface is mainly attributed to relatively low R _s and relatively large L _n . The annual average ratios of L _n /R _s on the TP were much higher than the global means, as well as for the average of Northern Hemisphere and other grasslands, indicating that energy loss due to the longwave radiation exchange between the TP surface and the atmosphere is more important than in other grassland ecosystems (Table 1).

Radiation efficiency (R _n /R _s ) displayed the opposite trend to L _n /R _s , with the lowest value in December and the highest value in August (Figure 3c). The annual average value of R _n /R _s was 0.40 for both of our sites, within the range of 0.30 to 0.44 measured for the TP (Ma et al., 2005; Zhang et al., 2010; Wang et al., 2012; Shang et al., 2015). Although the incident solar radiation on the TP is much higher than that of the same latitudes at low altitudes (Ye and Gao, 1979), the annual values of R _n /R _s on the TP were significantly lower than the global means, as well as the average of Northern Hemisphere and other grasslands. This finding suggests that the energy available in the alpine ecosystems for driving land-surface processes is not necessarily high due to high albedo and normalized effective terrestrial radiation.

3.3 Response of albedo to environmental conditions

The solar elevation angle (θ _h ) is a key environmental factor that influences the albedo and can be calculated from the longitude and latitude of a site, along with the Julian day and mean measurement time (Wang et al., 2005). To show accurately the influence of the solar elevation angle on surface albedo, the half-hourly radiation data associated with rain and snow days were removed. In addition, sharp peaks in surface albedo were also removed. The variation of surface albedo with the solar elevation angle is plotted in Figure 4. The albedo dramatically decreases with the increase of the solar elevation angle, especially when θ _h <30° at the alpine meadow site and when θ _h <40° at the alpine wetland site. The fitted equation form at our two sites can be expressed as follows:

$\alpha = a + b \cdot {{\rm{e}}^{ - c{\textit{θ} _h}}}$

(4)

which is consistent with the previous studies (Table 1) (Paltridge and Platt, 1981; Li and Hu, 2009; Neena et al., 2014; Zheng et al., 2014). Moreover, cloudiness is also an important factor that influences the albedo. As Figure 4 shows, under overcast sky conditions, the θ _h -α curve is significantly different from that under clear and cloudy sky conditions. When 10°<θ _h <30° for the alpine meadow site and 10°<θ _h <45° for the alpine wetland site,α under overcast sky conditions were lower than under clear and cloudy sky conditions. However, when 45°<θ _h for the alpine meadow site and 60°<θ _h for the alpine wetland site, under overcast sky conditions, α were higher than under clear and cloudy sky conditions. According to the above comparison, we can find that the coefficients of the relationship between θ _h and α not only depend on site characteristics but also are dramatically influenced by cloudiness (Table 2). Therefore, when one uses Equation (4) to parameterize the surface albedo, the coefficients must be determined from in situ observation; and the effects of cloudiness on albedo should be taken into account.

To show the influence of soil-water on surface albedo, the half-hourly albedo data that met the following criteria were adopted: (1) The solar elevation angle varying from 30° to 45° for the alpine meadow site and from 35° to 60° for the alpine wetland site were used to minimize the effects of solar elevation angle and cloudiness on surface albedo; and (2) the peak growing season (July and August) data were selected to eliminate the influence of vegetation change on surface albedo. The relationship between the surface albedo and soil-water content at the 5-cm depth is shown inFigure 5. The surface albedo decreases with increase in the soil-water content, consistent with the results showing the linear or exponential relations between surface albedo and soil-water content (Wang et al., 2005; Liu et al., 2008a; Guan et al., 2009; Li and Hu, 2009; Neena et al., 2014; Zheng et al., 2014); but the fitting parameters are significantly different from those in previous studies (Table 3). Although there are small differences between the two fitted curves for our two sites (Figure 5), the fit coefficients (R², Table 3) indicated that the optimal fitting formula was a linear equation and an exponential equation for the alpine meadow site and the alpine wetland site, respectively. Therefore, it is clear that the dependence of surface albedo on soil-water content is site-specific.

4 Conclusions

We examined the radiation balance and the response of albedo to environmental factors above an alpine meadow and an alpine wetland site on the Tibetan Plateau for the year 2014. The annual incident solar radiation was 6,447 and 6,012 MJ/(m²·a) for the alpine meadow and the alpine wetland, respectively. The ratios of net longwave radiation to incident solar radiation at our two alpine ecosystems were much higher than that for the global surface and the Northern Hemisphere, indicating that energy loss due to the longwave radiation exchange between the surface and the atmosphere is more important than in other ecosystems. Annual net radiation was 2,648 and 2,544 MJ/(m²·a) for the alpine meadow site and the alpine wetland site, respectively, and accounts for only about 40% of incident solar radiation. The annual average albedo was 0.22 and 0.18 for the alpine meadow surface and the alpine wetland surface, respectively. On a 30-min scale, surface albedo exponentially decreases with the increase of the solar elevation angle; and it tends to be a constant when the solar elevation angle is larger than 30° at the alpine meadow site and 40° for the alpine wetland site. And the relationship between surface albedo and the solar elevation angle for our two sites is significantly influenced by cloudiness. When minimizing the effects of the solar elevation angle, cloudiness, and vegetation change on surface albedo, surface albedo is found to decrease linearly with the increase of soil-moisture for our two sites.

Acknowledgments:

This work was jointly supported by the National Natural Science Foundation of China (Grant Nos. 91537106, 41405016, 41275016, 91537104, and 41605011). We would like to acknowledge the Zoige Plateau Wetland Ecosystem Research Station for providing data support for this study.