ORIGINAL PAPER

From brightening to dimming in sunshine durationover the eastern and central Tibetan Plateau (1961–2005)

Qinglong You & Shichang Kang &

Wolfgang-Albert Flügel & Arturo Sanchez-Lorenzo &

Yuping Yan & Jie Huang & Javier Martin-Vide

Received: 27 February 2009 /Accepted: 27 October 2009 /Published online: 19 November 2009# Springer-Verlag 2009

Abstract The Tibetan Plateau (TP) with an averageelevation of over 4,000 m asl is the highest and mostextensive highland in the world. We used monthly meansunshine duration from the Chinese Meteorological Ad-ministration to examine the spatial and temporal variabilityof sunshine duration at 71 stations with elevations above2,000 m asl in the eastern and central TP during the 1961–2005 period. The temporal evolution of the mean annualsunshine duration series shows a significant increase from1961 to 1982 at a rate of 49.8 h/decade, followed by adecrease from 1983 to 2005 at a rate of −65.1 h/decade,with an overall significant decrease at a rate of −20.6 h/decade during the whole 1961–2005 period, which ismainly due to the summer and spring seasons. Thisconfirms the evidence that sunshine duration in the TP

ranges from brightening to dimming in accordance withsunshine duration trends in the rest of China. The surfacesolar radiation downwards from ERA-40 reanalysis data inthe same region confirms the brightening/dimming phe-nomenon shown by the sunshine duration before/after the1980s. Otherwise, additional climatic variables such as lowcloud amount, total cloud amount, precipitation, relativehumidity and water vapor pressure, in most cases, exhibitsignificant negative correlation with sunshine duration in theTP on an annual and seasonal basis before and after 1982,respectively. The trends of these variables suggest thatchanges in some of them might be related to the brighteningand dimming detected with the use of sunshine durationmeasurements over the TP. We also hypothesize that theimpact of anthropogenic aerosols upon the climatic variables

Q. You : S. Kang (*) : J. HuangLaboratory of Tibetan Environment Changes and Land SurfaceProcesses, Institute of Tibetan Plateau Research,Chinese Academy of Sciences (CAS),18 Shuangqing Road,Beijing 100085, Chinae-mail: shichang.kang@itpcas.ac.cn

Q. You :W.-A. FlügelDepartment of Geoinformatics, Friedrich-Schiller University Jena,Löbdergraben 32,Jena 07743, Germany

S. KangState Key Laboratory of Cryospheric Science, CAS,320 Donggang West Road,Lanzhou 730000, China

A. Sanchez-LorenzoClimate Research Laboratory, Barcelona Science Park,C/Baldiri Reixac, No. 2, second floor,Barcelona 08028, Spain

Y. YanNational Climate Center,46 South street Zhongguancun Road,Beijing 100081, China

Q. You : J. HuangGraduate University of Chinese Academy of Sciences,19 Yuquan Road,Beijing 100049, China

A. Sanchez-Lorenzo : J. Martin-VideGroup of Climatology, University of Barcelona,4a Planta C/Montalegre, No. 6,Barcelona 08001, Spain

Theor Appl Climatol (2010) 101:445–457DOI 10.1007/s00704-009-0231-9

analyzed cannot be rejected, especially in the significantincrease in low cloud cover since approximately 1980.

1 Introduction

Global mean surface temperatures have risen by 0.74°C±0.18°C when estimated by a linear trend over the last100 years (1906–2005). The warming rate over the last50 years is almost double that recorded over the last century(IPCC 2007). Furthermore, the energy balance at theEarth’s surface determines the evolution of surface temper-ature and the water cycle, key parameters of climatechange, and ecosystems (e.g., IPCC 2007). Ohmura andLang (1989) found that solar radiation striking the Earth’ssurface had declined by more than 10% during the threeprevious decades. A decrease in downward solar radiation(a phenomenon known as global dimming (e.g., Stanhilland Cohen 2001)) of about 4% or about 7 W m−2 wasobserved from 1961 to 1990 at land stations around theworld (Liepert 2002). Since the late 1980s, a reversal in thistrend has been detected in many regions of the world andconsequently this dimming did not persist into the 1990s.Thus, a widespread brightening has been observed since thelate 1980s (Wild et al. 2005). However, recent studies(Alpert and Kishcha 2008; Alpert et al. 2005) suggest thatthe dimming is not global, but is rather confined only tolarge urban areas and consequently should be considered asa local or regional phenomenon.

On the other hand, at the regional scale, other studieshave detected trends in surface solar radiation or sunshineduration (a proxy measure of global solar radiation) overthe USA (Liepert 2002), different areas of western and

central Europe (Power 2003; Norris and Wild 2007;Sanchez-Lorenzo et al. 2007; 2008), in China (e.g., Cheet al. 2005, 2007; Liang and Xia 2005; Kaiser and Qian2002; Qian et al. 2006, 2007; Shi et al. 2008; Zhang et al.2004; Zheng et al. 2008) and India (Kumari et al. 2007),among others.

The Tibetan Plateau (TP), with an average elevation ofover 4,000 m asl and an area of approximately 2.5×106

km2, is the highest and most extensive highland in theworld. It has been called the “Ridge of the world” andpresents China’s highest solar radiation and sunshineduration mean values (Yeh and Gao 1979). Previous studies(Chen et al. 2003; Duan and Wu 2006; Liu and Chen 2000;You et al. 2008a) have shown significant warming in the TPduring the last half century, in phase with the global trendsresulting from increasing anthropogenic greenhouse gasemissions. Global warming has already had evidentenvironmental consequences in the TP such as acceleratedglacier melt, and thickening of the active layer inpermafrost areas (Zhang 2007).

Variations in solar radiation at the Earth’s surfaceprofoundly affect the human and terrestrial environment.The change in global solar radiation in the TP is sensitive toenvironmental evolution, and systematic analysis of long-term trends for the TP is still lacking due to the few solarradiation measurements available. It is therefore importantto analyze the temporal variability of sunshine durationseries and to examine its causes. The objective of this studyis to investigate the annual and seasonal (winter: DJF;spring: MAM; summer: JJA; autumn: SON) sunshineduration variations in the eastern and central TP during1961–2005. The European Centre for Medium-RangeWeather Forecasts (ECMWF) reanalysis (ERA-40 here-

Fig. 1 Distribution of 71stations (dot) and 56 ERA-40reanalysis grid points (cross inrectangle) in the eastern andcentral Tibetan Plateau and itsadjacent territories

446 Q. You et al.

Fig. 2 Regional annual and seasonal sunshine duration in the eastern and central TP during the 1961–2005 period. The smoother line is the 9-year smoothing average

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 447

after; ECMWF 2002) have been widely used in climatestudies. Therefore we also investigate the variability of surfacesolar radiation downwards in monthly ERA-40 reanalysisdata. Low cloud amount, total cloud amount, water vaporpressure, precipitation and relative humidity trends and therelationship between these and sunshine duration have alsobeen analyzed. The effect of anthropogenic aerosols onsunshine duration are also discussed.

2 Data and methods

Monthly mean sunshine duration and additional variablesfrom 71 stations (Fig. 1) were provided by the NationalClimate Center and the Chinese Meteorological Adminis-tration (CMA). There are 156 stations in the original dataset. Of these, 124 stations have maintained daily data since1961. Of the 124 stations, 38 stations were excluded owingto their location at an elevation below 2,000 m asl, then 12stations more were also excluded owing to problems in dataquantity (10 stations stopped operation during the 1980s–1990s and 2 stations showed abnormity due to thediscontinuity of data records). The distribution of theresulting stations is uneven and very sparse in the westernTP, which may influence regional trends. Therefore threestations distributed in the western TP were also excluded.The remaining 71 stations with elevations above 2,000 masl are distributed in the eastern and central TP. The periodof 1961–2004 was selected for analysis. Detailed stationselection procedures are described in recent papers by thepresent authors (You et al. 2008a, b).

The monthly mean surface solar-radiation-downwardsERA-40 reanalysis data were obtained from the ECMWFwebsite (http://www.ecmwf.int/). ERA-40 surface solar-radiation downwards is available from September 1957to August 2002 with a spatial resolution of 2.5°×2.5°(ECMWF 2002). The period of 1961–2001 was selectedfor the present study. The slightly shorter period forERA-40, due to lack of more current data, has limitedinfluence on our results. To compare with surfacestations, we selected grid points contained within therectangle outlined in Fig. 1, thus covering the samespatial domain.

The overall linear trends of the series were calculatedwith Sen’s estimator of the slope (Sen 1968), and theirsignificance was estimated by means of the Mann-Kendallnonparametric test. Both methods are robust to outliers andhave been widely used to compute trends in hydrologicaland meteorological studies. Parametric correlations betweensunshine duration and the additional climate variables arealso performed. In this paper, linear trends and correlationsare considered to be statistically significant when signifi-cant at the 5% level.

3 Results

3.1 Time evolution and trends of sunshine duration during1961–2005

Figure 2 shows the mean annual and seasonal time evolutionof sunshine duration in the eastern and central TP during the1961–2005 period, together with the 9-year smoothingaverage. The annual and seasonal mean trends along withthe number of stations with increasing/decreasing trends insunshine duration in the region are described in Table 1. On anannual basis, the mean TP sunshine duration series exhibits astatistically significant decreasing trend with a rate of−20.6 h/decade, and 27/5 stations have negative/positivesignificant linear trends, respectively. Most of the stationspresenting an increase are situated in the central TP (notshown), which is consistent with the detected decrease inprecipitation (You et al. 2008a). At seasonal level, thehighest rate of decrease in sunshine duration is found insummer (−7.4 h/decade) although the trend is not significant.Nonetheless, most of the stations (49) show a decrease, 19 ofthem statistically significant. The rates for spring, autumnand winter are −5.7, −3.4, −3.1 h/decade, respectively, andmore than 60% of stations show decreasing trends in theseseasons. Indeed, the linear trends detected are significant forthe spring and autumn mean TP sunshine duration series.

Figure 2 clearly shows that sunshine duration in the TPincreased in the first half of the study period and decreasedin the second half. We therefore subdivided the studyperiod into two phases: 1961–1982 and 1983–2005, and theannual and seasonal mean trends as well as the number ofincrease/decrease trends in sunshine duration are also listedin Table 1. A significant increasing trend of 49.8 h/decadeis found for sunshine duration during the 1961–1982 period

Table 1 Annual and seasonal mean trends along with the number ofincreasing/decreasing values in sunshine duration in the eastern andcentral TP for three periods: 1961–1982, 1983–2005 and 1961–2005

Annual Spring Summer Autumn Winter

1961–2005 −20.6a −5.7a −7.4 −3.4a −3.1Increase 23(5) 22(4) 22(2) 31(4) 26(7)

Decrease 48(27) 49(21) 49(19) 40(18) 45(17)

1961–1982 49.8a 14.4a 17.3 3.5 0.7

Increase 53(24) 47(14) 49(17) 45(7) 37(8)

Decrease 18(4) 24(2) 22(0) 26(2) 34(5)

1983–2005 −65.1a −10.5a −32.7a −10.4 −11.6Increase 16(5) 28(3) 9(1) 21(2) 28(2)

Decrease 55(37) 43(13) 62(24) 50(14) 43(12)

a The trends with a significance level greater than 95%

The number of stations (P < 0.05) is shown in parentheses. Unit is h/decade

448 Q. You et al.

Fig. 3 Regional annual and seasonal surface solar-radiation downwards in the eastern and central TP during the 1961–2001 period. The smootherline is the 9-year smoothing average

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 449

at the annual scale, with 53/18 stations presenting increas-ing/decreasing trends. All seasons show the same increas-ing trends, the clearest ones occurring in summer (17.3 h/decade) and spring (14.4 h/decade), although only the latterone is significant. The dramatic downward trend occursduring 1983–2005 with a rate of −65.1 h/decade, and 55/16stations show decreasing/increasing trends. The greatestand significant decreasing trends occur in summer (−32.7 h/decade) and spring (−10.5 h/decade).

3.2 Annual and seasonal time evolution of surface solarradiation downwards from ERA-40

Figure 3 shows the regional annual and seasonal surfacesolar-radiation downwards from ERA-40 reanalysis data inthe eastern and central TP during the 1961–2001 period. Onan annual basis, the surface solar-radiation downwardsdramatically decreased/increased before/after 1970, andthen shows a constant decreasing trend after the 1980suntil the end of the analyzed period, with the majority(about 70%) of grids having negative trend magnitudes. Inmost cases, the correlation coefficients between surfacesolar-radiation downwards at an annual scale and the sameat a seasonal scale are over 0.7, indicating that thevariability of surface solar-radiation downwards on aseasonal basis is similar to that at a annual scale. Figure 4shows the relationship between surface solar radiation fromERA-40 reanalysis data and sunshine duration from surfacestations on an annual basis in the eastern and central TPduring the 1961–2001 period, together with the correlationcoefficients and statistical significance. It is clear that thereis a significant positive correlation (R=0.43, P=0.005)

between surface solar-radiation downwards and sunshineduration in the TP during 1961–2001. We also calculatedthe correlation coefficients for the periods of 1961–1982and 1983–2001 and found that correlation coefficients havelarger values during 1983–2001 than those during 1961–1982 (Table 2). There is a slight positive correlation (R=0.31) between surface solar-radiation downwards andsunshine duration in the TP during 1961–1982, while ithas a significant positive correlation after 1982 (R=0.79).On the seasonal basis, surface solar-radiation downwardshas a significant positive correlation with sunshine durationduring 1961–1982, 1983–2001 and 1961–2001, and corre-lation coefficients have larger values in summer andautumn. With the exception of spring, correlation coef-ficients reach 0.8 during the 1983–2001 period. Therefore,the surface solar-radiation downwards from ERA-40 canidentify the brightening as shown in sunshine durationbefore the 1980s, and strengthen the dimming as indicatedby the sunshine duration decline after the 1980s.

3.3 Annual and seasonal time evolution and trendsof related meteorological elements

Table 3 shows correlation coefficients between sunshineduration and low cloud amount, total cloud amount, watervapor pressure, precipitation and relative humidity on anannual and seasonal basis in the eastern and central TP forthree periods: 1961–1982, 1983–2005 and 1961–2005. Forthe 1961–2005 period, sunshine duration in the region hasnegative correlation with low cloud amount, total cloudamount, water vapor pressure, precipitation and relativehumidity, which all passed the significant level with theexception of total cloud amount (Fig. 5), with closecorrelations occurring in summer and autumn. Noting thetwo subdivided phases, it is clear that the correlations passingthe significant level during 1983–2005 are higher than duringthe 1961–1982 sub-period, and the most significant correla-tions occur in summer and autumn, in accordance with theresults obtained during the 1961–2005 period.

Fig. 4 Relationship between surface solar-radiation downwards fromERA-40 reanalysis data and sunshine duration from surface stationson an annual basis in the eastern and central TP during the 1961–2001period. R stands for correlation coefficients and P for statisticalsignificance

Table 2 Correlation coefficients between sunshine duration andsurface solar-radiation downwards on the annual and seasonal basisin the eastern and central TP for three periods: 1961–1982, 1983–2005and 1961–2001

Annual Spring Summer Autumn Winter

1961–2005 0.43a 0.32 0.65b 0.58a 0.46a

1961–1982 0.31 0.36 0.48a 0.56b 0.29

1983–2001 0.79a 0.28 0.88b 0.88b 0.79a

a The correlation coefficients with a significance level greater than95%b The highest correlation coefficients

450 Q. You et al.

In order to examine the factors contributing to sunshineduration trends over the eastern and central TP, the annualvariations in low cloud amount, total cloud amount, watervapor pressure, precipitation and relative humidity in theeastern and central TP during 1961–2005 are shown in Fig. 6.The annual and seasonal trends during 1961–1982, 1983–2005 and 1961–2005 are presented in Table 4.

On an annual basis, there are significant increasing trendsin water vapor pressure and precipitation, and a significantdecrease in total cloud amount during the 1961–2005 period(Fig. 6). Equally, all the variables, except total cloud cover,have an overall decrease and increase during the 1961–1982and 1983–2005 sub-periods, respectively—the latter increasebeing more pronounced than the decreasing trends during thefirst part of the period analyzed. On a seasonal basis, in mostcases, the time evolution in each season is similar to that of theannual series, and, likewise, the 1983–2005 trends are morepronounced than during the 1961–1982 sub-period, whichresulted in the increasing/decreasing trends from 1961 to2005. The variations in spring and summer mostly contributeto the annual series. Taking precipitation as an example, wefound that the significant increasing trend in precipitation is6.66 mm/decade, mainly due to increases in spring, with adecreasing trend in the 1960s and slight increases since the1970s. Approximately 73% of stations (not shown), present-ing increasing trends in precipitation, are mostly located onthe southern TP and in the north of the Qinghai Province,whereas 27% of the stations presenting decreasing trends arelocated on the central TP (You et al. 2008a). Otherwise, weobserved a significant increase in water vapor pressure on anannual and seasonal basis during the whole 1961–2005period, mainly resulting from the sharp increase detectedfrom the 1980s to the present. It is also evident that, in most

of the series, variability and trends exhibited a clear shift inlate 1970s or at the beginning of the 1980s.

Although the increasing precipitation and water vaporpressure over the eastern and central TP are in agreementwith the detected decrease in sunshine duration, it isinteresting to note that an abrupt decrease/increase intotal/low cloud cover has been reported since the mid-1970s. This inconsistency between trends of sunshineduration and total cloud amount, both variables showing adecrease over the TP rather than the expected opposite signin their trends, are in accordance with the results shown forthe rest of China. Compared with the total cloud amount,low cloud amount has the same variability as the sunshineduration, which can be attributed to the increased aerosolconcentrations in the atmosphere (Zheng et al. 2008). Thisindicates that sunshine duration is more sensitive to lowcloud amount than total cloud amount.

4 Discussion and conclusions

The present paper analyzes the spatial and temporal variabilityof sunshine duration based on 71 stations with an elevationabove 2,000 m asl in the eastern and central TP during the1961–2005 period. We found that sunshine duration has asignificant increasing trend on an annual and seasonal basisduring the 1961–1982 period, mostly due to the pronouncedincrease taking place in summer. After the 1980s, sunshineduration dramatically decreases at annual and seasonal levels.Thus, sunshine duration in the TP changes from brightening todimming during the 1961–2005 period. The surface solar-radiation downwards from ERA-40 reanalysis also dramati-cally increased before the 1980s, and then shows a constant

Annual Spring Summer Autumn Winter

Low cloud amount 1961–2005 −0.76a −0.68a −0.92b −0.84a −0.84a

1961–1982 −0.59a −0.65a −0.84a −0.82a −0.84b

1983–2005 −0.91a −0.67a −0.96b −0.87a −0.89a

Total cloud amount 1961–2005 −0.07 −0.17 −0.72b −0.61a −0.62a

1961–1982 −0.53a −0.61a −0.85a −0.90b −0.86a

1983–2005 −0.43a −0.18 −0.91b −0.78a −0.68a

Water vapor pressure 1961–2005 −0.65b −0.43a −0.63a −0.54a −0.36a

1961–1982 −0.26 −0.03 −0.30 −0.41 −0.091983–2005 −0.71a −0.46a −0.76b −0.57a −0.49a

Precipitation 1961–2005 −0.51a −0.59a −0.78b −0.48a −0.57a

1961–1982 −0.24 −0.45a −0.76b −0.49a −0.381983–2005 −0.62a −0.56a −0.80b −0.48a −0.71a

Relative humidity 1961–2005 −0.61a −0.45a −0.72a −0.79b −0.52a

1961–1982 −0.51a −0.47a −0.77a −0.89b −0.50a

1983–2005 −0.64a −0.18 −0.78a −0.81b −0.54a

Table 3 Correlationcoefficients between sunshineduration and low cloud amount,total cloud amount, water vaporpressure, precipitation andrelative humidity on the annualand seasonal basis in the easternand central TP for three periods:1961–1982, 1983–2005 and1961–2005

a The correlation coefficientswith a significance level greaterthan 95%b The highest correlationcoefficients

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 451

Fig. 5 Relationship between sunshine duration and low cloud amount, total cloud amount, water vapor pressure, precipitation and relative humidity onan annual basis in the eastern and central TP during the 1961–2005 period. R stands for correlation coefficients and P for statistical significance

452 Q. You et al.

Fig. 6 Regional annual low cloud amount, total cloud amount, water vapor pressure, precipitation and relative humidity in the eastern and centralTP during the 1961–2005 period. The smoother line is the 9-year smoothing average

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 453

decreasing trend after the 1980s until the end of the analyzedperiod. In most cases, the correlations between surface solar-radiation downwards and sunshine duration are very high. Thesurface solar radiation downwards from ERA-40 confirms thebrightening/dimming before/after the 1980s.

This temporal behavior in sunshine duration over the TPis opposite to the results found for most areas of the worldbased on solar radiation or sunshine duration measurements(Stanhill and Cohen 2001; Liepert 2002; Wild et al. 2005;Norris and Wild 2007 Sanchez-Lorenzo et al. 2008; amongothers), which show a decrease from the 1960s to the early1980s (global dimming), followed by a positive trend up tothe end of the twentieth century (global brightening). Themost accepted reason at present for this dimming/brighten-ing phenomenon is a change in the transmissivity of theEarth’s atmosphere due to an increase/decrease in world-wide anthropogenic aerosol concentrations. Even so, thetransition from brightening to dimming sunshine durationin the TP is in accordance with previous studies in EastChina (Kaiser and Qian 2002; Qian et al. 2006; Zheng et al.2008) and India (Kumari et al. 2007), which accommodatesmost of the world’s population and has been home to aboom in economic development since the 1980s, with theresulting regional increase in air pollution. For east China,aerosol loading has been increasing significantly during thepast decades (Luo et al. 2001), and the increased anthro-pogenic aerosol loading is believed to have resulted in lesssunshine duration (Kaiser and Qian 2002; Qian et al. 2006).

Aerosol loading plays an important role in the variationof sunshine duration in the TP. Previous studies have foundthat an increase in aerosol particles is expected to enhancethe scattering and absorption of incoming solar radiationand thus impair visibility (Liang and Xia 2005). Goodvisibility anomalies in western China (35–50°N, 75–95°E,

including our study region; Liang and Xia 2005; Xia et al.2006) during 1961–2000 indicates that the occurrence ofgood visibility significantly decreased and leveled offbefore/after the mid-1980s, indicating that aerosol loadingincreased during 1961–2000.

In order to check the differences of atmosphericcirculation and its possible relations with climaticvariable trends over the TP, we show the annual mean500 hPa latitudinal wind, longitudinal wind and wind-field differences between 1983–2005 and 1961–1982(Fig. 7). NCEP/NCAR reanalysis data are used in thisstudy (Kalnay et al. 1996). It is clear that the latitudinal/longitudinal wind is decreasing/increasing, indicatingthat both the easterly flow from east China andsoutherly flow from South Asia have strengthened. Thisis consistent with the differences of the wind fieldbetween 1983–2005 and 1961–1982 and is the result ofrapid increase in the emissions of aerosol in easternChina and South Asia (Luo et al. 2001). Consequently,these intriguing results may hint at the possibility of animpact of anthropogenic aerosols emissions on thedynamics of the atmospheric circulation and should beexplored further.

We also checked the China Compendium of Statisticsregarding, for example, industrial waste gas and sootemissions and the number of civilian vehicles in theTibetan Plateau. According to the China Compendiumof Statistics 1949–2004 compiled by Department ofComprehensive Statistics of the National Bureau ofStatistics, concerning Autonomous Tibet, the numberof civilian vehicles owned was 2,711 in 1961; 13,543 in1982; and surged up to 96,657 in 2003; the industrialwaste gas and soot emissions were 6.3 (100 million m3)and 0.8 (10,000 tons) in 1990, respectively, while they

Annual Spring Summer Autumn Winter

Low cloud amount (%/decade) 1961–2005 0.03 0.04a 0.05 0.02 −0.011961–1982 −0.05 −0.06 −0.12 −0.02 0.01

1983–2005 0.19a 0.10 0.41a 0.23a 0.16

Total cloud amount (%/decade) 1961–2005 −0.09a −0.08a −0.07 −0.09a −0.10a

1961–1982 −0.03 0.03 −0.11 −0.03 0.04

1983–2005 −0.04 −0.16a 0.10 0.01 −0.04Water vapor pressure (hPa/decade) 1961–2005 0.10a 0.09a 0.10a 0.08a 0.10a

1961–1982 0 −0.01 −0.13 −0.02 0.08a

1983–2005 0.20a 0.16a 0.34a 0.22 0.16a

Precipitation (mm/decade) 1961–2005 6.66a 4.82a 0.98 0.60 0.98a

1961–1982 −2.32 5.29 −13.27 5.34 2.21

1983–2005 12.52 6.86a 8.16 −2.98 0.83

Relative humidity (%/decade) 1961−2005 0.22 0.35 −0.18 −0.06 0.87a

1961−1982 −0.29 0.00 −0.95a −0.43 0.92

1983–2005 0.79 0.56 0.56 0.78 1.89a

Table 4 Trends of low cloudamount, total cloud amount,water vapor pressure,precipitation and relativehumidity on an annual andseasonal basis in the eastern andcentral TP for three periods:1961–1982, 1983–2005 and1961–2005

a The trends with a significancelevel greater than 95%

454 Q. You et al.

increased to 134 (100 million m3) and 1.58 (10,000 tons),respectively by 2004. In the TP, industrial and agricul-tural activities, population and energy consumption haverapidly increased since 1961, especially after the 1980s.This has led to a rapid increase in the emissions ofaerosol, although the speed of development is lowerthan that in eastern China, indicating that increase ofaerosol loading could account for, at least, part ofvariation of sunshine duration in the TP.

Although we speculate that anthropogenic aerosolsinfluence variability in sunshine duration over the TP, we

do not rule out other possible causes. In order to examinewhat factors influence the variability in sunshine durationin the region, we also analyze regional trends of low cloudamount, total cloud amount, water vapor pressure, precip-itation and relative humidity, and their correlations withsunshine duration on an annual and seasonal basis. It isclear that low cloud amount, total cloud amount, watervapor pressure, precipitation and relative humidity pre-sented significant negative trends during the 1961–1982sub-period, while exhibiting significant positive trends,with the exception of total cloud amount, during the

Fig. 7 Annual mean 500 hPa latitudinal wind (a), longitudinal wind (b) and wind filed differences (c) between 1983–2005 and 1961–1982. Theunit is m/s. Data are derived from NCEP/NCAR reanalysis data (available from http://www.cdc.noaa.gov)

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 455

1983–2005 sub-period. These results are consistent with thedecreasing trend of diurnal temperature range in the TP(You et al. 2008a), which present negative correlations withsunshine duration before and after 1982, with most of thecorrelation coefficients having passed the significant level.Sunshine duration is normally closely related to changes incloud amount, and increasing/decreasing cloud amountwould be expected to result in decreasing/increasingsunlight to the ground. In our study, sunshine duration hasa significant positive/non-significant correlation with low/total cloud amount on an annual basis. Consequently, thedecrease in sunshine duration over the TP cannot beexplained solely by changes in total cloud cover, whichhave shown a negative significant trend since the 1980s,and might be related to the increase detected in the lowcloud amount, precipitation and water vapor pressure series.Therefore, the variability in related climatic variables canbe partly linked to the sunshine duration trends, althoughwe cannot rule out the possible impact of anthropogenicaerosols, particularly in relation to the decrease in totalcloud cover. It is difficult to use observations to quantita-tively detect the change in cloud cover induced byincreased aerosols, because the knowledge of cloud typesand physical/radiative properties is not available. It hasbeen suggested that the increase in pollution in Chinaduring the last few decades might be related to the decreasedetected in cloud cover by means of an inhibition of cloudformation resulting from increased use of anthropogenicaerosols (Qian et al. 2006).

Solar dimming may have profoundly counterbalancedthe increase in downward longwave energy from theenhanced greenhouse effect, and may have been effectivein masking greenhouse warming (Wild et al. 2007). The TPunderwent significant warming during the 1961–2005period, and presents a dramatic increase, particularly sincethe 1980s (Chen et al. 2003; Duan and Wu 2006; Liu andChen 2000; You et al. 2008a), while the reversal frombrightening to dimming occured at the start of the 1980s.A regional climate model study also confirms thatenhanced climatic warming over the TP is due to thedoubling of carbon dioxide (Chen et al. 2003). Thus, thewarming detected in the TP might be underestimated, asdimming from the early 1980s to the present may be partlymasking the full dimension of the warming effect causedby greenhouse gases. Equally, the detected upward trendsin water vapor pressure or relative humidity are inagreement with the trends observed over the global oceansand some land areas in the Northern Hemisphere since themid-1970s (IPCC 2007), which are closely associated withthe increase in surface temperature. Likewise, atmosphericbrown clouds also contribute to warming trends in Asia(Ramanathan et al. 2007) and might have an impact in theTP if transported by atmospheric circulation.

Acknowledgements This study is supported by the National NaturalScience Foundation of China (40771187, 40830743), the NationalBasic Research Program of China (2005CB422004), Chinese Acad-emy of Sciences (KZCX2-YW-145), and the Sixth FrameworkProgram Priority (036952). The authors thank the National ClimateCenter and the China Meteorological Administration, for providingthe meteorological data for this study. Qinglong You is supported bythe “DAAD-CAS Joint Scholarship Program” and funded by the CASSpecial Grant for Postgraduate Research, Innovation and Practice.Arturo Sanchez-Lorenzo and Javier Martin-Vide are supported by theSpanish Ministry of Science and Innovation (MICINN) projectsIPIBEX (CGL2005-07664-C02-01) and NUCLIEREX (CGL2007-62664/CLI).

References

Alpert P, Kishcha P (2008) Quantification of the effect of urbanizationon solar dimming. Geophys Res Lett 35, L08801. doi:10.1029/2007GL033012

Alpert P et al (2005) Global dimming or local dimming? Effect ofurbanization on sunlight availability. Geophys Res Lett 32,L17802. doi:10.1029/2005GL023320

Che HZ et al (2005) Analysis of 40 years of solar radiation datafrom China, 1961–2000. Geophys Res Lett 32(6):L06803.1–L06803.5

Che HZ et al (2007) Analysis of sky conditions using 40 yearrecords of solar radiation data in China. Theor Appl Climatol89:83–94

Chen B et al (2003) Enhanced climatic warming in the Tibetan Plateaudue to doubling CO2: a model study. Clim Dyn 20:401–413

Duan AM, Wu GX (2006) Change of cloud amount and the climatewarming on the Tibetan Plateau. Geophys Res Lett 33, L22704.doi:10.1029/2006GL027946

European Center for Medium-Range Weather Forecasts (ECMWF)(2002) Workshop on re-analysis, 5–9 November 2001, ERA-40Proj. Rep. Ser. 3, ECMWF, Reading, UK, 443 pp

IPCC (2007) Summary for policymakers of climate change 2007: thephysical science basis. Contribution of Working Group I to theFourth Assessment Report of the Intergovernmental Panel onClimate Change (M). Cambridge University Press, Cambridge, UK

Kaiser DP, Qian Y (2002) Decreasing trends in sunshine duration overChina for 1954–1998: indication of increased haze pollution?Geophys Res Lett 29, 2042. doi:10.1029/2002GL016057

Kalnay E et al (1996) The NCEP/NCAR 40-year reanalysis project.Bull Am Meteorol Soc 77(3):437–471

Kumari BP et al (2007) Observational evidence of solar dimming:offsetting surface warming over India. Geophys Res Lett 34,L21810. doi:10.1029/2007GL031133

Liang F, Xia XA (2005) Long-term trends in solar radiation and theassociated climatic factors over China for 1961–2000. AnnGeophys 23:2425–2432

Liepert BG (2002) Observed reductions of surface solar radiation atsites in the United States and worldwide from 1961 to 1990.Geophys Res Lett 29(10), 1421. doi:10.1029/2002GL014910

Liu XD, Chen BD (2000) Climatic warming in the Tibetan Plateauduring recent decades. Int J Climatol 20:1729–1742

Luo YF et al (2001) Characteristics of the spatial distribution andyearly variation of aerosol optical depth over China in last30 years. J Geophys Res Atmos 106:14501–14513

Norris JR, Wild M (2007) Trends in aerosol radiative effects overEurope inferred from observed cloud cover, solar “dimming”,and solar “brightening”. J Geophys Res 112, D08214.doi:10.1029/2006JD007794

456 Q. You et al.

Ohmura A, Lang H (1989) Secular variation of global radiation inEurope. In: Lenoble J, Geleyn J-F (eds) IRS ’88: Currentproblems in atmospheric radiation. Deepak, Hampton, VA,USA, pp 298–301

Power HC (2003) Trends in solar radiation over Germany and anassessment of the role of aerosols and sunshine duration. TheorAppl Climatol 76:47–63

Qian Y et al (2006) More frequent cloud-free sky and less surfacesolar radiation in China from 1955 to 2000. Geophys Res Lett33, L01812. doi:10.1029/2005GL024586

Qian Y et al (2007) Variability of solar radiation under cloud-free skiesin China: the role of aerosols. Geophys Res Lett 34, L12804.doi:10.1029/2006GL028800

Ramanathan V et al (2007) Warming trends in Asia amplified bybrown cloud solar absorption. Nature 448:575–578

Sanchez-Lorenzo A et al (2007) Recent spatial and temporalvariability and trends of sunshine duration over the IberianPeninsula from a homogenized data set. J Geophys Res Atmos112(D20):115.1–115.18. doi:10.1029/2007JD008677

Sanchez-Lorenzo A et al (2008) Spatial and temporal trends insunshine duration over Western Europe (1938–2004). J Climate21:6089–6098

Sen PK (1968) Estimates of regression coefficient based on Kendall’stau. J Am Stat Assoc 63:1379–1389

Shi GY et al (2008) Data quality assessment and the long-term trendof ground solar radiation in China. J Appl Meteorol Climatol47:1006–1016

Stanhill G, Cohen S (2001) Global dimming: a review of the evidencefor a widespread and significant reduction in global radiation

with discussion of its probable causes and possible agriculturalconsequences. Agric For Meteorol 107:255–278

Wild M et al (2005) From dimming to brightening: decadal changes insolar radiation at Earth’s surface. Science 308:847–850

Wild M et al (2007) Impact of global dimming and brightening onglobal warming. Geophys Res Lett 34, L04702. doi:10.1029/2006GL028031

Xia XA et al (2006) Analysis of downwelling surface solar radiationin China from National Centers for Environmental Predictionreanalysis, satellite estimates, and surface observations. J Geo-phys Res Atmos 111:D09103.1-D09103.9

Yeh TC, Gao YX (1979) Meteorology of the Qinghai-Xizang (Tibet)Plateau (in Chinese). Science Press, Beijing

You QL et al (2008a) Changes in daily climate extremes in the easternand central Tibetan Plateau during 1961–2005. J Geophys ResAtmos 113:D07101.1–D07101.17

You QL et al (2008b) Relationship between trends in temperatureextremes and elevation in the eastern and central Tibetan Plateau,1961–2005. Geophys Res Lett 35, L04704

Zhang T (2007) Perspectives on environmental study of response toclimatic and land cover/land use change over the Qinghai-Tibetan plateau: an introduction. Arct Antarct Alp Res 39:631–634

Zhang YL et al (2004) Analysis of 40 year records of solar radiationdata in Shanghai, Nanjing and Hangzhou in Eastern China. TheorAppl Climatol 78:217–227

Zheng XB et al (2008) Long-term trends in sunshine duration overYunnan-Guizhou Plateau in Southwest China for 1961–2005.Geophys Res Lett 35, L15707

From brightening to dimming in sunshine duration over the eastern and central Tibetan Plateau (1961–2005) 457