contribución de radiación solar a variabilidad de temperaturas decadal sobre tierra

7/27/2019 Contribucin de radiacin solar a variabilidad de temperaturas decadal sobre tierra

1/7

EARTH,

ATMOSPHERIC,

www.pnas.org/cgi/doi/10.1073/pnas.1311433110 PNAS Early Edition | 1 of 6

Contribution of solar radiation to decadal temperature

variability over landKaicun Wanga,1 and Robert E. Dickinsonb

aState Key Laboratory of Earth Surface Processes and Resource Ecology, College of Global Change and Earth System Science, Beijing Normal University, Beijing100875, China; and bDepartment of Geological Sciences, Jackson School of Geosciences, The University of Texas at Austin, Austin, TX 78712

Edited by Mark H. Thiemens, University of California, San Diego, La Jolla, CA, and approved August 2, 2013 (received for review June 18, 2013)

Global air temperature has become the primary metric for judging

global climate change. The variability of global temperature on a

decadal timescale is still poorly understood. This paper examines

further one suggested hypothesis, that variations in solar radia-tion

reaching the surface (Rs) have caused much of the observed

decadal temperature variability. Because Rs only heats air during

the day, its variability is plausibly related to the variability of di-

urnal temperature range (daily maximum temperature minus its

minimum). We show that the variability of diurnal temperature

range is consistent with the variability of Rs at timescales from

monthly to decadal. This paper uses long comprehensive datasets

for diurnal temperature range to establish what has been the

contribution ofRs to decadal temperature variability. It shows that

Rs over land globally peaked in the 1930s, substantially decreased

from the 1940s to the 1970s, and changed little after that. Reduc-

tion ofRs caused a reduction of more than 0.2 C in mean temper-

ature during May to October from the 1940s through the 1970s,

and a reduction of nearly 0.2 C in mean air temperature during

November to April from the 1960s through the 1970s. This cooling

accounts in part for the near-constant temperature from the 1930s

into the 1970s. Since then, neither the rapid increase in tempera-

ture from the 1970s through the 1990s nor the slowdown of

warming in the early twenty-first century appear to be signifi-cantly

related to changes ofRs.

global dimming | global brightening | global warming |surface incident solar radiation | decadal variability

lobal temperature has become the primary metric forjudging global climate change, although many other factors

are recognized to be of comparable importance. The overallincrease of global temperature over the last century has beenlargely attributed to the increase of greenhouse gases (1). Lesswell understood are the reasons for the variability of this increaseon a decadal timescale. In particular, warming from 1900 to 1940was followed by three decades of flat or slightly decreasingtemperature, then three decades of very rapid temperature in-crease, then so far in this century, very little additional increase.The two most plausible explanations for the decadal variabilityare natural climate variability and variable degrees of coolingfrom anthropogenic releases of sulfur gas producing sulfateaerosols (2). This effect has long been proposed as a mechanism

to counter greenhouse warming (3), has become the basis formany geoengineering proposals (4), and has been used to attri-bute the lack of warming so far this century to the rapid growthof aerosols in Asia (5).

Besides the difference in sign of their temperature effects,sulfate aerosols are distinguished from greenhouse gases in thatthey only affect daytime radiation, i.e., surface incident solarradiation (Rs). Some kinds of natural variability can also actthrough affecting Rs, i.e., those involving cloud properties.

Changes of aerosol loading and cloud properties likely causedthe rapid decrease ofRs, measured at the surface from the 1950sto the 1980s, referred to as global dimming, and its partialrecovery after that (6). The plausible suggestion was made byWild et al. (7) that the rapid warming in the late twentieth

century was a consequence of the cessation of global dimming,possibly in part from the imposition of controls on sulfur emis-sion in the industrialized nations (8, 9).

This paper examines further the hypothesis that variations in Rshave caused much of the observed decadal variability in the rateof warming. Direct measurements ofRs cannot be quanti-tativelyrelated to such variability because they have been limited in theirgeographical coverage. The approach used here is to examine aglobal land dataset of diurnal temperature range (DTR). Thisconcept is not new, indeed, Wild et al. (7) noted (compare withtheirfigure 2) that the global pattern ofDTRwas similar to that oftheir global dimming and brightening. The present paper develops

the longest and most comprehensive dataset for DTRpossible,and, with some plausible assumptions, establishes what thecontribution ofRs has been to decadal temperature variability. Itindicates that a decrease ofRs from the 1940s through the 1970sreduced the global temperature trend over that period. However,global temperature does not appear to have been significantlyaffected by changing Rsafter that. The method is limited in that itis only applicable over land. As the effects of aerosols are likelyto be less over ocean, es-pecially in the Southern Hemisphere, thisapproach may exag-gerate the actual effect of aerosols on globaltemperature trends.

Results

Relationship Between Rs and DTR. This section establishes that lo-cally DTRis highly correlated with Rs,but that spatial and sea-

sonal variability precludes direct use of this correlation to infer Rswhere it is not already measured. In the absence of weathervariability, near-surface air temperature Ta over land decreaseswith time at night from longwave radiative cooling and reachesTminbefore sunrise. After sunrise, the surface is heated byRsand

Significance

Global air temperature has become the primary metric for

judging global climate change. The variability of global tem-

perature on a decadal timescale is still poorly understood. This

paper shows that surface incident solar radiation (Rs) over

land globally peaked in the 1930s, substantially decreased

from the 1940s to the 1970s, and changed little after that. The

cooling effect of this reduction of Rs accounts in part for thenear-constant temperature from the 1930s into the 1970s.

Since then, neither the rapid increase in temperature from the

1970s through the 1990s nor the slowdown of warming in the

early twenty-first century appear to be significantly related to

changes ofRs.

Authorcontributions:K.W. designed research; K.W. performed research; K.W. and R.E.D.

analyzed data; and K.W. and R.E.D. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

1To whom correspondence should be addressed.E-mail: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.

1073/pnas.1311433110/-/DCSupplemental.

G
http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110mailto:[email protected]:[email protected]:[email protected]://www.pnas.org/lookup/suppl/doi:10http://www.pnas.org/lookup/suppl/doi:10mailto:[email protected]://www.pnas.org/cgi/doi/10.1073/pnas.1311433110


2/7

1.5

1

0.5

0

AnnualDTRAnomaly(C)

0.5

1.5

1

0.5

0

AnnualDTRAnomaly(C)

0.5

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1311433110 Wang and Dickinson

Fig. 1. Correlation coefficients between monthly anomalies ofDTRand Rs. The

Rs observations are from the GEBA, and the homogenized maximum and

minimum air temperature at 2 m are from the Global Historical Clima-tology

Network (GHCN). Both datasets cover the period from 1950 to 2005. Each

point in the figure represents a weather station where both Rs and DTR areavailable for more than 120 mo. There are 524 stations in total.

. . . . . . . . . .

this heat is transferred as sensible heat H to the overlying air, raising

Tato Tmax in early afternoon. Therefore, changes ofDTR= Tmax Tmin, have been interpreted as directly related to changes ofRs (6, 7,1013). Here we explain how Rs and DTRconnect physically andhow their relationship varies with environment.

Fig. 1 shows the correlation of monthly anomalies of Rs col-lected by the Global Energy Balance Archive (GEBA) (14) withDTRfrom 1950 to 2005 at 524 globally distributed stations (see SIText and Fig. S1 forDTRdata sources and their quality control).The correlation coefficients between Rs and DTRare the highest inhumid areas and lower in arid or semiarid areas because thefraction of absorbed Rs generating H also depends on variable soilmoisture resulting from the frequency and in-tensity oprecipitation (15, 16). Besides its dependence on sur-face wetness(17), the partitioning of surface-absorbed Rsbetween H and latent

heat flux (1E) depends on land-cover conditions (18, 19) andatmospheric evaporative demand (20). In humid areas, both H and1E generally increase with Rs (21, 22), but under warm conditionsthe latter increases more (23). In arid or semiarid regions, 1E islimited by soil water supply and H can account for a higher portionof surface absorbed Rs. Fig. 2 shows, as expected from the abovediscussion, that the sensitivity of DTRto Rs is higher in arid orsemiarid areas than in humid areas.

Surface aridity changes seasonally for most monsoon areas,i.e., where it is wet only in a rainy season, but its interannual

Annual Rs Anomaly (Wm2

)


)

China, Cold Seasons


)

Europe, Cold Seasons


)

.

.

AnnualDTRAnomaly(C)

.

AnnualDTRAnomaly(C)

. . . . . . . . .

Fig. 2. The sensitivity ofDTRto Rs(in C per Wm2) calculated from the monthly

anomalies ofRsand DTR. The data used here are the same as in Fig. 1.

Fig. 3. Scatterplots of annual anomalies of regional DTR as a function ofannual anomaly of Rs during warm seasons (May to October) and cold sea-sons (November to April) from 1950 to 2005. The correlation coefficients are0.61 and 0.83 over China and 0.86 and 0.73 over Europe during the warm andcold seasons, respectively.

.

.

.

variability is expected to be much less than such seasonalchanges. To reduce the impact of seasonality, we used monthlyand annual anomalies rather than absolute values of DTRandRs. In the following discussion, we also divide a year into borealwarm seasons (May to October) and boreal cold seasons(November to April).

The correlations and sensitivity shown in Figs. 1 and 2 are thelowest in coastal areas. Evidently the impact of Rs on DTR inthese areas is masked by the impact of energy advection with

regular alteration between land breezes and ocean breezes. Thismasking can be substantially reduced by regional averaging ofDTRand Rs (6, 24).

Figs. 3 and 4 compare Europes and Chinas regional averageannual anomalies of DTR with those of Rs. These quantitiesagree quite well, partly because of their better data density anddata continuity (Fig. S3). The agreement between regional DTRand Rs over Europe has also been confirmed by both dataanalysis (10) and model simulation (24). In China, the decreaseof Rs is in good agreement with the reduction of DTRbefore1990 (11). However, Rs in China increased suddenly during theearly 1990s but not DTR and sunshine duration (25). The in-troduction of new pyranometers from 1990 to 1993 introducedthis inhomogeneity into the Rs observations (25, 26).

Fig. 4 also shows that DTRhas had a larger temporal vari-ability than Rs, a consequence of the annual variability of pre-cipitation leading to variations in the partitioning of surfaceabsorbed Rs between 1E and H. The Intergovernmental Panel onClimate Change (IPCC) Fourth Assessment Report (AR4) con-cluded that precipitation has had large annual variability duringthe last century, but that its long-term trend and thus its impact onthe long-term trend ofDTRhas been negligible (1), as con-firmedby Figs. 3 and 4 and the following sections. The impact of annualvariability of precipitation is largely removed by using 5-ysmoothing of the anomalies ofDTRas in the following.

Variability ofDTR, a Proxy of Rs, from 1900 to 2010. This sectionestablishes what is available as a global record for DTR vari-ability. For estimation ofDTRover land with optimum spatial andtemporal coverage and the highest quality, we combined three datasources (2729) (see SI Text fordetailed information)

China, Warm Seasons Europe, Warm Seasons
http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110http://www.pnas.org/cgi/doi/10.1073/pnas.1311433110


3/7

China, Cold Seasons

.

.

urope, o easons

.

.

AnnualDTRAnomaly(C)

AnnualDTRAnomaly(C)

AnnualRsAnomaly(Wm2)

AnnualRsAnomaly(Wm2)

China, Warm Seasons. .

.

Europe, Warm Seasons

.

Wang and Dickinson PNAS Early Edition | 3 of 6

EARTH,ATMOSPHERIC,

A s i a

1900 1950 2000

0 .5

0 . 5

0

Au st ra li a

1900 1950 2000

0.5

0

0.5

FiveyearAverageAnomalyofDiuranlTemperatureRange(DTR,

C)

South America

North America

0 . 5

0

1900 1950 2000 1900 1950 2000

A f r i ca0.5

0

Europe

1900 1950 2000 1900 1950 2000

0 .5

0

0 .5

0 .5

0

0.5

Fig. 4. Regionally averaged annual anomalies of RS (in blue) and DTR (in

green) during boreal warm seasons (May to October) and boreal cold sea-sons

(November to April) from 1950 to 2005. Data used here are the same as in Fig.

3. Equivalent plots for the United States are given in Fig. S2.

for the past 110 y. Monthly anomalies ofDTRwere derived byremoving its seasonal cycle. Observations of DTR had thehighest density in North America. To mitigate the impact of thedifferent data densities, monthly anomalies ofDTRwere binnedinto 5 5 grids. Given the low correlation between DTRandRs in coastal areas, we only selected the grids with more than50% of their area over land, as shown in Fig. S4. The monthlyanomalies at each grid were averaged into regional monthlyanomalies, and then into annual values and 5-y average annualanomalies at each region, as plotted in Fig. 5.

Europe is the only region where measurements of both DTRand Rs extend back to the 1920s (6). DTRgenerally increased inEurope from the 1920s to the 1950s. After the late 1950s, itbegan to decrease until the 1980s, and since the 1990s increased.

These variations ofDTRare consistent with those of observed Rs(6, 24, 30). The better agreement of warm-season variability ofDTRwith that ofRs is consistent with the largerRs during warmseasons.

Attempts have been made to correlate annual Rs and DTRbothat regional and global scales (6, 7). However, existing studieshave not recognized that DTR and Rs have different seasonalcycles; Rs is largest in summertime as a result of higher solarelevation. However, DTR is relatively low in moist summersbecause of the small fraction of Rs that is partitioned into H.Therefore, annual variability ofDTRis primarily determined byits variability during seasons other than summer. DTR and Rsagree well both for warm and cold seasons, and variability of Rsover warm seasons is more representative of its annual vari-ability. The reported annual variability of Rs, therefore, agrees

better with DTR over warm seasons over Europe (and otherregions) than that over an entire year or cold seasons. VariabilityofDTRover warm and cold seasons is substantially different atboth the regional scale (Fig. 5) and the global scale (Fig. 6). Forthis reason, it is essential to consider these differences in recon-structing variability ofRs from DTR.

In Asia, DTR substantially decreased from the 1950s to the1980s, was stable until 2000, and then decreased again, consistentwith Rs derived from sunshine duration (25) and the dimming ofdirectly measured Rs between 1960 and 1990 in China (11, 31). Asalready mentioned, after the 1990s, direct observations of Rsbecame inconsistent with those of DTRand sunshine (31), a re-sult of the urban bias ofRs observations. When averaged over allstations (400 stations) rather than over the 50 urban stations inChina with direct observations ofRs, Rs derived from

sunshine duration was stable during the 1990s and decreasedafter 2000 (25).

DTR substantially decreased in North America from 1900 to2010, consistent with the increase of cloudiness, in particular, oflow clouds (32), and decrease of sunshine duration (33). Cloudcover alone accounted for up to 63% of the regional annual DTRvariability in the United States from 1902 to 2002, with cloud-cover trends especially driving DTR in northern United States(34). Aerosol loading over North America was relatively light

(35) and rather stable during the past few decades (8). Obser-vations at six stations in the United States showed that Rs sig-nificantly increased from 1995 to 2007 (36, 37), primarily in the1990s (38).

As there is a good agreement between Rs and DTR, changes ofRs are expected to be similar to those of DTR, especially duringthe warm seasons. Fig. 5 shows the variability of DTRover landduring the past century, and hence provides qualitative estimatesof Rs variability over this period. However, it is difficult to re-construct Rs quantitatively using the variability of DTRbecauseof the changes of their relationship with time (e.g., from wet todry seasons; Fig. 3) and region (e.g., from humid to arid regions)(Figs. 1 and 2). Below, we describe another approach for usingDTRto estimate the impact ofRs on Ta.

Estimation of the Impact of Rs on Ta from 1900 to 2010. Elevatedgreenhouse gases (GHG) have increased atmospheric downwardlongwave radiation (Ld) (39, 40) and Ta (41) during the twentiethcentury. However, variability in radiative forcing from aerosolsand clouds complicates the attribution of the observed climatechange to the elevated GHG. The previous sections haveestablished a more comprehensive climatology forDTRthan thatavailable previously and its long-term variability is highly con-sistent with that ofRs. This climatology allows us to address thequestion of how much of the observed temperature change hasbeen a result of changes of Rs. For the following analysis, weassume: (i) Tmin is not changed by Rs; and (ii) DTR is onlychanged by changes of Rs (elaborated on in Discussion andConclusions).

The globally averaged anomaly of DTR is calculated directlyfrom its grid values. Daily mean air temperature Ta is commonlyestimated by Ta = 0.5 (Tmax + Tmin). As DTR= Tmax Tmin, we

Fig. 5. Five-year average of annual anomaly (black) of regional DTRfrom 1900

to 2010 averaged from the monthly anomalies at 5 5 grids (Fig. S4), whichis calculated from weather stations. For comparison, anomalies during boreal

warm seasons (May to October, red) and boreal cold seasons (No-vember to

April) are shown.


4/7

ImpactofRsonTa(C)

1

0.5

0

. 5

1900 1920 1940 1960 1980 2000

.

.

.

.

MayOctoberNovemberAprilEntire Year

Y e a r MeanAirTemperature(Ta)(C)

4 of 6 |www.pnas.org/cgi/doi/10.1073/pnas.1311433110 Wang and Dickinson

A

B

Fig. 6. (A) The 5-y smoothed impact of Rs on mean air temperature (Ta) over

global land during boreal warm seasons (May to October, red), boreal cold

seasons (November to April, blue), and an entire year (black). The mean air

temperatures over global land are also shown in B.

obtain Ta = Tmin+ 0.5 DTR. For the given assumptions, theimpact ofRs on daily mean Ta is 0.5 DTR (Fig. 6). Theseassumptions can be inaccurate for various reasons, e.g., changesof daytime radiation can be stored and released to changenighttime temperature. Observations from global flux networksshow that storage fraction is less than 10% of Rs at most

surfaces (42, 43). Allowing for this effect would likely amplifyour esti-mate of the impact ofRs on Ta over global land by afactor of-1.1.

We calculate the impact of Rs on mean Ta during the threetime periods: (i) 19002010 (the whole time period when dataare available); (ii) 19401984 (the global dimming period) and(iii) 19852010 (the global brightening period). The results aresummarized in Table 1.

Table 1 and Fig. 6 indicate that a reduction in Rshas reduced Taand that it decreased most rapidly during the dimming period of19401984. The rate of temperature increase during the cold

seasons has been reported to be much higher than that duringthe boreal warm seasons (May to October) (44). Fig. 6 showsthat warm-season Rs substantially decreased from the 1940s toearly 1950s and during the 1970s, resulting in a reduction ofmore than 0.2 C in Ta. Similarly, cold-season Rs substantiallydecreased from the 1960s through the 1970s, resulting in a de-crease of nearly 0.2 C in Ta. A subsequent increase ofRs wasonly significant over Europe. In conclusion, the variations ofRspartly accounted for the near absence of warming from

midcentury through the 1970s. The maximum cooling seen inthe early 1980s and early 1990s were consistent with the effectsexpected from the El Chichn and Pinatubo volcanoes,respectively. Fig. 6 also shows that the results are substantial-lydifferent for warm seasons, cold seasons, and the entire yearwhen using DTR to quantify the impact of Rs on air tem-perature (7).

Discussion and Conclusions

This paper shows, using direct Rs observations (6, 45) and sun-shine duration observations (25), that the interannual variabilityofDTRcan be used as a proxy for the long-term variability ofRs. In principle, this relationship should also be applicable tomodel simulations. AR4 climate models (46) show a weakmonotonic increase ofDTRfrom 1950 (44), compare with their

figure 5, suggesting that many of the models examined applied aslow constant ramp-up of aerosol forcing rather thanconcentrated increases before 1980 as indicated here. Changes ofDTR are expected to be directly related to H from surface tooverlying air but the magnitudes of these turbulent fluxes are notreadily estimated (22). Many parameters affect the relationshipbe-tween Rs and H, and consequently, the relationship betweenRsandDTR.

Impacts of land-cover and land-use change (i.e., urbanizationand irrigation) have been ignored here. In developing countries,such as China and India (47), there has been substantial ur-banization and increased irrigation activity (48) since 1900 withopposing and possibly largely cancelling effects on DTR(16, 49,50), so with impacts likely to be important locally, but likely to

be small at a regional scale (1). Precipitation had a large annualvariability but its long-term trend was negligible during the lastcentury (1), and so likely also its impact on the long-term trend ofDTR. At annual timescale or station-scale changes of pre-cipitation and land-cover/land introduce substantial uncertain-ties. Therefore, use ofDTRfor estimates of variability ofRsand

Table 1. The impact of Rs on daily mean air temperature (Ta) during three periods, 19002010,

19852010, and 19401984 (in C per 100 y)

Time periods Global land North America South America Europe Africa Asia Australia


5/7

Yearly

19002010a 0.11* 0.11* 0.68* 0.01 0.04 0.50* 0.084

19852010b 0.07 0.21 0.23 0.46* 0.34 0.57* 0.80

19401984 0.36* -0.46* 1.04* 0.24* 0.29* 0.53* 0.08

Warm seasons

19002010a 0.11* 0.15* 0.47 0.01 0.03 0.43* 0.06

19852010b 0.19 0.31 0.03 0.75* 1.10* 0.22 1.96*

19401984 -0.45* 0.62* 0.82 0.29* 0.43* 0.54* 0.08

Cold seasons

19002010a 0.12* 0.07 0.82* 0.03 0.14 0.52* 0.0919852010b 0.09 0.02 0.28 0.27 0.49 0.50 0.41

19401984 0.29* 0.22 1.58* 0.20* 0.28 0.55* 0.03

Negative values indicate that Rs reduced the rate of warming caused by the elevated GHG, and positive values

mean that Rs amplified the warming rate by GHG. We also divide the data into boreal warm seasons (May to

October) and cold seasons (November to April). The asterisk represents impact of Rs is statistically significant (i.e.,

pass the Students t confidence test at = 0.05).aTime periods for different regions are different and may cover only a fraction of 1900 2010 (Fig. 5).

bTime periods for different regions are different and may cover only a fraction of 1985 2010 (Fig. 5).


6/7

Wang and Dickinson PNAS Early Edition | 5 of 6

EARTH,ATMOSPHERIC,

its impact on Ta should be confined to decadal timescale andregional space scale.Because of the sparse distribution of measurement stations (1)

and changes in measurement methods (38) and instruments (25,51), direct observations cannot provide a reliable estimate ofRsover land during the past century, nor do current climate modelsgenerate long-term variability of Rs (52). This studyqualitatively reconstructs Rsover land from 1990 to 2010 usingthe latest homogenized DTRobservations at globally distributed

weather stations. It infers that Rs over land globally peaked inthe late 1930s, substantially decreased from the 1940s to the1970s, and changed little after that. These estimates areconsistent with observations ofRs and sunshine duration wherethese observa-tions are available.

More importantly, the DTRobservations allow us to estimatethe impact ofRs on the observed changes ofTa. Only changesbefore 1984 appear related to the observed temperature trendsand DTRvariability after 1995 indicates a negligible global im-pact of Rs variability. The small impact of Rs on Ta may bepartly a result of the low sensitivity of Ta to Rs, much lowerthan the sensitivity of Ta to longwave radiation caused bygreenhouse gases (53).The surface energy budget directly determines the Earths surface

climate and its changes, but on more local scales strongly interactswith transport processes. In consequence, most existing studieshave focused on the energy balance at the top of the atmosphere(5), which is indirectly related to surface Ta, depending on howclouds (54, 55), aerosols, and other feedbacks work. This paperprovides a direct and simple method to estimate the variability ofRsover land, which is applied from 1900 to

1.Trenberth KE, et al. (2007) Observations: Surface and atmospheric climate change.

Climate Change 2007: The Physical Science Basis. Contribution of Working

Group I to the Fourth Assessment Report of the Intergovernmental Panel on

Climate Change, eds Solomon S, et al. (Cambridge Univ Press), pp 236247.2.Booth BBB, Dunstone NJ, Halloran PR, Andrews T, Bellouin N (2012) Aerosols impli-

cated as a prime driver of twentieth-century North Atlantic climate variability. Nature

484(7393):228232.3.Rasool SI, Schneider SH (1971) Atmospheric carbon dioxide and aerosols: Effects of

large increases on global climate. Science173(3992):138141.4.Wei T, et al. (2012) Developed and developing world responsibilities for historical

climate change and CO2 mitigation. Proc Natl Acad Sci USA 109(32):1291112915.5.Kaufmann RK, Kauppi H, Mann ML, Stock JH (2011) Reconciling anthropogenic cli-mate

change with observed temperature 1998-2008. Proc Natl Acad Sci USA 108(29):

1179011793.6.Wild M (2009) Global dimming and brightening: A review. J Geophys Res114(D10):

D00D16.

7.Wild M, Ohmura A, Makowski K (2007) Impact of global dimming and brightening on

global warming. Geophys Res Lett34(4):L04702.

8.Wang KC, Dickinson RE, Liang S (2009) Clear sky visibility has decreased over land

globally from 1973 to 2007. Science323(5920):14681470.9.Streets DG, et al. (2009) Anthropogenic and natural contributions to regional trends in

aerosol optical depth, 1980-2006. J Geophys Res114(D10):D00D18.

10. Makowski K, Wild M, Ohmura A (2008) Diurnal temperature range over Europe

be-tween 1950 and 2005.Atmos Chem Phys8(21):64836498.11. Liu B, Xu M, Henderson M, Qi Y, Li Y (2004) Taking China s temperature: Daily

range, warming trends, and regional variations, 19552000. J Clim17(22):44534462.12. Dai A, DelGenio AD, Fung IY (1997) Clouds, precipitation and temperature

range. Nature386(6626):665666.13. Dai A, Trenberth KE, Karl TR (1999) Effects of clouds, soil moisture,

precipitation, and water vapor on diurnal temperature range. J Clim12(8):24512473.14. Gilgen H, Wild M, Ohmura A (1998) Means and trends of shortwave irradiance

at the surface estimated from global energy balance archive data. J Clim11(8):20422061.

15. Zhou LM, et al. (2009) Spatial dependence of diurnal temperature range trends

on precipitation from 1950 to 2004. Clim Dyn32(2-3):429440.16. Zhou LM, Dickinson RE, Tian YH, Vose RS, Dai YJ (2007) Impact of vegetation

removal and soil aridation on diurnal temperature range in a semiarid region:

Application to the Sahel. Proc Natl Acad Sci USA 104(46):1793717942.17. Wang KC, Liang S (2008) An improved method for estimating global

evapotranspi-ration based on satellite estimation of surface net radiation, vegetation

index, tem-perature, and soil moisture. J Hydrometeorol9(4):712727.18. Wang KC, et al. (2007) Influences of urbanization on surface characteristics as

derived from the Moderate-Resolution Imaging Spectroradiometer: A case study for the

Beijing metropolitan area. J Geophys Res112(D22):D22S06.

2010 and estimates the impact of this variability on surfacetemperature change.

Changes ofRs are primarily determined by changes of cloudsand aerosols. Aerosols are known to have accounted for vari-ability ofRs in Europe and China (25, 56), while clouds havebeen used to explain changes ofRs in the United States (36, 37)during the last two decades. Natural variability from clouds isexpected to be more regional and of shorter timescale than the

trends from aerosols, but otherwise we are not able to separatetheir effects. This paper also does not address the mechanismsthrough which clouds and aerosols respond to climate change(57), i.e., through changes of cloud-cover fraction or cloudheight (58).

Our analysis of impact ofRson Tadoes not account for warmingeffect of solar radiation absorbed by aerosols, i.e., from blackcarbon (5961). To zeroth order, aerosol absorption within thedaytime boundary layer will return the solar energy removed fromthe surface, so will not change DTRbut will contribute to warmingTa. Our analysis, in principle, cannot include the warming ofabsorbing aerosols in the aerosol layer although their scatteringand absorption effects on surface Rsare included.

AC KN OW LE DG ME NT S. Chinese homogenized dai ly maximum and

minimum temperature at 549 stations were provided by Prof. Zhongwei

Ya n. GE BA su rf ac e in ci de nt so la r ra di at io n da ta we re ki nd ly pr ov id ed by

Prof. Martin Wild. We thank Dr. Qian Ma for processing some data for this

study. This study was supported by the National Basic Research Program of

China (2012CB955302), the National Natural Science Foundation of China

(41175126), and the US Department of Energy (BER) Grant DE-FG02-

09ER64746.

19.Wang KC, Wang P, Li ZQ, Cribb M, Sparrow M (2007) A simple method to estimate

actual evapotranspiration from a combination of net radiation, vegetation index, and

temperature. J Geophys Res112(D15):D15107.

20.Wang KC, Dickinson RE, Liang S (2012) Global atmospheric evaporative demand over

land from 1973 to 2008. J Clim25(23):83538361.21.Wang KC, Dickinson RE, Wild M, Liang S (2010) Evidence for decadal variation in global

terrestrial evapotranspiration between 1982 and 2002: 2. Results. J Geophys Res

115(D20):D20113.

22.Wang KC, Dickinson RE (2012) A review of global terrestrial evapotranspiration:

Observation, modeling, climatology, and climatic variability. Rev Geophys 50(2):

RG2005.

23.Wang KC, Li ZQ, Cribb M (2006) Estimation of evaporative fraction from a combina-tion

of day and night land surface temperatures and NDVI: A new method to determine the

Priestley-Taylor parameter. Remote Sens Environ102(3-4):293305.24.Makowski K, et al. (2009) On the relationship between diurnal temperature range and

surface solar radiation in Europe. J Geophys Res114(D10):D00D07.

25.Wang KC, Dickinson RE, Wild M, Liang S (2012) Atmospheric impacts on climatic

variability of surface incident solar radiation.Atmos Chem Phys12(20):95819592.26.Tang WJ, Yang K, Qin J, Cheng CCK, He J (2011) Solar radiation trend across China in

recent decades: A revisit with quality-controlled data. Atmos Chem Phys 11(1):

393406.27.Lawrimore JH, et al. (2011) An overview of the Global Historical Climatology Network

monthly mean temperature data set, version 3. J Geophys Res116(D19):D19121.

28.Menne MJ, Williams CN, Vose RS (2009) The U.S. historical climatology network

monthly temperature data, version 2. Bull Am Meteorol Soc90(7):9931007.29.Li Z, Yan Z (2009) Homogenized daily mean/maximum/minimum temperature series for

China from 1960-2008.Atmos Ocean Sci Lett2(4):237243.30.Ohmura A (2006) Observed long-term variations of solar irradiance at the earths

surface. Space Sci Rev125(1-4):111128.31.Shi GY, et al. (2008) Data quality assessment and the long-term trend of ground solar

radiation in China. J Appl Meteorol Climatol47(4):10061016.32.Sun B, Karl TR, Seidel DJ (2007) Changes in cloud-ceiling heights and frequencies over

the United States since the early 1950s. J Clim20(15):39563970.33.Angell JK (1990) Variation in United States cloudiness and sunshine duration between

1950 and the drought year of 1988. J Clim3(2):296308.34.Lauritsen RG, Rogers JC (2012) U.S. diurnal temperature range variability and regional

causal mechanisms, 19012002. J Clim25(20):72167231.35.Wang KC, Dickinson RE, Su L, Trenberth KE (2012) Contrasting trends of mass and

optical properties of aerosols over the Northern Hemisphere from 1992 to 2011. Atmos

Chem Phys12(19):93879398.36.Long CN, et al. (2009) Significant decadal brightening of downwelling shortwave in the

continental United States. J Geophys Res114(D10):D00D06.

37.Augustine JA, Dutton EG (2013) Variability of the surface radiation budget over the

United States from 1996 through 2011 from high-quality measurements. J Geophys

Res118(1):4353.


7/7

6 of 6 |www.pnas.org/cgi/doi/10.1073/pnas.1311433110 Wang and Dickinson

38.Wang KC, Dickinson KE, Ma Q, Augustine JA, Wild M (2013) Measurement methods

affect the observed global dimming and brightening. J Clim26(12):41124120.

39.Wang KC, Liang S (2009) Global atmospheric downward longwave radiation over

land surface under all-sky conditions from 1973 to 2008. J Geophys Res114(D19):

D19101.

40.Wang KC, Dickinson RE (2013) Global atmospheric downward longwave radiation at the

surface from ground-based observations, satellite retrievals and reanalyses. Rev

Geophys51(2):150185.41.Dickinson RE, Cicerone RJ (1986) Future global warming from atmospheric trace gases.

Nature319(6049):109115.42. Jacobsen A (1999) Estimation of the soil heat flux/net radiation ratio based on

spectral vegetation indexes in high -latitude Arctic areas. Int J Remote Sens20(2):445461.

43.Clothier BE, et al. (1986) Estimation of soil heat flux from net radiation during the

growth of alfalfa.Agric For Meteorol37(4):319329.44.Wallace JM, Fu Q, Smoliak BV, Lin P, Johanson CM (2012) Simulated versus observed

patterns of warming over the extratropical Northern Hemisphere continents during the

cold season. Proc Natl Acad Sci USA109(36):1433714342.45.Wild M, et al. (2009) Global dimming and brightening: An update beyond 2000. J

Geophys Res114(D10):D00D13.

46.Zhou LM, Dickinson RE, Dai AG, Dirmeyer P (2010) Detection and attribution of an-

thropogenic forcing to diurnal temperature range changes from 1950 to 1999:

Comparing multi-model simulations with observations. Clim Dyn35(7-8):12891307.

47.Zhou LM, et al. (2004) Evidence for a significant urbanization effect on climate in China.Proc Natl Acad Sci USA101(26):95409544.

48.Mukherji A, et al. (2009) Revitalizing Asias Irrigation: To Sustainably Meet To-

morrows Food Needs (Food and Agricultural Organization of the United Nations,

Rome)Rome,.

49.Geerts B (2002) On the effects of irrigation and urbanisation on the annual range of

monthly-mean temperatures. TheorAppl Climatol72(3-4):157163.50.Sacks WJ, Cook BI, Buenning N, Levis S, Helkowski JH (2009) Effects of global irriga-

tion on the near-surface climate. Clim Dyn33(2-3):159175.51.Wang KC, Augustine J, Dickinson RE (2012) Critical assessment of surface incident solar

radiation observations collected by SURFRAD, USCRN and AmeriFlux networks from

1995 to 2011. J Geophys Res117(D23):D23105.

52.Wild M, Schmucki E (2011) Assessment of global dimming and brightening in IPCC-

AR4/CMIP3 models and ERA40. Clim Dyn37(7):16711688.53.Stanhill G (2011) The role of water vapor and solar radiation in determining

temperature changes and trends measured at Armagh, 1881-2000. J Geophys Res116(D3):D03105.

54.Spencer RW, Braswell WD (2010) On the diagnosis of radiative feedback in the pres-

ence of unknown radiative forcing. J Geophys Res115(D16):D16109.

55.Dessler AE (2010) A determination of the cloud feedback from climate variations over

the past decade. Science330(6010):15231527.56.Streets DG, Wu Y, Chin M (2006) Two-decadal aerosol trends as a likely explanation of

the global dimming/brightening transition. Geophys Res Lett33(15):L15806.57.Kerr RA (2010) Climate change. El Nio lends more confidence to strong global

warming. Science330(6010):1465.

58.Davies R, Molloy M (2012) Global cloud height fluctuations measured by MISR on Terrafrom 2000 to 2010. Geophys Res Lett39(3):L03701.

59.Jacobson MZ (2012) Investigating cloud absorption effects: Global absorption prop-

erties of black carbon, tar balls, and soil dust in clouds and aerosols. J Geophys Res117(D6):D06205.

60.Ramanathan V, et al. (2005) Atmospheric brown clouds: Impacts on South Asian cli-

mate and hydrological cycle. Proc Natl Acad Sci USA102(15):53265333.61.Ramanathan V, et al. (2007) Warming trends in Asia amplified by brown cloud solar

absorption. Nature448(7153):575578.

contribución de radiación solar a variabilidad de temperaturas decadal sobre tierra

Documents