at SciVerse ScienceDirect

Atmospheric Environment 77 (2013) 178e184

Contents lists available

Atmospheric Environment

journal homepage: www.elsevier .com/locate/atmosenv

Beryllium-7 atmospheric deposition and soil inventory on thenorthern Loess Plateau of China

Fengbao Zhang a,b, Bo Zhang b, Mingyi Yang a,b,*

a State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Institute of Soil and Water Conservation, Northwest A&F University,Yangling, Shaanxi 712100, Chinab Institute of Soil and Water Conservation, Chinese Academy of Science and Ministry of Water Resources, Yangling, Shaanxi 712100, China

h i g h l i g h t s

� 7Be concentrations in rainfall were artificially high due to 7Be dry deposition.� 7Be deposition input showed the highest in summer and the lowest in winter.� 7Be inventories in undisturbed soil were obviously unimodal over the year.

a r t i c l e i n f o

Article history:Received 11 August 2012Received in revised form12 April 2013Accepted 3 May 2013

Keywords:Loess Plateau7BeDeposition fluxSoil inventory

* Corresponding author. State Key Laboratory of Soilon the Loess Plateau, Institute of Soil and Water Consversity, No. 26 Xinong Road, Yangling District, ShaaTel.: þ86 (0)29 87012884; fax: þ86 (0)29 87016082.

E-mail address: ymyzly@163.com (M. Yang).

1352-2310/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.atmosenv.2013.05.002

a b s t r a c t

Beryllium-7 is a potentially powerful tracer of soil erosion, but information on 7Be atmospheric depo-sition and associated soil inventories on the Loess Plateau of China is not readily available. In the studyreported in this paper, we measured the 7Be inventories in undisturbed soil at different sampling timeson the northern Loess Plateau of China for three years, between 2010 and 2012, and estimated the 7Bedeposition fluxes and the daily 7Be inventories in undisturbed soil. The annual 7Be depositionfluxes during this period varied between 1303 � 119 and 2222 � 147 Bq m�2, with a mean of1759 � 416 Bq m�2. There is a marked seasonality for the 7Be deposition fluxes with the maximum insummer, approximately 50% to the annual deposition flux, and the minimum in winter, approximately 5%to the annual deposition flux. Precipitation amounts can explain more than 70% of the variation in 7Bedeposition flux. 7Be deposition in the form of dustfall, dew and frost make a significant contribution tothe 7Be deposition flux in the study region. The daily 7Be inventories in undisturbed soil varied markedlythrough time and ranged between 89.2 and 941.8 Bq m�2 with a mean of 392 � 210 Bq m�2. Theydemonstrated a unimodal distribution over the year, with the highest values in August or September andthe lowest in late winter or early spring.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Cosmogenic 7Be (T1/2 ¼ 53.3 d) is produced in the uppertroposphere and mainly in the stratosphere as a product of thespallation reaction of oxygen and nitrogen nuclei with high-energy cosmic ray particles (Lal et al., 1958). After 7Be is pro-duced, it rapidly forms BeO or Be(OH)2 by ionic reactions andbecomes associated with sub-micrometre aerosol particles(Papastefanou and Ioannidou, 1995; Cho et al., 2007). Subse-quently, 7Be enters the marine and terrestrial environments

Erosion and Dryland Farmingervation, Northwest A&F Uni-nxi Province 712100, China.

All rights reserved.

through wet and dry deposition processes (Wallbrink and Murray,1994; Ioannidou and Papastefanou, 2006; Doering and Akber,2008). Many investigators (Olsen et al., 1985; Dibb, 1989; Harveyand Matthews, 1989; Todd et al., 1989; Caillet et al., 2001;Graham et al., 2003) have suggested a linear relationship betweenthe amount of precipitation and 7Be deposition flux and haveshown that precipitation plays a dominant role in 7Be depositionflux. Annual, seasonal and monthly variations of 7Be depositionfluxes, which are determined by the 7Be production rate in theatmosphere, the extent of stratospheric to tropospheric air ex-change, tropospheric circulation patterns, latitude, and the effi-ciency of the removal of 7Be by wet and dry deposition, have alsobeen reported for many areas of the world (Narazaki et al., 2003;González-Gómez et al., 2006; Baskaran and Swarzenski, 2007; JuriAyub et al., 2009). Upon reaching the land surface, 7Be is rapidlyand strongly fixed by soil particles and other ground cover and

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184 179

activities are readily measured by gamma spectrometry (Bondiettiet al., 1984; Wallbrink and Murray, 1996; Papastefanou et al., 1999;Zhang et al., 2011). 7Be in bare soil is restricted to an upper surfacelayer of approximately 20 mm, predominantly within the first10 mm, and it decreases exponentially with increased depth(Wallbrink and Murray, 1996; Blake et al., 1999; Walling et al.,1999; Wilson et al., 2003; Yang et al., 2006). 7Be has been usedsuccessfully to document soil erosion occurring on bare soils sincethe 1990s in various places around the world (Wallbrink andMurray, 1993; Blake et al., 1999; Walling et al., 1999; Wilsonet al., 2003; Yang et al., 2006; Sepulveda et al., 2008; Wallinget al., 2009; Schuller et al., 2010).

Soil erosion is a major environmental and agricultural problemon the Loess Plateau of China. There is an urgent need to quantifyamounts of soil loss and 7Bemeasurement offer a valuable means ofquantifying short-term erosion rates. For such studies, a good un-derstanding of the local deposition flux of 7Be and the temporalvariation of the 7Be soil inventory is necessary. However, infor-mation on 7Be deposition flux and soil inventories is relativelyscarce on the Loess Plateau. Although a large amount of data on 7Beatmospheric deposition is available for other areas of the world(Walling, 2013), 7Be deposition fluxes are site specific and stronglydependent on location, and particularly latitude and local meteo-rological conditions. The underlying objectives of the study re-ported here were to document 7Be deposition fluxes and toreconstruct the temporal distribution of 7Be fallout inputs andassociated daily 7Be inventories in undisturbed soil on the northernLoess Plateau by measuring 7Be inventories in undisturbed soil atregular intervals and to examine the short-term variations and thefactors that may possibly influence them. No previous publicationhas reported the 7Be deposition fluxes for this region and this studytherefore provides information on inputs of 7Be to the northernLoess Plateau to add to existing databases.

2. Methods for determining 7Be deposition flux

Generally, 7Be deposition flux has been determined by collectingthe rainfall and dry fallout for individual rainfall events or at regularintervals, recovering the 7Be from the collected rainfall by chemicalmeans, measuring the 7Be activity by gamma spectrometry, andthen calculating the 7Be fallout input as the product of the 7Beconcentration in the collected rainfall and the amount of precipi-tation during the rainfall event or sampling interval (Wallbrink andMurray,1994; Caillet et al., 2001). However, the chemical procedurerequired to separate 7Be from rainfall is complex and timeconsuming. It is known that the 7Be inventory in undisturbed soildirectly reflects the 7Be fallout input and is controlled by the 7Befallout input and the decay of 7Be. Variations of the 7Be inventory inundisturbed soil can reflect the variations of 7Be fallout input.Previous investigators have also shown that undisturbed soil sitesserve as effective natural repositories for atmospherically derivedradionuclides and that soil inventories from such sites can be usedto estimate past atmospheric fluxes (Hardy et al., 1973; Moore andPoet, 1976; Nozaki et al., 1978).Walling et al. (2009) calculated themean 7Be concentration in rainwater (Cm in Bq l�1) using Eq. (1),based on measurements of the 7Be inventories in undisturbed soiland records of daily rainfall. They then used this value to calculatethe 7Be deposition flux and reconstructed the temporal distributionof 7Be fallout input during the periods between inventory mea-surements. Eq. (1) was expressed as:

Cm ¼hAref

�T��Aref

�t¼ 0

�exp

��lT

�i.ZT

0

IðtÞexp½�lðT�tÞ�dt

(1)

where Aref (Bq m�2) is the 7Be inventory in undisturbed soil, I(l m�2) is the daily rainfall, l is the daily decay constant for 7Be, andT (d) is the number of days in the study period. In this equation,Walling et al. (2009) assumed that the 7Be concentration in rain-water at a given site is constant at the weekly or monthly timescale,that all the 7Be fallout is delivered as wet deposition and that drydeposition can be ignored during the inter-sampling period.Walling et al. (2009) also demonstrated that it is reasonable toassume that the mean 7Be concentration in rainfall remainsessentially constant on the weekly or monthly timescale. However,dry deposition (dustfall, dew and frost) may be significant in semi-arid and arid regions and will influence the results obtained usingthis approach. Further details are provided below.

3. Study site and experimental procedures

3.1. Study site

The study site was located in the Dunshan watershed at theAnsai Research Station of Soil and Water Conservation, ChineseAcademy of Science, situated within Ansai County, Shaanxi Prov-ince, on the northern Loess Plateau region of China (109�1902300E,36�5103000N). The climate is warm and semi-arid. The mean annualprecipitation is approximately 510 mm, most of which falls duringthe period July to September and causes severe soil erosion. Theannual mean temperature is 8.8 �C, and the annual evaporationranges from 1500 to 1800 mm. The soil type is Huangmian soil(Calcaric Cambisols, FAO), developed by wind deposits and char-acterized by a yellow colour. Dustfall occurs frequently in this re-gion, particularly in spring and early summer.

3.2. Soil sampling and laboratory procedures

Five adjacent experimental plots with a 0� slope in the AnsaiResearch Station were selected as our experimental sites for sam-pling soil. On December 31, 2009, these plots were manually tilledto mix the upper 25 cm layer. Extraneous materials such as stones,roots or grass fragments were removed, and the surface wassmoothed by replicating local cultivation practices. Aref (t ¼ 0) wasassumed to be zero after tilling. To maintain the surface of theseplots bare, herbicide was sprayed on to prevent the growth ofvegetation during the study period. Further disturbance of the plotswas prevented. Soil samples were routinely collected from April 1,2010, to December 25, 2012. In each plot, three soil samples werecollected using a scraper-plate to a depth of 2.5 cmwithin an area of10 � 10 cm2 and these were mixed to provide a single compositesoil sample. A total of five mixed soil samples were sent back to thelaboratory to measure the 7Be activity on each sampling occasion.Few samples were collected in winter and early spring because ofthe low amount of precipitation. All soil samples were air-dried,weighed, dispersed and passed through a 1 mm sieve. The sam-ples of sieved soil were then packed into identical plastic column-shaped boxes prior to detecting 7Be activity.

Measurements of the 7Be activity in soil were undertaken bygamma spectrometry using a high-resolution, low-background,low-energy, hyperpure n-type germanium coaxial r-ray detector(EG&G ORTEC, Oak Ridge, TN, USA). Further details of the 7Bemeasurement procedure are provided by Yang et al. (2006) andZhang et al. (2011). The measured 7Be activity was always correctedto the sampling day using the decay constant. The mean inventoryof the five soil samples was used to represent the 7Be inventory inundisturbed soil on each sampling occasion. Statistical analyseswere performed using SPSS PASW Statistics (Version 18.0)software.

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184180

3.3. Meteorological data

A record of daily precipitation during the study period wasobtained from the Ansai Research Station of Soil and Water Con-servation, Chinese Academy of Science.

Fig. 1. Relationship between measured 7Be inventories in undisturbed soil anddecayed cumulative precipitation.

4. Results and discussion

4.1. 7Be atmospheric deposition

The measured 7Be inventory in undisturbed soil during thestudy period varied between 92.5 � 15.6 and 860.8 � 38.5 Bq m�2

(Table 1). The 7Be inventory in undisturbed soil reflects the cu-mulative input of 7Be after decay. If we assume that the cumulativeprecipitation has the same half life as 7Be, a significant linearrelationship is obtained between the measured 7Be inventories inundisturbed soil and the decayed cumulative precipitation (Fig. 1).The significant linear relationship (r2 ¼ 0.88) implies that 7Be wetdeposition by precipitation controlled the 7Be deposition input andit is reasonable to use Eq. (1) to estimate the 7Be concentration inrainfall at the study site for the timescale of the individual samplinginterval.

The mean 7Be concentrations in rainfall during the samplingintervals which ranged from 19 to 161 days were calculatedusing Eq. (1) and are presented in Table 1. The calculated values ofmean 7Be concentration in rainfall for each sampling interval variedwidely within the range1.2 � 0.2 to 10.1 � 1.7 Bq l�1, with oneexceptionally high value of 20.8 � 2.6 Bq l�1. The higher valueswere related to periods with low precipitation, which may occur inany month from November to June at the study site. The calculatedmean 7Be concentration in rainfall quickly decreased withincreasing precipitation during the study period (Fig. 2), showing asimilar response to those reported by Caillet et al. (2001) andBaskaran et al. (1993) who studied the 7Be concentration in rainfallduring individual rainfall events. After monthly averaging, the

Table 1Data for the 7Be inventories in undisturbed soil, 7Be deposition fluxes, and the calculate

Sampling date Samplinginterval (d)

Aref (Bq m�2 � SD)a

1-Apr-2010 90 92.5 � 15.61-Jun-2010 61 265.7 � 20.220-Jun-2010 19 352.8 � 28.610-Jul-2010 20 396.5 � 30.431-Jul-2010 21 354.8 � 28.520-Aug-2010 20 539.8 � 50.211-Sep-2010 22 455.3 � 44.530-Sep-2010 19 383.9 � 36.45-Nov-2010 36 308.4 � 27.315-Apr-2011 161 140.2 � 20.424-May-2011 39 331.5 � 18.326-Jun-2011 33 355.1 � 29.511-Aug-2011 46 503.0 � 45.25-Sep-2011 25 853.0 � 44.225-Sep-2011 20 860.8 � 38.515-Oct-2011 20 724.7 � 40.830-Nov-2011 40 617.4 � 39.507-Mar-2012 98 206.5 � 23.212-May-2012 66 234.5 � 27.307-Jun-2012 26 387.0 � 32.401-Jul-2012 24 592.9 � 35.720-Jul-2012 19 552.2 � 30.511-Aug-2012 22 493.4 � 41.503-Sep-2012 23 594.2 � 50.226-Sep-2012 23 697.6 � 48.515-Oct-2012 19 622.7 � 25.625-Dec-2012 71 302.2 � 32.4

a The data in the grey column were influenced by manual tilling.

lower values of 7Be concentration in rainfall were associated withthe rainy season (July, August and September). Using theIndependent-Sample Test, sampling intervals with <30 mm pre-cipitation were shown to be characterized by significantly higher(with 95% confidence) mean 7Be concentrations in rainfall thansampling intervals with >30 mm during the study period.

Direct measurements of 7Be concentration in rainfall have beenreported for other locations around the world. The range of ourvalues is slightly greater than those reported for some other areas,e.g., 1.11e5.55 Bq l�1 in The Netherlands (Bleichrodt and VanAbkoude, 1963), 0.59e2.74 Bq l�1 at Chilton and Milford Haven inthe United Kingdom (Peirson, 1963), 0.02e5.9 Bq l�1 in south-eastern Australia (Wallbrink and Murray, 1994), and 0.73e5.05 Bq l�1 in Cantabria, Spain (Ródenas et al., 1997), but similar tothose observed by Baskaran et al. (1993), with 0.11e24.84 Bq l�1 inTexas (USA), and by Caillet et al. (2001), with 0.93e10.45 Bq l�1 in

d mean 7Be concentrations in precipitation during the sampling intervals.

Precipitation(mm)

Cm (Bq L�1 � SD) 7Be flux (Bq m�2 � SD)

16.5 8.1 � 1.4 133.2 � 22.591.5 3.6 � 0.2 331 � 19.423.3 7.2 � 0.6 168.4 � 14.936.6 3.8 � 0.3 140.1 � 9.412.8 4.6 � 0.5 58.7 � 5.9

109.5 2.7 � 0.3 291.0 � 30.950.2 1.2 � 0.2 61.8 � 8.412.7 2.5 � 0.1 31.2 � 1.822.4 3.9 � 0.3 87.1 � 5.817.7 8.6 � 1.4 152.4 � 20.773.2 4.1 � 0.1 302.9 � 6.08.6 20.8 � 2.6 178.3 � 18.4

140.0 3.0 � 0.3 424.0 � 32.6169.6 3.5 � 0.1 595.9 � 11.579.7 2.9 � 0.1 231.3 � 4.153.3 1.2 � 0.2 66.4 � 9.964.9 4.1 � 0.3 265.3 � 16.93.7 9.3 � 3.3 34.5 � 12.4

49.7 3.8 � 0.5 187.4 � 22.326.2 10.0 � 0.6 261.4 � 15.485.2 3.9 � 0.25 328.6 � 12.753.3 1.8 � 0.1 97.8 � 2.975.2 1.3 � 0.3 99.5 � 23.5

121.2 2.2 � 0.2 264.7 � 22.584.5 3.5 � 0.2 298.8 � 13.18.1 10.8 � 1.7 87.7 � 13.8

15.5 6.0 � 2.4 93.0 � 37.6

Fig. 2. Magnitude of precipitation vs. 7Be concentrations in precipitation at the sam-pling interval.

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184 181

Geneva, Switzerland. Overall, the values obtained in our studywererelatively high when compared with other reported data, especiallyin winter, spring and early summer. Possible physically-based rea-sons for the higher calculated values of mean 7Be concentration inrainfall associated our study include: (1) Enhanced air exchangebetween the stratosphere and troposphere in winter and earlyspring increases 7Be concentration in the troposphere in the mid-latitude region (Feely et al., 1989; Doering and Akber, 2008;Heikkilä et al., 2008), (2) The climate of our study site inwinter andspring is controlled by the Siberian High which originates from thenorthern part of the Asian continent and Siberia and consists of acold and dry air mass. Some reports have indicated that this coldand dry air mass contains a high concentration of 7Be and increasesthe local 7Be concentration in air (Narazaki et al., 2003; Lee et al.,2004; Yamamoto et al., 2006; Cho et al., 2007), (3) Snow andsleet, which have been reported to bemore efficient than rain dropsat scavenging 7Be from air mass (McNeary and Baskaran, 2003;Ioannidou and Papastefanou, 2006), are the main forms of precip-itation from December to April at the study site, (4) When cleanedby precipitation the local atmosphere has sufficient time to berecharged by the 7Be-enriched air masses because of the lowerprecipitation and increased inter-arrival time of rainfall events,which maintains high 7Be concentrations in the local atmosphereduring winter and spring (Caillet et al., 2001), (5) The thunder-showers which frequently occur on the Loess Plateau in the sum-mer and early autumn months (Yuan et al., 2007) can enhancestratosphericetropospheric exchange (Burchfield et al., 1983),which can result in higher 7Be concentrations in rainfall.

In addition, the higher 7Be concentration in rainfall shown byour study could also reflect the assumptions of Eq. (1) which ig-nores 7Be dry deposition (dustfall, dew and frost). 7Be deposition inthe form of dustfall, dew and frost, which is included in wetdeposition by precipitation when calculating the mean 7Be con-centrations in rainfall using Eq. (1), is significant in the study area(Li et al., 2008; Wang and Zhang, 2011). This is particularly the casefor dustfall which mainly occurs at times of lower precipitation inwinter and spring at the study site, and could be expected to resultin artificially high values for the calculated mean 7Be concentrationin rainfall in the study. The likely effects of dry deposition on 7Beatmospheric deposition are discussed below.

A summary of the values of 7Be deposition flux calculated as theproduct of the calculatedmean 7Be concentration in rainfall and thevalues of daily precipitation for each sampling interval are pre-sented in Table 1. Inter-annual and intra-annual variations of 7Bedeposition fluxes are significant at the study site. The meanmonthly 7Be deposition flux was 146.6 � 146.4 Bq m�2 with a

coefficient of variation of 99.9%. The maximum values were foundin August or September and the minimum in the winter months(Fig. S1). There is a marked seasonality in the seasonal values of 7Bedeposition flux, which varied markedly from 25.0 � 9.6 to985.2 � 60.1 Bq m�2 with a mean of 440 � 309 Bq m�2. Maximumvalues occurred in summer andminimumvalues inwinter (Fig. S2).The contribution of 7Be deposition in summer to the annualdeposition fluxwas 46.4� 6.7%, but inwinter it was only 3.9� 2.1%,and in spring and in autumn about 25.9 � 3.9% and 23.8 � 11.9%,respectively. The annual 7Be deposition flux between 2010 and2012 was 1303 � 119, 2222 � 147 and 1753 � 149 Bq m�2,respectively (Fig. S3). Inter-annual and intra-annual variations of7Be deposition fluxes were most certainly controlled by the dis-tribution of rainfall during the study period at our study site(Figs. S1e3). Many studies have demonstrated a strong positivecorrelation between precipitation and 7Be deposition flux (Olsenet al., 1985; Dibb, 1989; Todd et al., 1989; Dueñas et al., 2001),and our study showed a similar trend. Whether considering the 7Bedeposition flux associated with individual rainfall events, monthlyvalues, values for each sampling interval or seasonal values, astrong linear correlation with precipitation is evident (Fig. 3),despite the existence of dry deposition on the Loess Plateau. Thiscorrelation is statistically significant at the level of 95%. Precipita-tion amount could explain more than 70% (¼r2) of the variation inthe 7Be deposition flux. These linear regressions provide a meanvalue for the mean 7Be concentration in rainfall at the study site of2.71 Bq l�1. This value is very close to the value of 2.6 Bq l�1 derivedfrom the slope of linear regression between 7Be inventories inundisturbed soil and the decayed cumulative precipitation. Thisresult demonstrates that using the 7Be inventories measured inundisturbed soil at different sampling occasions to estimate the 7Bedeposition flux is meaningful.

Compared with the deposition fluxes measured for some othersites with comparable latitude and annual precipitation(Papastefanou and Ioannidou, 1995; Dueñas et al., 2001; González-Gómez et al., 2006), and even for sites with larger annual precipi-tation (Walton and Fried, 1962; Lee et al., 1985; Igarashi et al., 1998;Kim et al., 1999; Hirose et al., 2004; Momoshima et al., 2006), theannual 7Be deposition flux for our study site was greater (Table 2).The greater 7Be deposition flux may reflect the various physicalfactors identified above that will increase 7Be concentrations inrainfall. In addition, 7Be deposition in the form of dustfall, dew andfrost to landscapes is significant at our study site and willcontribute to the deposition flux. Based on 15 observation sites onthe Loess Plateau, Li et al. (2008) reported a mean annual dustdeposition of 254.8 t km�2. Using the mean of the highest 7Be ac-tivities in the dust measured by Kaste et al. (2011) in Owens Valley,California (400 Bq kg�1) and by Belmaker et al. (2011) in the JudeanDesert (543 Bq kg�1) in conjunction with the annual mean dustdeposition rate on the Loess Plateau, the 7Be deposition flux in theform of dustfall can be estimated to be approximately121 Bq m�2 a�1, which is approximately 6.9% of the annual 7Bedeposition flux for the study site. Kaste et al. (2011) indicated thatthe dust that they collected originated frommore local sources andthat 7Be concentrations associated with mobilized dust in arid re-gions is likely to be low. At our study site any dustfall is likely toreflect regional sources and the dust could have been airborne forseveral days and transported over relatively long distances. Thiswould give an opportunity for the dust to scavenge additional 7Befrom the atmosphere. As a result the 7Be concentrations in anydustfall at our study may be greater than those reported by Kasteet al. (2011) and Belmaker et al. (2011). The contribution of dust-fall to the annual 7Be deposition flux at the study site mighttherefore be greater than the value of 6.9% estimated above. Dew inthe semi-arid Loess Plateau is equivalent to approximately 45 mm

Fig. 3. Correlation between 7Be deposition fluxes and individual (a), monthly (b), sampling interval (c) and seasonal (d) precipitation amount.

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184182

of precipitation (Wang and Zhang, 2011). Using the mean 7Beconcentration in rainfall of 2.6 Bq l�1 derived from the slope oflinear regression of between 7Be inventory in undisturbed soil andthe decayed cumulative precipitation, the annual 7Be depositionflux in the form of dew can be estimated to be approximately117 Bq m�2, which is equivalent to approximately 6.7% of theannual 7Be deposition flux at the study site. Frost is also anothersource of 7Be deposition. The frost-free season at our study site isrelatively short (157 d). However, precipitation in the form of frostis not measured. We suggest that the 7Be deposition flux associatedwith frost is likely to be similar to that associated with dew. 7Bedeposition in the form of dustfall, dewand frost should therefore beconsidered as an important process in the removal of 7Be from theatmosphere and to make a significant contribution to the 7Bedeposition flux in the northern Loess Plateau.

7Be deposition inputs in the form of dustfall, dew and frost willexert a significant influence on the estimate of the mean 7Be con-centration in rainfall when using Eq. (1). It will result in an artifi-cially high 7Be concentration in rainfall. However, use of this

Table 2Annual 7Be deposition fluxes at different locations around the world.

Location Latitude Deposition fluxes(Bq m�2 a�1)

Precipit(mm a�

Málaga, Sprain 36�430 412 308Thessaloniki, Greece 40�380 776 430Granada, Sprain 37�100 469 452Ansai, China 36�510 1759 502WESTwood, USA 40�000 717 787Arkansas, USA 38�000 867 1070Tsukuba, Japan 36�030 1322 1362Bermuda 33� 1483 1400Nagasaki, Japan 32�450 1474 1561Kumamoto, Japan 32�480 1590 1900

approach will have less effect on the estimates of 7Be depositionflux. On the contrary, including 7Be deposition input in the form ofdustfall, dew and frost will make the estimate of 7Be deposition fluxmore accurate than that obtained using direct measurements of the7Be activity in rainfall. Furthermore, it is difficult to accuratelyseparate the contributions of wet and dry deposition whenmeasuring the 7Be concentration in rainfall by collecting the rainfallfor individual rainfall event or at regular intervals. The 7Be con-centrations in rainfall reported by some investigators also includethe contribution of 7Be dry deposition (Baskaran et al., 1993;Ródenas et al., 1997; Caillet et al., 2001).

4.2. Temporal variation of the 7Be inventories in undisturbed soil

To eliminate the effect of manual tilling the study plots at thebeginning of the study, the estimates of the daily 7Be inventory inthe undisturbed soil were considered valid from approximately 265days (5 7Be half-lives) after the tilling. After this time, the measured7Be inventory in undisturbed soil varied between 140.2 � 20.4 and

ation1)

Monitoring period Author

1992e1998 Dueñas et al. (2001)1987e1993 Papastefanou and Ioannidou (1995)1995e1998 González-Gómez et al. (2006)2010e2012 This worke Walton and Fried (1962)1979.9e1981e11 Lee et al. (1985).1986e1993 Igarashi et al. (1998)1996e1997 Kim et al. (1999)2000 Hirose et al. (2004)2002e2003 Momoshima et al. (2006)

Fig. 4. 7Be inventories in undisturbed soil between 2010 and 2012. The dates beforeSeptember 30, 2010, were influenced by manual tilling.

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184 183

860.8 � 38.5 Bq m�2 (Table 1) at different sampling occasions. The7Be inventories in undisturbed soil, which reflect a combination ofwet and dry fallout of 7Be and radioactive decay, will vary markedlythrough time. These measured values only reflect the broad trendof 7Be inventories in undisturbed soil during the study period at thestudy site. The true range of the 7Be inventories associated withundisturbed soil through the study period would be nearlyimpossible to document. However, estimates of the daily 7Be in-ventory in undisturbed soil can be calculated by combining theestimates of 7Be activity in precipitation for the individual samplingintervals derived from the measured inventories using Eq. (1) withthe daily precipitation record. The synthesized dataset is presentedin Fig. 4. From September 30, 2010 to December 30, 2012, the dailyinventory values ranged between 89.2 and 941.8 Bq m�2 with amean of 392� 210 Bqm�2, and themonthlymean inventory valuesranged between 102.1 � 10.6 and 856.4 � 36.5 Bq m�2. Althoughsome of the estimates of the daily 7Be inventory in undisturbed soilwere influenced by the manual tilling in 2010 and there were dif-ferences in the highest values and lowest values of the 7Be in-ventories in undisturbed soil between the three years, the intra-annual distribution pattern of the 7Be inventories in undisturbedsoil was obviously unimodal distribution and tended to repeat agenerally regular pattern over a one-year cycle, with the highestvalues in August or September and the lowest in latewinter or earlyspring (Fig. 4). These variations were markedly controlled by the7Be fallout input and its temporal distribution.

The 7Be inventories in undisturbed soil in other regions havebeen estimated to range between 230 and 330 Bq m�2 by Youngand Silker (1974), 300e700 Bq m�2 by Olsen et al. (1985), 140e342 Bq m�2 byWallbrink and Murray (1994) and 181e1401 Bq m�2

by Walling et al. (2009). Kaste et al. (2011) and Belmaker et al.(2011) have also estimated the steady-state 7Be inventories in soilto be 188 Bq m�2 and 12 Bq m�2 in their respective study regions.However, because the 7Be deposition fluxes, the distribution of 7Befallout input through time, the sampling interval and the distri-bution of sampling time during the study period are different forthe different studies, direct comparison with our dataset is oflimited value. Furthermore, Fig. 4 shows that in the north LoessPlateau it is not easy to approach a steady-state equilibrium con-dition between atmospheric 7Be fluxes and radioactive decay inundisturbed soil.

5. Conclusion

We use the measured 7Be inventories in undisturbed soil atdifferent sampling times to estimate the 7Be deposition fluxes andthe daily 7Be inventories in undisturbed soil on the northern LoessPlateau. The calculated mean 7Be concentrations in rainfall, which

quickly decreased with increasing precipitation, are artificially highdue to the 7Be deposition input in the form of dustfall, dew andfrost which provide significant contributions to the 7Be depositionflux in the study region. Inter-annual and intra-annual variations of7Be deposition fluxes were very marked with the maximum insummer and the minimum in winter and were strongly controlledby the precipitation regime. The annual 7Be deposition fluxes weresignificantly greater than those for some other sites with compa-rable latitude and annual precipitation. The daily and monthly 7Beinventories in undisturbed soil exhibited a clear unimodal distri-bution over the year due to the 7Be fallout input and its temporaldistribution. Highest values are obtained in August or Septemberand the lowest ones in late winter or early spring.

Acknowledgements

This work was supported by the National Natural ScienceFoundation of China (No. 40901127, 41171228), the Natural ScienceFoundation of Shannxi Province (No. S2012JC6612), and the specialfunds of Northwest A&F University for the operational costs of basicscientific research (QN2011146, QN2011072). The authors aregrateful to the Ansai Research Station of Soil and Water Conserva-tion, Chinese Academy of Science for providing the experimentalsites and the meteorological data and Prof. D. E. Walling forimproving the English. This paper benefited greatly from thecomments on the original manuscript provided by two anonymousreferees and the editor. Their valuable input is also gratefullyacknowledged.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.atmosenv.2013.05.002.

References

Baskaran, M., Coleman, C.H., Santschi, P.H., 1993. Atmospheric depositional fluxes of7Be and 210Pb at Galveston and College Station, Texas. Journal of GeophysicalResearch 98, 20555e20572.

Baskaran, M., Swarzenski, P.W., 2007. Seasonal variations on the residence timesand partitioning of short-lived radionuclides (234Th, 7Be and 210Pb) and depo-sitional fluxes of 7Be and 210Pb in Tampa Bay, Florida. Marine Chemistry 104,27e42.

Belmaker, R., Lazar, B., Stein, M., Beer, J., 2011. Short residence time and fasttransport of fine detritus in the Judean Desert: clues from 7Be in settled dust.Geophysical Research Letters 38, L16714.

Blake, W.H., Walling, D.E., He, Q., 1999. Fallout beryllium-7 as a tracer in soil erosioninvestigations. Applied Radiation and Isotopes 51, 599e605.

Bleichrodt, J.F., Van Abkoude, E.R., 1963. On the deposition of cosmic-ray-producedberyllium 7. Journal of Geophysical Research 68, 5283.

Bondietti, E.A., Hoffman, F.O., Larsen, I.L., 1984. Air-to-vegetation transfer rates ofnatural submicron aerosols. Journal of Environmental Radioactivity 1, 5e27.

Burchfield, L.A., Akridge, J.D., Kuroda, P.K., 1983. Temporal distributions of radio-strontium isotopes and radon daughters in rainwater during a thunderstorm.Journal of Geophysical Research: Oceans 88, 8579e8584.

Caillet, S., Arpagaus, P., Monna, F., Dominik, J., 2001. Factors controlling 7Be and210Pb atmospheric deposition as revealed by sampling individual rain events inthe region of Geneva, Switzerland. Journal of Environmental Radioactivity 53,241e256.

Cho, Y., Lee, W., Chung, K., Choi, G., Lee, C., 2007. Seasonal variation and activity sizedistribution of 7Be in ambient air. Journal of Radioanalytical and NuclearChemistry 274, 531e538.

Dibb, J.E., 1989. Atmospheric deposition of beryllium 7 in the Chesapeake Bay re-gion. Journal of Geophysical Research 94, 2261e2265.

Doering, C., Akber, R., 2008. Beryllium-7 in near-surface air and deposition atBrisbane, Australia. Journal of Environmental Radioactivity 99, 461e467.

Dueñas, C., Fernández, M.C., Carretero, J., Liger, E., Cañete, S., 2001. Atmosphericdeposition of 7Be at a coastal Mediterranean station. Journal of GeophysicalResearch 106, 34059e34065.

Feely, H.W., Larsen, R.J., Sanderson, C.G., 1989. Factors that cause seasonal variationsin Beryllium-7 concentrations in surface air. Journal of Environmental Radio-activity 9, 223e249.

F. Zhang et al. / Atmospheric Environment 77 (2013) 178e184184

González-Gómez, C., Azahra, M., López-Peñalver, J.J., Camacho-García, A.,Bardouni, T.E., Boukhal, H., 2006. Seasonal variability in 7Be depositional fluxesat Granada, Spain. Applied Radiation and Isotopes 64, 228e234.

Graham, I., Ditchburn, R., Barry, B., 2003. Atmospheric deposition of 7Be and 10Be inNew Zealand rain (1996e98). Geochimica et Cosmochimica Acta 67, 361e373.

Hardy, E.P., Krey, P.W., Volchok, H.L., 1973. Global inventory and distribution offallout plutonium. Nature 241, 444e445.

Harvey, M.J., Matthews, K.M., 1989. 7Be deposition in a high-rainfall area of NewZealand. Journal of Atmospheric Chemistry 8, 299e306.

Heikkilä, U., Beer, J., Alfimov, V., 2008. Beryllium-10 and beryllium-7 in precipitationin Dübendorf (440 m) and at Jungfraujoch (3580 m), Switzerland (1998e2005).Journal of Geophysical Research 113, 11104.

Hirose, K., Honda, T., Yagishita, S., Igarashi, Y., Aoyama, M., 2004. Deposition be-haviors of 210Pb, 7Be and thorium isotopes observed in Tsukuba and Nagasaki,Japan. Atmospheric Environment 38, 6601e6608.

Igarashi, Y., Hirose, K., Otsuji-Hatori, M., 1998. Beryllium-7 deposition and itsrelation to sulfate deposition. Journal of Atmospheric Chemistry 29, 217e231.

Ioannidou, A., Papastefanou, C., 2006. Precipitation scavenging of 7Be and 137Csradionuclides in air. Journal of Environmental Radioactivity 85, 121e136.

Juri Ayub, J., Di Gregorio, D.E., Velasco, H., Huck, H., Rizzotto, M., Lohaiza, F., 2009.Short-term seasonal variability in 7Be wet deposition in a semiarid ecosystemof central Argentina. Journal of Environmental Radioactivity 100, 977e981.

Kaste, J.M., Elmore, A.J., Vest, K.R., Okin, G.S., 2011. Beryllium-7 in soils and vege-tation along an arid precipitation gradient in Owens Valley, California.Geophysical Research Letters 38, L09401.

Kim, G., Alleman, L.Y., Church, T.M., 1999. Atmospheric depositional fluxes of traceelements, 210Pb, and 7Be to the Sargasso Sea. Global Biogeochemical Cycles 13,1183e1192.

Lal, D., Malhotra, P.K., Peters, B., 1958. On the production of radioisotopes in theatmosphere by cosmic radiation and their application to meteorology. Journalof Atmospheric and Terrestrial Physics 12, 306e328.

Lee, L.Y.L., Kwok, R.C.W., Cheung, Y.P., Yu, K.N., 2004. Analyses of airborne 7Beconcentrations in Hong Kong using back-trajectories. Atmospheric Environ-ment 38, 7033e7040.

Lee, S.C., Saleh, A.I., Banavali’, A.D., Jonooby, L., Kurodap, K., 1985. Beryllium-7deposition at Fayetteville, Arkansas, and excess polonium-210 from the 1980eruption of Mount St. Helens. Geochemical Journal 19, 317e322.

Li, J., Dong, Z., Wang, X., 2008. Amount of spring dustfall and its environmentalsignificance in east part of northern China. Journal of Desert Research 28, 195e201 (in Chinese with English abstract).

McNeary, D., Baskaran, M., 2003. Depositional characteristics of 7Be and 210Pb insoutheastern Michigan. Journal of Geophysical Research (Atmospheres) 108,4210.

Momoshima, N., Nishio, S., Kusano, Y., Fukuda, A., Ishimoto, A., 2006. Seasonalvariations of atmospheric210Pb and7Be concentrations at Kumamoto, Japan andtheir removal from the atmosphere as wet and dry depositions. Journal ofRadioanalytical and Nuclear Chemistry 268, 297e304.

Moore, H.E., Poet, S.E., 1976. 210Pb fluxes determined from 210Pb and 226Ra soilprofiles. Journal of Geophysical Research 81, 1056e1058.

Narazaki, Y., Fujitaka, K., Igarashi, S., Ishikawa, Y., Fujinami, N., 2003. Seasonalvariation of 7Be deposition in Japan. Journal of Radioanalytical and NuclearChemistry 256, 489e496.

Nozaki, Y., DeMaster, D.J., Lewis, D.M., Turekian, K.K., 1978. Atmospheric 210Pb fluxesdetermined from soil profiles. Journal of Geophysical Research: Oceans 83,4047e4051.

Olsen, C.R., Larsen, I.L., Lowry, P.D., Cutshall, N.H., Todd, J.F., Wong, G.T.F.,Casey, W.H., 1985. Atmospheric fluxes and marsh-soil inventories of 7Be and210Pb. Journal of Geophysical Research 90, 10487e10496.

Papastefanou, C., Ioannidou, A., 1995. Aerodynamic size association of 7Be inambient aerosols. Journal of Environmental Radioactivity 26, 273e282.

Papastefanou, C., Manolopoulou, M., Stoulos, S., Ioannidou, A., Gerasopoulos, E.,1999. Soil-to-plant transfer of 137Cs, 40K and 7Be. Journal of EnvironmentalRadioactivity 45, 59e65.

Peirson, D.H., 1963. Beryllium 7 in air and rain. Journal of Geophysical Research 68,3831e3832.

Ródenas, C., Gómez, J., Quindós, L.S., Fernández, P.L., Soto, J., 1997. 7Be concentra-tions in air, rain water and soil in Cantabria (Spain). Applied Radiation andIsotopes 48, 545e548.

Schuller, P., Walling, D.E., Iroum, A.S., Castillo, A., 2010. Use of beryllium-7 to studythe effectiveness of woody trash barriers in reducing sediment delivery tostreams after forest clearcutting. Soil and Tillage Research 110, 143e153.

Sepulveda, A., Schuller, P., Walling, D.E., Castillo, A., 2008. Use of 7Be to documentsoil erosion associated with a short period of extreme rainfall. Journal ofEnvironmental Radioactivity 99, 35e49.

Todd, J.F., Wong, G.T.F., Olsen, C.R., Larsen, I.L., 1989. Atmospheric depositionalcharacteristics of beryllium 7 and lead 210 along southeastern Virginia coast.Journal of Geophysical Research 94, 11106e11116.

Wallbrink, P.J., Murray, A.S., 1993. Use of fallout radionuclides as indicators oferosion processes. Hydrological Processes 7, 297e304.

Wallbrink, P.J., Murray, A.S., 1994. Fallout of 7Be in south eastern Australia. Journal ofEnvironmental Radioactivity 25, 213e228.

Wallbrink, P.J., Murray, A.S., 1996. Distribution and variability of 7Be in soils underdifferent surface cover conditions and its potential for describing soil redistri-bution processes. Water Resources Research 32, 467e476.

Walling, D.E., 2013. Beryllium-7: the Cinderella of fallout radionuclide sedimenttracers? Hydrological Processes 27, 830e844.

Walling, D.E., He, Q., Blake, W., 1999. Use of 7Be and 137Cs measurements todocument short- and medium-term rates of water-induced soil erosion onagricultural land. Water Resources Research 35, 3865e3874.

Walling, D.E., Schuller, P., Zhang, Y., Iroumé, A., 2009. Extending the timescale forusing beryllium 7 measurements to document soil redistribution by erosion.Water Resources Research 45, W02418.

Walton, A., Fried, R.E., 1962. The deposition of beryllium 7 and phosphorus 32 inprecipitation at north temperate latitudes. Journal of Geophysical Research 67,5335e5340.

Wang, S., Zhang, Q., 2011. Atmospheric physical characteristics of dew formation insemi-arid in loess plateau. Acta Physica Sinica 60, 059203 (in Chinese withEnglish abstract).

Wilson, C.G., Matisoff, G., Whiting, P.J., 2003. Short-term erosion rates from a 7Beinventory balance. Earth Surface Processes and Landforms 28, 967e977.

Yamamoto, M., Sakaguchi, A., Sasaki, K., Hirose, K., Igarashi, Y., Kim, C.K., 2006.Seasonal and spatial variation of atmospheric 210Pb and 7Be deposition: featuresof the Japan Sea side of Japan. Journal of Environmental Radioactivity 86, 110e131.

Yang, M.Y., Walling, D.E., Tian, J.L., Liu, P.L., 2006. Partitioning the contributions ofsheet and rill erosion using beryllium-7 and cesium-137. Soil Science Society ofAmerica Journal 70, 1579e1590.

Young, J.A., Silker, W.B., 1974. The determination of air-sea exchange and oceanicmixing rates using 7Be during the Bomex experiment. Journal of GeophysicalResearch 79, 4481e4489.

Yuan, H., Hou, J., Du, J., Li, M., 2007. The climate characteristics of sudden localstorms in loess plateau. Journal of Catastrophology 22, 101e104 (in Chinesewith English abstract).

Zhang, F., Yang, M., Zhang, B., 2011. Beryllium-7 activity concentration in plants onthe loess Plateau, China. Journal of Radioanalytical and Nuclear Chemistry 289,353e359.