SEARCH

SEARCH BY CITATION

Keywords:

  • Solar radiation;
  • climate change;
  • urban effects

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[1] Daily values of global radiation Eg[DOWNWARDS ARROW] measured in a network of six sites within the greater Tel Aviv region indicated a maximum urban dimming effect of 7% with a similar day-to-day range in differences during the week that was significantly negatively related to the number of vehicles using the roads. Interannual variability of Eg[DOWNWARDS ARROW] differed between three regions of Israel having very diverse population densities, both in their overall trends and in their rates of change during periods of dimming and brightening. During the 50 years of measurement the overall trends in all regions were negative; the rates of dimming were related to the logarithm of the mean population densities with the greatest net rate of dimming, −3.4% per decade, measured at the downwind edge of the densely populated Tel Aviv conurbation. Rates of change in the dimming and brightening periods were not related to the rates of change in population; maximum rates of both dimming and brightening were measured at the Dead Sea, a desert region with a low and unchanging population. Trends in maximum, minimum, mean, and diurnal range air temperatures were significant and differed between regions, but these differences were not clearly related to those in Eg[DOWNWARDS ARROW] or in population density. The results indicate that in Israel the maximum urban effects on Eg[DOWNWARDS ARROW] were insufficient to explain the large interannual variability measured during the last 50 years and that population density was not a robust proxy for global radiation change.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[2] Analysis of interannual changes in solar radiation measured at the Earth's surface Eg[DOWNWARDS ARROW] during the last 50 years shows significant temporal variations; predominantly reductions or “dimming” during the third quarter of the last century followed by a partial recovery or “brightening” in the last quarter. The spatial extent of these temporal changes is important because of the major role Eg[DOWNWARDS ARROW] plays in determining the heat, water and carbon balances at the Earth's surface and the fact that the total planetary radiative forcing due to this phenomenon depends on how changes reported at individual sites are integrated spatially.

[3] Studies of these spatial variations showed that the greatest temporal changes occurred in the midlatitudes of the Northern Hemisphere which contain the major centers of population and industrial activity [Gilgen et al., 1998]. Stanhill and Cohen [2001] reported a highly significant spatial relationship between 1114 annual values of Eg[DOWNWARDS ARROW] measured at 854 sites during the years 1958, 1965, 1975, 1985 and 1992 and the anthropogenic carbon emissions from the 1° cells surrounding the measurement sites. No significant relationship between Eg[DOWNWARDS ARROW] and the population density of the cells was found.

[4] Many studies of the influence of individual cities on solar radiation have been made on the basis of comparisons between measurements in the city center and those from the outskirts. Results reported from Quebec [East, 1968], Los Angeles [Peterson et al., 1978], Moscow [Abakumova, 1980], St. Louis [Peterson and Stoffel, 1980], Cairo [Robaa, 2006] and Taipei [Chou et al., 2006] show values of urban dimming ranging between 3% and 18%.

[5] The relationship between the rate of dimming and the size of the population surrounding the measurement site was studied by Nunez [1993] using satellite derived calculations and ground-based measurements at 15 Asian locations, the data showed a highly significant logarithmic relationship between Eg[DOWNWARDS ARROW] depletion and population size. More recently, Alpert and Kishcha [2008] quantified this relationship for three latitudinal zones, comparing rates of global dimming during the 1964–1989 period and the population density in the 1° cell surrounding the pyranometer sites in the year 2000 and concluded that dimming was confined to the most densely populated one third of the Earth's land surface; that is, less than 10% of the Globe's surface was affected.

[6] Stanhill and Moreshet [1994], using measurements in Israel, confirmed the effect of human activity on solar dimming and indicated that road traffic played a major role, presumably through its aerosol emissions. On an annual basis a highly significant correlation was established between the decrease in Eg[DOWNWARDS ARROW] and the increase in the number of vehicles passing within 1 km of the Israel Meteorological Service's observatory at Bet Dagan in the densely populated central coastal plain. The causal nature of this correlation and the magnitude of the effect was demonstrated by the 10% increase in Eg[DOWNWARDS ARROW] measured on the Day of Atonement, a day without road traffic.

[7] Solar brightening (the slowing, cessation or reversal of solar dimming), reported by Wild et al. [2005] as occurring some 25 years ago, has also been explained as resulting from reductions in anthropogenic emissions of aerosols [Streets et al., 2006; Wild et al., 2007].

[8] The importance of the question raised in the title of this study lies in the consequences of the large changes in solar radiation measured in the last half century for the Earth's hydrological and carbon cycles and its air temperature. Are the effects of changes in solar irradiance at the Earth's land and water surface significant on a global scale or are they confined to highly populated urban land surfaces? And if dimming is caused by aerosols which can move large distances from their anthropogenic sources according to atmospheric circulation patterns [Ramanathan et al., 2007] can population density serve as a robust proxy for analyzing the areal extent of solar changes at the Earth's surface?

[9] In this study we address these questions by examining the spatial differences in Eg[DOWNWARDS ARROW] measured within the Tel Aviv conurbation and by comparing the interannual trends of Eg[DOWNWARDS ARROW] and maximum and minimum air temperatures measured in Israel during the last half century in three regions with very different population densities.

2. Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

2.1. Measurements

[10] The spatial variation of Eg[DOWNWARDS ARROW] in the central coastal plain region of Tel Aviv was measured during 1991 and 1992 with a small network of six pyranometer stations; all instruments were calibrated before exposure by reference to the pyranometer used at the Israel Meteorological Services observatory at Bet Dagan, which was taken as the standard. Random instrumental differences for daily totals averaged 0.33% and never exceeded 1.4 W m−2. The six sites were selected to cover the range of exposure to urban pollution in Israel's densely populated central coastal plain and varied from a cliff top overlooking the Mediterranean Sea, southwest, i.e., upwind of the city, to a rooftop site overlooking Tel Aviv's central bus station (Figure 1a).

image

Figure 1. Sites of global radiation and air temperature measurements in Israel. (a) Greater Tel Aviv pyranometer network, 1991–1992; letters represent the sites whose coordinates are listed in Table 1. The horizontal and vertical lines indicate 32°N and 35°E, respectively. (b) Sites of annual values of Eg[DOWNWARDS ARROW] (open circles) and air temperature (black dots) measurements. (c) Wind direction frequency measured at Bet Dagan during daylight hours, 1964–1983; the length of the vectors represents their percent frequency, the width represents the wind force, and the encircled number at the center of the wind rose gives the percentage frequency of calms.

Download figure to PowerPoint

[11] The temporal as well as spatial variation of Eg[DOWNWARDS ARROW] in Israel was analyzed using a database consisting of 341 mean annual values of Eg[DOWNWARDS ARROW] measured between 1954 and 2007 at 23 sites. Each value represents a complete year of data measured with a calibrated thermopile pyranometer in which both the data and instrumentation were quality controlled. This database was not homogeneous; the number of measurement sites increased significantly with time, on an average by three sites per decade. By contrast, spatial variation, quantified as the standard error of intersite variation in annual Eg[DOWNWARDS ARROW], was unchanged during the period; the decrease of 0.8 W m−2 per decade was not statistically significant and was only one eighth of the average spatial variation, 6.4 W m−2, equivalent to 2.7% of the mean annual flux, 232.5 W m−2. Thus in Israel the spatial variation of Eg[DOWNWARDS ARROW] was of similar size as the accuracy of measurement to be expected for annual values [World Meteorological Organization, 1997].

[12] Annual mean values of mean temperature Ta, maximum temperature Tmax, minimum temperature Tmin, and diurnal temperature range DTR, were obtained from the Israel Meteorological Service climate stations at which, or close to the sites where Eg[DOWNWARDS ARROW] was measured; in the case of the Dead Sea region, the station was Sedom. For the pooled national series, the values used were the average of five stations representing the major climate regions, the station sites are shown in Figure 1b.

[13] The four Israeli solar radiation and air temperature time series studied represent regions with very different population densities, ranging from a maximum of 6000 per km2 in the Tel Aviv area to a minimum of 1 per km2 in the Dead Sea region, 100 km downwind from and 400 m lower than Tel Aviv.

[14] The Eg[DOWNWARDS ARROW] series examined were: (1) the pooled series of all Israeli measurement sites having an average altitude of 166 m, (2) the pooled series from Jerusalem in the central hill region consisting of 53 years of data from three nearby urban sites whose average altitude was 785 m, (3) the pooled series from the central coastal plain consisting of a 46-year series from two nearby sites both downwind of Tel Aviv, Israel's major conurbation, with an average altitude of 35 m, and (4) the pooled series from the sparsely populated Dead Sea region of the Jordan Rift Valley consisting of 49 years of data from five sites whose mean weighted altitude was 390 m below sea level. Locations of the pyranometer sites are shown in Figure 1b and a wind rose representative of the region in Figure 1c.

[15] Annual values of population densities relevant to the series were obtained from the Israel Central Bureau of Statistics, using both its Annual Statistical Abstract and data available online from its Information Center (cbs@cbs.gov.il). Population densities, in units of persons per km2, represent the number of persons (including an estimate of tourists) divided by the total areas of the four regions; these areas were 22145, 1650, 652 and 171 km2 respectively for Israel, Dead Sea, Jerusalem and Tel Aviv. In the case of the Dead Sea region the population density was so small and had changed so little that the current value of one person per km2 was taken as constant.

[16] We note that these population densities are based on municipal and district borders and differ from those calculated for 1° cells of the Earth surface, which represent population densities averaged over much larger areas, except in the case of Israel as a whole whose total area is equivalent to approximately three 1° cells. The area of the Dead Sea region equals one fifth, that of Jerusalem 0.07 and Tel Aviv 0.02 of a 1° cell.

2.2. Statistical Analysis

[17] All four time series of Eg[DOWNWARDS ARROW] had statistically significant (P < 0.05) trends as shown by the nonparametric Mann-Kendall test recommended by the World Meteorological Organization for the analysis of time series of climate data [Sneyers, 1990]. The time trends in air temperatures were also significant as shown by this nonparametric test. Having established their significance the magnitude of the time trends were determined using standard parametric methods of linear and quadratic regression analysis.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

3.1. Spatial Variation

[18] The mean daily values of Eg[DOWNWARDS ARROW] measured in the greater Tel Aviv station network, normalized to values measured at the Israel Meteorological Service's Bet Dagan Observatory on the same day, are presented as ratios in Table 1 together with their standard deviations. The difference between sites with the maximum and minimum values was 7%, with the maximum ratio of 1.062 ± 0.088 measured at a cliff site on the sea coast upwind of the city and the minimum value of 0.988 ± 0.093 measured at a roof top, road side site 2 km downwind from the reference observatory site. A further 7 km downwind at a rural site Eg[DOWNWARDS ARROW] increased to 1.04 ± 0.03.

Table 1. Spatial Variation of Daily Global Radiation in Israel's Central Coastal Plain
 Site A Bet Dagan IMS ObservatorySite B Bet Dagan ARO RoofSite C Tel Aviv Town Hall RoofSite D Tel Aviv Bus Station RoofSite E Palmahim Dining Hall RoofSite F Rehovot Weizmann Institute
Coordinates
Longitude, °E34.4934.4934.4834.4834.4334.49
Latitude, °N32.0031.5932.0532.0231.5631.54
Elevation, m303815152045
Pyranometer height, m1.5185025151.5
Season (days of year)1–365289–3163–155207–28045–160180–280
67–155
Number of days36518365678492
 
Ratios of Eg[DOWNWARDS ARROW] to Values Measured at Reference Site A
Mean1.0000.9881.0421.0411.0621.039
Standard deviation 0.0310.0930.0890.0400.088
 
Ratio of Daily Means of Eg[DOWNWARDS ARROW] to Values Measured at Reference Site A
Mean1.0000.0961.0451.0381.0541.037

[19] Daily values of Eg[DOWNWARDS ARROW] measured for each week day were normalized to the weekly mean and then averaged for the six network sites (Y); the ratios were negatively and highly significantly correlated (R2 = 0.890) with the average, normalized number of vehicles using the roads for that day of the week (X). This value varied between a maximum of 1.09 on Sunday, the first working day of the week, to a minimum of 0.78 on Saturday, the weekend day of rest according to the 1992 Statistical Abstract of Israel 43 (cbs@cbs.gov.il). The linear equation Y = −0.206 (±0.032) X + 1.206 (±0.010) indicated a decrease in Eg[DOWNWARDS ARROW] of 0.2% for each 1% increase in road traffic.

3.2. Temporal Variation

[20] Mean annual values of Eg[DOWNWARDS ARROW] for the four series are shown in Figure 2 with lines representing quadratic equations fitted to the data. The parameters of both quadratic and linear functions of year of measurement are presented in Table 2. The highest average annual value of solar irradiance was recorded for the Jerusalem series, 2.5% above that of the pooled Israel national series. The lowest irradiance was recorded in the Central Coastal Plain series, with an average annual value 2.7% less than that of the Dead Sea series (Table 2).

image

Figure 2. Time series of mean annual global radiation measured in Israel in W m −2. Lines represent fitted quadratic relationships listed in Table 2. Data are from all Israel, 1954–2007 (diamonds, blue line), Jerusalem, 1954–2007 (squares, red line), Tel Aviv, 1957–2007 (asterisks, black line) and Dead Sea, 1960–2007 (triangles, green line).

Download figure to PowerPoint

Table 2. Linear and Quadratic Time Trends in Annual Global Radiation in Israela
 Israel All DataJerusalemBet Dagan-LodDead Sea
  • a

    P values: *, P < 0.05; ***, P < 0.001.

Years1954–20071954–20071957–20071960–2007
Mean Irradiance, W m−2232238219225
 
Linear trend    
Slope W m−2 a−1−0.575−0.710−0.738−0.276
Standard error of slope0.0650.0840.0980.136
As % per decade−2.48−2.98−3.28−1.26
R20.59***0.60***0.55***0.19*
 
Quadratic trend Eg[DOWNWARDS ARROW] = a · (year2) + b · (year) + c
      a0.02140.011010.031500.03695
Standard error of a0.00360.006330.00580.0058
      b−85.384−44.310−125.63−146.9
      Standard error of b14.40425.09122.9223.04
      c8534744821125452146217
      R20.75***0.62 N.S.0.73***0.63***

[21] Overall linear trends during the last 50 years were negative for all four series indicating net dimming. The greatest reduction in Eg[DOWNWARDS ARROW] both absolutely and relatively occurred in the central coastal plain followed by Jerusalem and then the pooled national series with the least reduction in the Dead Sea series (Figure 2). The difference between the slope of the Dead Sea series and those of the other three series was statistically significantly (P < 0.05).

[22] The use of quadratic equations to analyze separately time trends in periods of dimming and brightening explained more interannual variation in Eg[DOWNWARDS ARROW] than the linear relationships did for all four time series. The maximum rates of dimming (represented in the quadratic equations by the negative linear coefficient b), and of brightening (represented by the positive quadratic coefficient a) occurred in both cases at the Dead Sea, followed by the coastal plain and the pooled national series. Minimum rates of change, both of dimming and brightening, occurred in the Jerusalem series. Differences in the rates of both dimming and brightening were significant (P < 0.05) between two pairs of sites; the Dead Sea and Coastal plain series and the Jerusalem and the pooled national series (Table 2).

[23] Highly significant but different relationships were found between annual values of Eg[DOWNWARDS ARROW] and population densities: the relationships were negative in the early, dimming period followed by less marked positive associations in the later, brightening period. The saturation population density at which the relationship changed sign varied between 200 persons per km2 in the pooled national series and 6000 persons per km2 in the greater Tel Aviv series (Figure 3).

image

Figure 3. Relationships between global radiation and population density in Israel, annual values during dimming (solid diamonds) and brightening (open squares) periods. (a) All Israel mean values, 1954–2007. (b) Tel Aviv, 1956–2007. (c) Jerusalem, 1954–2007.

Download figure to PowerPoint

[24] Parameters of the trends in annual values of air temperatures, listed in Table 3, show that the largest overall temperature changes occurred in the Bet Dagan series, with lesser changes in the pooled national series and the least change in the Jerusalem series. This regional order varied for the four different indices of air temperature change examined.

Table 3. Linear Time Trends in Annual Air Temperatures in Israela
 Israel Pooled DataJerusalemBet DaganSedom
  • a

    Temperatures given in °C.

Year1967–20031964–20041964–20041960–2002
Mean maximum24.5521.5025.2530.28
   Slope, °C per year0.0270.031−0.0020.004
   Standard error of slope0.0070.0090.0070.006
   As % per decade1.091.45−0.010.09
   R20.3020.2550.0020.012
 
Mean minimum14.9313.5213.6621.26
   Slope, °C per year0.0330.023−0.0360.041
   Standard error of slope0.0060.0060.0120.009
   As % per decade2.201.733.791.94
   R20.4370.2910.200.441
 
Mean average19.7717.5319.4725.77
   Slope, °C per year0.0300.027−0.0190.02
   Standard error of slope0.0060.0070.0090.007
   As % per decade1.511.541.020.79
   R20.3810.2880.1220.263
 
Mean diurnal range9.678.0311.639.02
   Slope, °C per year−0.0060.0070.054−0.033
   Standard error of slope0.0030.0050.0060.007
   As % per decade−0.010.0910.415−3.66
   R20.1020.0510.7040.447

[25] Interannual variations in the four indices of air temperature were not simply related to those in global radiation. Significant correlations were only found in two of the series, those of the Central Coastal Plain and at the Dead Sea and for three of the four temperature indices: average, minimum and diurnal temperature range. The five, out of 16 possible statistically significant correlations are shown in Figures 4a and 4b with fitted linear relationships.

image

Figure 4. Statistically significant relationships between mean annual values of air temperature, in °C, and global radiation, in W m−2. Symbols indicate average (diamond) and minimum (triangle) temperatures and daily temperature range (crosses). (a) Tel Aviv series, 1956–2007: Ta = −0.019, Eg[DOWNWARDS ARROW] + 25.26, P < 0.05; Tmin = −0.036, Eg[DOWNWARDS ARROW] + 21.39, P < 0.01; DTR = 0.034, Eg[DOWNWARDS ARROW] + 4.33, P < 0.001. (b) Dead Sea series, 1960–2007: Tmin = −0.070, Eg[DOWNWARDS ARROW] + 37.88, P < 0.05; DTR = 0.070, Eg[DOWNWARDS ARROW] −7.50, P < 0.01.

Download figure to PowerPoint

4. Discussion and Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[26] The maximum between-site spatial variation in global radiation over the greater Tel Aviv area was seven percent, with the lowest Eg[DOWNWARDS ARROW] values measured downwind rather than within the city center. Averaged over the urban sites the mean reduction, 4%, falls within the range of values of urban dimming previously reported in the literature. The seven percent range in day of the week values of Eg[DOWNWARDS ARROW] measured in the Tel Aviv network showed the mean daily values to be linearly and negatively related to the mean amounts of road traffic on each week day. The maximum size of these two urban dimming effects measured in 1991 and 1992, was less than half the 16.9% overall long-term reduction measured at Bet Dagan during 50 years, indicating the importance of other, nonurban causal factors. Changes in cloud cover can be excluded as such a factor as observations at Bet Dagan showed cloud cover to have slightly decreased over the period [Stanhill and Moreshet, 1992]. This was also the case at Sedom.

[27] The overall linear rates of solar dimming for the four time series measured in Israel during the last 50 years listed in Table 2 (Y, W m−2 a −1), were negatively related to the logarithm of the mean population density of the four regions (X, persons km−2). The relationship Y = −0.131(±0.016) log10X – 0.286(±0.043), was highly significant (R2 = 0.969, P < 0.01). Although only based on four points, the negative offset of this relationship indicates that a reduction in Eg[DOWNWARDS ARROW] would have occurred even if the region was uninhabited. The above relationship differs from the linear spatial one derived by Alpert and Kishcha from their analysis of the data from the 40°S–40°N latitudinal zone [Alpert and Kishcha, 2008] but resembles that established by Nunez for 19 sites in the Pacific region [Nunez, 1993].

[28] The more meaningful temporal rather than spatial relationships of Eg[DOWNWARDS ARROW] to population densities shown in Figure 3, shows no evidence for a simple causal relationship common to the three Israel series. Neither were common relationships with population densities evident when the data was separated into negative relationships during the earlier dimming period and the weakly positive relationships occurring during the more recent brightening period.

[29] The fact that the greatest changes in Eg[DOWNWARDS ARROW] occurred in the very sparsely populated Dead Sea desert region argues against a purely local anthropogenic cause of solar forcing at the Earth's surface. Changes in the sparse cloud cover of this desert region can be excluded as the nonanthropogenic cause of solar changes as observations at Sedom show a very small overall decrease in cloud cover of 0.12 oktas per decade, which did not coincide with changes in Eg[DOWNWARDS ARROW].

[30] By contrast the aerosol load at the Dead Sea, enhanced by the extreme depth of the atmospheric column, is large as shown by direct measurement [Levin et al., 2005] and indicated by the high fraction of diffuse radiation [Stanhill, 1987] and very high values of circumsolar radiation and sky anisotropy measured at the Dead Sea [Stanhill, 1985].The presence of a high aerosol load at Sedom is also shown by the low value of the cloudless clearness index (global as a fraction of extraterrestrial radiation), derived from the relationship between monthly values of the ratio of actual to possible sunshine duration and the clearness index: At Sedom this relationship extrapolated to maximum sunshine duration yielded a clearness index of 0.70 ± 0.03 for cloudless sky conditions, significantly below the average reported for cloudless conditions [Linacre, 1992]. The reason for the high aerosol load is probably topographical; the unique depth of the Dead Sea Rift valley, the lowest point on the Earth's surface, trapping aerosols emitted from upwind sources which include Israel's densely populated coastal plain, the Mediterranean Sea and further upwind, southeast Europe.

[31] That the sources of the aerosols trapped at the Dead Sea are not exclusively anthropogenic can be seen from the reports of haze in the region documented 150 years ago in the accounts and illustrations of early explorers [Lynch, 1849; de Saulcy, 1853]. A recent study of the haze particles at the Dead Sea showed that a large fraction are sulfates which can be of either anthropogenic, industrial or of nonanthropogenic, biological origin [Levin et al., 2005]. While industrial emissions have certainly increased in the period under consideration the biological component, originating from the dimethyl sulfate gas emitted from biological activity at the sea surface, has probably decreased; it is now almost zero from the salt saturated Dead Sea while emissions from the Eastern Mediterranean have probably declined owing to the sea's increased salinity and pollution.

[32] The failure of population density to serve as a robust proxy for spatial and temporal changes in global radiation demonstrated in this study reinforces previous reports of significant time trends in Eg[DOWNWARDS ARROW] measured at seven very sparsely inhabited sites [Stanhill and Moreshet, 1994] as well as within the Arctic [Stanhill, 1995] and Antarctic [Stanhill and Cohen, 1997] polar circles. Together they confirm studies documenting the importance of the long-range transport of aerosols in determining radiation forcing at the Earth's surface [Ramanathan et al., 2007].

[33] This conclusion is supported by the evidence presented demonstrating that maximum rates of radiation change occurred in the sparsely inhabited Dead Sea region while local urban effects were insufficient to explain the solar dimming measured during the last 50 years even in Israel's most densely populated region.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

[34] We thank D. Faiman, A. Furspan, I. Gertman, A. Ianetz, A. Kudish, V. Lyubansky, S. Oram, A. Rimmer, and Y. Tsipris, who provided much of the solar radiation, population density, and air temperature data analyzed in this study, and the editor and reviewers for their helpful comments on an earlier draft.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion and Conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
jgrd15581-sup-0001-t01.txtplain text document1KTab-delimited Table 1.
jgrd15581-sup-0002-t02.txtplain text document1KTab-delimited Table 2.
jgrd15581-sup-0003-t03.txtplain text document1KTab-delimited Table 3.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.