Research ArticleEARTH SCIENCES

Continental igneous rock composition: A major control of past global chemical weathering

See allHide authors and affiliations

Science Advances  08 Mar 2017:
Vol. 3, no. 3, e1602183
DOI: 10.1126/sciadv.1602183
  • Fig. 1 Plot of εHf of detrital zircon grains versus U-Pb date.

    Black dots represent the nonresampled εHf data from a screened subset of database S1. Moving median and quartiles at 1-My increments for the resampled database S1 (red line and green lines) and for the nonresampled database S1 (black line). The resampled data set includes both the debiasing resampling and the uncertainty propagation resampling (see Materials and Methods and the Supplementary Materials). The uncertainty propagation resampling uses U-Pb date (±2%) and εHf (±0.62).

  • Fig. 2 Plot of εHf values of inherited zircon grains versus whole-rock εSr values for Phanerozoic igneous rocks.

    (A) Nonscreened database S2. (B) Screened database S2. Screening procedure removed rock with crystallization age older than 300 Ma and with 87Rb/86Sr superior to 40. Regression models fitted to the data (red lines) with 95% confidence interval (blue lines). RMSE, root mean square error.

  • Fig. 3 Time series analysis between N(87Sr/86Sr)seawater (blue) and the median (red) and mean (green) (87Sr/86Sr)i-zig, both decomposed to their slow-varying trend and fast-varying residues.

    N(87Sr/86Sr)seawater ratio (A) and (87Sr/86Sr)i-zig ratio (B) as in Fig. 1. (C) CCF. The first column corresponds to the original data at a 1-My interpolated sampling interval, the second column corresponds to the signal filtered for 700-My or longer periodicity, and the third column corresponds to the residual filtered between 30- and 700-My periodicity. The range of lags tested was set at ±200 My to include slowly exhumed plutonic rocks. A negative lag corresponds to the N(87Sr/86Sr)seawater ratio lagging the (87Sr/86Sr)i-zig ratio.

  • Fig. 4 Interpretation of the marine Sr budget.

    Maximum correlation coefficient through time (A) between the N(87Sr/86Sr)seawater ratio (B) and the (87Sr/86Sr)i-zig ratio (green) (C). Curves (B) and (C) are obtained by summing the slow-varying trend and fast-varying residues of the mean value obtained in Fig. 3 (A and B, respectively). The CCF is calculated for a 150-My time window moving in 10-My increments, with lags ranging between 0 and 20 My. (D) Frequency of collisional orogenies (46), major LIPs (23, 314), and major glaciations. Shaded blue bars represent ocean opening stages.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/3/e1602183/DC1

    fig. S1. Map of sediment locations.

    fig. S2. Sampling bias: Histogram of zircon grain count per sediment from the screened data set (24,377 individual zircon grains from 470 individual sediments).

    fig. S3. Geographic bias and geographic weights: Map of sediment sample locations for the screened data set with associated geographic resampling weights.

    fig. S4. Geographic debiasing validation.

    fig. S5. Resampled εHf using combined weights (red) superimposed on nonresampled εHf (black) from detrital zircons from Cenozoic and Mesozoic sediment data compiled in database S1.

    fig. S6. Map of sediment sample locations with associated final resampling weights combining geographic and sampling weights.

    fig. S7. Resampling weights visualization.

    fig. S8. Sensitivity of the smoothed εHf trend to the uncertainty in U-Pb date.

    fig. S9. Sensitivity of the smoothed εHf trend to the uncertainty in εHf values.

    fig. S10. Evolution of the (87Sr/86Sr)seawater ratio in the last 1000 My from the analysis of 87Sr/86Sr ratio of carbonates.

    fig. S11. Sensitivity of time series analysis to debiasing procedures.

    fig. S12. Sensitivity of time series analysis to data screening.

    fig. S13. Sensitivity of time series analysis to N(87Sr/86Sr)seawater curve parameterization by assuming a linear increase in Rb/Sr ratio of the crust in the last 1000 My.

    fig. S14. Sensitivity of time series analysis to N(87Sr/86Sr)seawater curve parameterization by assuming a linear decrease in Rb/Sr ratio of the crust in the last 1000 My.

    database S1. Excel data table with worldwide U-Pb and Hf isotopes in detrital zircons compiled from the literature.

    database S2. Excel data table with 87Sr/86Sr and 143Nd/144Nd ratio data from whole igneous rock and U-Pb and Hf isotope data from their hosted zircons compiled from the literature.

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. Map of sediment locations.
    • fig. S2. Sampling bias: Histogram of zircon grain count per sediment from the screened data set (24,377 individual zircon grains from 470 individual sediments).
    • fig. S3. Geographic bias and geographic weights: Map of sediment sample locations for the screened data set with associated geographic resampling weights.
    • fig. S4. Geographic debiasing validation.
    • fig. S5. Resampled εHf using combined weights (red) superimposed on nonresampled εHf (black) from detrital zircons from Cenozoic and Mesozoic sediment data compiled in database S1.
    • fig. S6. Map of sediment sample locations with associated final resampling weights combining geographic and sampling weights.
    • fig. S7. Resampling weights visualization.
    • fig. S8. Sensitivity of the smoothed εHf trend to the uncertainty in U-Pb date.
    • fig. S9. Sensitivity of the smoothed εHf trend to the uncertainty in εHf values.
    • fig. S10. Evolution of the (87Sr/86Sr)seawater ratio in the last 1000 My from the analysis of 87Sr/86Sr ratio of carbonates.
    • fig. S11. Sensitivity of time series analysis to debiasing procedures.
    • fig. S12. Sensitivity of time series analysis to data screening.
    • fig. S13. Sensitivity of time series analysis to N(87Sr/86Sr)seawater curve parameterization by assuming a linear increase in Rb/Sr ratio of the crust in the last 1000 My.
    • fig. S14. Sensitivity of time series analysis to N(87Sr/86Sr)seawater curve parameterization by assuming a linear decrease in Rb/Sr ratio of the crust in the last 1000 My.
    • Legends for database S1 and S2

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • database S1. Excel data table with worldwide U-Pb and Hf isotopes in detrital zircons compiled from the literature.
    • database S2. Excel data table with 87Sr/86Sr and 143Nd/144Nd ratio data from whole igneous rock and U-Pb and Hf isotope data from their hosted zircons compiled from the literature.

    Files in this Data Supplement:

Navigate This Article