Research ArticleGEOCHEMISTRY

The prevalence of kilometer-scale heterogeneity in the source region of MORB upper mantle

See allHide authors and affiliations

Science Advances  22 Nov 2017:
Vol. 3, no. 11, e1701872
DOI: 10.1126/sciadv.1701872
  • Fig. 1 Observations and the model for melting of an upwelling and chemically heterogeneous mantle beneath mid-ocean spreading centers.

    (A) Histogram of 143Nd/144Nd in abyssal peridotites (AP) from Mid-Atlantic Ridge (37), Mid-Cayman Rise (37), and Southwest Indian Ridge (4, 6, 37); selected residual abyssal peridotites; and spatially associated MORB (see data compilation). (B) Cartoon illustrating the double-porosity ridge model for melting and melt migration in a heterogeneous mantle beneath a mid-ocean spreading center. The ridge model consists of bundles of double-porosity columns. Enriched heterogeneities are sketched as orange blobs in one column.

  • Fig. 2 Stretching and smearing of an enriched heterogeneity in an upwelling melting column.

    Spatial variations of (A and B) 143Nd/144Nd and (C and D) 176Hf/177Hf at three selected times are shown by blue (residue) and red lines (matrix melt). The initial size of the heterogeneity is 4.4 km (A and C) or 1.1 km (B and D). The initial Nd and Hf in the heterogeneity are five times more enriched than in the depleted mantle. Pairs of facing triangles are material markers that track the locations of major element particles in the residue at the selected times. The small reduction in the marker distance is due to partial melting (fig. S3B). The degree of melting is linearly proportional to the vertical distance above the solidus (z = 0). The dashed lines mark the isotope ratio of the enriched endmember (see movie S1).

  • Fig. 3 The effect of the size and trace element abundance of mantle heterogeneity on the variations of isotope ratios in the pooled melt, erupted melt, and residues.

    (A to D) The enrichment factor (EMX), proportion of EM in the source, and the initial size of enriched heterogeneities are listed above each panel. Gray outlines at the lower left and upper right corner are the range of isotope ratios of EM and DMM, respectively. Gray dots are global MORB data (39). The pentagram is the average of global MORB. Contours show the probability density function of the composition of the erupted melt predicted by the model. Histograms show the marginal distribution of 143Nd/143Nd and 176Hf/177Hf in the source, residue, pooled melt, and erupted melt. Dashed lines on histograms mark the concentration average of the source.

  • Fig. 4 Size-sensitive mixing for the residue and the melt.

    The 143Nd/144Nd of the residue (A) and the erupted melt (B) as a function of the enrichment strength of the heterogeneity (L*EMX). Colored dots are the 143Nd/144Nd at the peak of marginal distribution in the residue or the erupted melt. Vertical bars are the half-height width of the marginal distribution of 143Nd/144Nd. The thick gray lines are the concentration average of the source. Dashed lines in (B) are the range of MORB. If the enrichment strength is smaller than 20 km [arrow in (A)], the peak of marginal distribution in the residue would converge to the source average, and no offset would be observed. If the enrichment strength is larger than 60 km [arrow in (B)], the range of isotopic compositions in the erupted melt would be larger than that observed in the MORB.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/11/e1701872/DC1

    table S1. Errors and numerical orders.

    table S2. Physical parameters in working equations (Eqs. 33 to 35 and 41 to 43).

    fig. S1. Histograms of Nd isotopes in abyssal peridotites, residual abyssal peridotites, and local MORB at individual locations.

    fig. S2. Lack of correlation between the REE concentration and 143Nd/144Nd of clinopyroxenes in abyssal peridotites.

    fig. S3. Spatial variations of porosity, residue, and matrix melt velocities in the 1D upwelling melting column.

    fig. S4. The 176Hf/177Hf of the residue and the erupted melt as a function of the enrichment strength of the heterogeneity (L*EMX).

    fig. S5. The effect of chemical exchange rate and melt extraction on 143Nd/144Nd of the residue and the erupted melt.

    fig. S6. The composition of melt collection, residues, pooled melt, and sources in the Hf-Nd isotope correlation diagram.

    fig. S7. A comparison of the predicted compositions of the residue and erupted melt in a case with Hf offset but without Nd offset.

    movie S1. Spatial and temporal variations of 143Nd/144Nd and 176Hf/177Hf in the melting column.

    References (4153)

  • Supplementary Materials

    This PDF file includes:

    • table S1. Errors and numerical orders.
    • table S2. Physical parameters in working equations (Eqs. 33 to 35 and 41 to 43).
    • fig. S1. Histograms of Nd isotopes in abyssal peridotites, residual abyssal peridotites, and local MORB at individual locations.
    • fig. S2. Lack of correlation between the REE concentration and 143Nd/144Nd of clinopyroxenes in abyssal peridotites.
    • fig. S3. Spatial variations of porosity, residue, and matrix melt velocities in the 1D upwelling melting column.
    • fig. S4. The 176Hf/177Hf of the residue and the erupted melt as a function of the enrichment strength of the heterogeneity (L*EMX).
    • fig. S5. The effect of chemical exchange rate and melt extraction on 143Nd/144Nd of the residue and the erupted melt.
    • fig. S6. The composition of melt collection, residues, pooled melt, and sources in the Hf-Nd isotope correlation diagram.
    • fig. S7. A comparison of the predicted compositions of the residue and erupted melt in a case with Hf offset but without Nd offset.
    • Legend for movie S1
    • References (41–53)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • movie S1 (.mp4 format). Spatial and temporal variations of 143Nd/144Nd and 176Hf/177Hf in the melting column.

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article