Research ArticleOCEANOGRAPHY

Coral reef islands can accrete vertically in response to sea level rise

See allHide authors and affiliations

Science Advances  10 Jun 2020:
Vol. 6, no. 24, eaay3656
DOI: 10.1126/sciadv.aay3656
  • Fig. 1 Reef island response to SLR.

    (A) Aerial photograph of Fatato, Funafuti atoll, Tuvalu; white dashed line indicates central profile line. (B) Experimental setup in the physical and numerical model. (C to E) Physical model data (black circles) and numerical model results (black line) of (C) incident significant wave height Hs,INC, (D) infragravity significant wave height Hs,IG, and (E) mean water level wl for a run with Hs = 4 m, Tp = 9.9 s, and hreef = 1 m. Vertical black dashed lines represent the reef platform edge. (F) Measured and modeled reef island morphology after 50 hours (Hs = 4 m and Tp = 9.9 s in the numerical model; representing 7 hours, Hs = 0.08 m, and Tp = 1.3 s in the physical model) with sea level raised from hreef = 2.5 m to hreef = 3 m for the optimal combination of the relevant model parameters (D50 = 14 mm; K = 0.005 m s−1; ϕ = 25o). (G to I) Measured (red circles) and modeled (black line) time series of (G) island crest elevation zcrest, (H) island crest position xcrest, and (I) overwash discharge Qcrest. The 1:50 scale physical experiment results are plotted at the prototype scale.

  • Fig. 2 Results of 3-hour numerical model simulations of reef island response to SLR.

    (A to C) Sensitivity of island response to different forcing scenarios and environmental conditions: (A) island sediment size (D50) and associated hydraulic conductivity (K), (B) sea level (MSL), and (C) offshore significant wave height (Hs). Red dashed line and black solid line represent profile at the start and end of the 3-hour model run. (D to F) Correlation between overtopping discharge averaged over the full 3-hour model run across the moving island crest (Qcrest) and morphological response parameters: (D) change in island crest position (Δxcrest), (E) change in crest elevation (Δzcrest), and (F) sediment discharge across the crest (Qsed). The vertical dashed line in (D) to (F) represents Qcrest = 0.001 m3 m−1 s−1 or 10 liters m−1 s−1.

  • Fig. 3 Results of 0.75-m SLR simulations.

    (A to C) Island adjustment to (A) 0.75 m of gradual SLR over 108 hours and typical annual storm wave conditions, (B) 0.75 m of gradual SLR over 108 hours with typical annual storm wave conditions and extreme wave perturbations, and (C) fifteen 0.05-m step changes in sea level over a total of 1620 hours and typical annual storm wave conditions. The horizontal blue dashed lines in (A) to (C) represent the sea level at the start and end of the simulation. (D to F) Change in island crest elevation (Δzcrest) and sea level (ΔMSL) relative to the start of each simulation for model simulations shown in (A) to (C), respectively. (G to I) Hourly averaged and hourly maximum water discharge across the moving island crest Qcrest for model simulations shown in (A) to (C), respectively, with the dashed line indicating Qcrest = 0.01 m3 m−1 s−1. The vertical arrows in (H) represent the 3-hour episodes of extreme wave action (Hs = 3 to 3.8 m).

  • Fig. 4 Conceptual diagram of reef island morphological adjustment to future SLR under different environmental and management scenarios.

    Island response is driven by extrinsic factors (rate of SLR, storm characteristics, and overtopping/overwashing balance) and controlled by intrinsic factors (presence/absence of conglomerate platform beneath the island, reef growth, size of the island, and sediment supply). The most appropriate adaptation strategy (managed realignment, nourishment, coastal defense, and relocation) to deal with island change is strongly determined by the type of island response to SLR. For example, an island that is narrowing, but maintaining freeboard, could benefit more from nourishment than coastal defense. If an island is already completely defended, preventing overtopping and overwashing, the only way to maintain habitation is upgrading the coastal defenses (or relocation). The width of the black bars represents the magnitude/importance/relevance of the factor in question.

  • Table 1 Hydrodynamic parameters used in numerical modeling.

    zstart, profile at start of the simulation; Hs, significant wave height; Tp, wave period; hreef, water level on reef; SLR, sea level rise; D50, sediment size; K, hydraulic conductivity; φ, sediment transport phase angle; Ttest, test time.

    Test serieszstartHs (m)Tp (s)hreef (m)SLR (m)D50 (mm)K (m s−1)φ (°)Ttest (hours)
    XB1After D249.931VariableVariableVariable50
    XB2After D249.9312–150.002–0.125–353
    XB3After D239.92–40–2140.005253
    XB4After D22–49.931140.005253
    XB5Actual (2013)2.6/2.29.92→2.750→0.75140.00525108
    XB6Actual (2013)2.6/2.2 + >3 m9.92→2.750→0.75140.00525108
    XB7Actual (2013)2.6/2.29.92→2.750→0.75140.005251620
  • Table 2 Results of the numerical model validation of the physical model run D3 (test series XB1).

    D50, sediment size; K, hydraulic conductivity; φ, sediment transport phase angle; BSS, Brier skill score (goodness of fit; cf. Eq. 4); Δzcrest, difference between modeled and measured island crest elevation; Δxcrest, difference between modeled and measured island crest position. For all model runs, Hs = 4 m, Tp = 9.9 s, hreef = 3 m, and Ttest = 50 hours. “X” denotes that the island was destroyed during the simulation. Bold values for BSS, Δzcrest, and Δxcrest represent the best performance of the numerical model in each of the groups of model runs. Positive values for Δzcrest and Δxcrest mean that the modeled crest is higher and further landward, respectively.

    Test seriesD50 (mm)K (m s−1)φ (°)BSS (−)Δzcrest (m)Δxcrest (m)
    Sensitivity to sediment transport phase angle φ
    XB1_06_005_2560.00525XXX
    XB1_06_005_2660.00527XXX
    XB1_06_005_2960.005290.790.0614.0
    XB1_06_005_3060.005300.890.258.2
    XB1_06_005_3160.005310.860.404.9
    XB1_06_005_3360.005330.740.600.4
    XB1_06_005_3560.005350.630.74−1.9
    Sensitivity to hydraulic conductivity K
    XB1_10_000_25100.000250.53−0.8321.8
    XB1_10_005_25100.005250.68−0.5617.0
    XB1_10_010_25100.010250.830.106.2
    XB1_10_020_25100.020250.660.420.3
    XB1_10_030_25100.030250.500.54−2.7
    XB1_10_040_25100.040250.420.63−4.3
    XB1_10_050_25100.050250.360.71−5.9
    Sensitivity to sediment size D50
    XB1_06_005_2560.00525XXX
    XB1_08_005_2580.00525XXX
    XB1_10_005_25100.005250.68−0.5617.0
    XB1_12_005_25120.005250.88−0.209.7
    XB1_14_005_25140.005250.770.082.1
    Sensitivity to sediment size D50 and hydraulic conductivity K
    XB1_06_010_2560.01025XXX
    XB1_07_020_2570.020250.830.318.3
    XB1_08_030_2580.030250.590.520.6
    XB1_10_050_25100.050250.360.71−5.9
    XB1_12_070_25120.070250.210.86−7.3
    XB1_15_000_25150.000250.84−0.173.5
    XB1_15_050_25150.050250.240.76−7.4

Supplementary Materials

  • Supplementary Materials

    Coral reef islands can accrete vertically in response to sea level rise

    Gerd Masselink, Eddie Beetham, Paul Kench

    Download Supplement

    This PDF file includes:

    • Figs. S1 and S2
    • Tables S1 to S4

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article