Fig. 1

(a) Image of the natural Meiji Atoll, modified from Zhao (1996) and the corresponding Google Earth satellite image. The black circles indicate the locations, and black numbers indicate the water depths measured during the scientific expedition on Nansha Islands (1987-1989) by the Chinese Academy of Sciences (Zhao, 1996). Please note that the widths of reef flats are not on a precise scale, and the locations and boundaries of lagoonal patch reefs are approximate. Section 0 (the purple line) is a section that was investigated during the scientific expedition on the Nansha Islands (1987-1989). Section 1 (the blue line) is a section that passes through the location of the NK-1 well and extends in the windward and leeward directions. The modeled results along section 1 are the focus and presented in this work; (b) Water depths along section 0; (c) The Pleistocene/Holocene unconformity (the red curve) in a west-east cross section of Meiji Atoll, modified from Liu et al (2019). The numbers on the left side of the boreholes (i.e., the vertical blue lines) indicate the depths (m) below sea level, and the numbers on the right side of the boreholes indicate the depths (m) below seafloor. Please note that the distance was not labelled under uniform scaling"

Fig. 2

Environmental threshold functions for shallow (top row), moderate-deep (middle row) and deep (bottom row) assemblages (Salles et al, 2018b, 2018c)"

Fig. 3

The annual precipitation data used in the simulations (b) indicated by the ratio of clay/feldspar (a) recorded from ODP Site 1143. The clay/feldspar ratio is from Wan et al (2006)"

Tab. 1

Model parameters used in the simulations"

Fig. 4

Three-dimensional view of the modeled Meiji Atoll since 2.65 Ma. The light orange areas indicate the reef rims or flats that reach mean sea level, and the light orange spots in the lagoon are the modeled lagoonal patch reefs that reach mean sea level. The bluer areas have greater water depths. Detailed water depths are presented for a cross section (i.e., section 1) in Fig. 6"

Fig. 5

(a) Modeled age/depth (thin blue x symbols and solid blue line) compared with the measured age/depth (black circles, red triangles, green diamonds and dark red stars) from the NK-1 well; (b) Modeled and measured age/depth with the sea-level changes (dark green x symbols and dark green solid line) (Bates et al, 2014) used for the area of the Meiji Atoll. In Fig. 5a and 5b, the black circles, red triangles, green diamonds and dark red stars indicate the mean strontium isotope age, U-Th age, C-14 age and paleomagnetic age (2.6 Ma at 166 m and 0.78 Ma at 52 m beneath the reef rim surface), respectively. In Fig. 5a, the horizontal black lines and red lines through black circles and red triangles, respectively, indicate the ranges of measured ages. Several black circles and red triangles have indistinguishable horizontal lines because the ranges of these measured ages are too narrow to discern. (1), (2), (3) and (4) correspond to the four phases of reef accretion as described in section 2"

Fig. 6

Modeled age/depth for section 1. (a) Modeled age/depth for the strata on section 1; (b) modeled age/depth for the stratum interfaces on section 1. The white numbers indicate the ages of modeled strata"

Fig. 7

Modeled reef morphology at 0.45 Ma (a) and 0.40 Ma (b) on section 1"

Fig. 8

Modeled water depths of the top surface of the reef. (a) Map highlighting the modeled top surface of the reef that is less than 5 m deep; (b) map highlighting the modeled top surface of the reef that is less than 50 m deep"

Tab. 2

Deviation analysis between simulated results and measured data"

Tab. 3

Summary of the simulation tests in this study"

Fig. 9

Modeling results for section 1 from two simulation tests. (a) using a uniform dissolution factor of Fdr = Fdl =3.75×10-5 for both the margin and interior zones; and (b) using a rainfall-runoff-related dissolution factor of Ff = 2.5×10-4"

Fig. 10

Simulation tests using the depth-dependent production function (D0), the horizontal reduction function (U0) and the sediment input controlling function from Warrlich et al (2002). A spatially nondifferential dissolution rate, related to annual rainfall amount (i.e., Fdr = Fdl =3.75×10-5) was used for both the margin and interior zones. (a) The function describing the depth dependency of carbonate production (D0); (b) the function describing the reduction in carbonate production rate due to deviation from fully marine waters (U0); (c to e) the modeling results for section 1 with U0 included, using maximum vertical accretion rates of 0.01, 0.0075 and 0.005 m·a-1, respectively; (f) modeling results for section 1 without U0, using a maximum vertical accretion rate of 0.01 m·a-1. h is the water depth (m). hb is the bottom depth of maximum carbonate production rate in the function D0 (m). σD is the scale depth for carbonate production in the function D0 (m). dp_m is the distance from a given location to its nearest margin on the platform top (m)"

Fig. 11

Modeling results on section 1 under different assumptions. (a) 0.45 Ma, (b) 0.40 Ma, and (c) 0 Ma, assuming no subsidence; and (d) 0 Ma, assuming a nonflat-topped initial basement"

Tab. 4

Comparison of the drowning periods of the reef body in Core NK-1 from Meiji Atoll with the interglacial or deglacial drowning periods identified by Droxler et al (2021)"