The Effects of a Constructed Closure of the Bering Strait on AMOC Tipping Behavior

Abstract

The Atlantic Meridional Overturning Circulation (AMOC) is a major tipping element in the present-day climate, and could potentially collapse under sufficient freshwater or CO2-forcing. While the effect of the Bering Strait on AMOC stability has been well studied, it is unknown whether a constructed closure of this Strait can prevent an AMOC collapse under climate change. Here, we show in an Earth system Model of Intermediate Complexity that an artificial closure of the Strait can extend the safe carbon budget of the AMOC, provided that the AMOC is strong enough at the closure time. Specifically, for this model, an equilibrium AMOC with a reduction below (6.1 +/- 0.5)% from pre-industrial has an additional budget up to 500PgC given a sufficiently early closure, while for a weaker AMOC a closure reduces this budget. This indicates that constructing this closure could be a feasible climate intervention strategy to prevent an AMOC collapse.

Figures (15)

• Figure 1: The Bering Strait Dam. The proposed Bering Strait Dam (BSD, black lines) consisting of three separate dams: a western section connecting mainland Russia with Big Diomede Island ($\sim 38$ km), a middle section connecting Big Diomede Island to Little Diomede Island ($\sim 4$ km), and an eastern section connecting Little Diomede Island to Alaska, USA ($\sim 38$ km) (A), with corresponding depth profile along these transects (B), and the BSD in a regional overview showing it severing the Arctic-Pacific connection (C). Bathymetry data from gebco.
• Figure 2: The hysteresis experiment. The quasi-equilibrium simulations for an open Strait (blue, green), and a closed Strait (red, orange), consisting of simulations where the hosing flux $F_H$ increases (blue, red), and decreases (green, orange). The asterisks mark the estimated tipping points of the AMOC collapses (A&C), while the vertical line (dashed, black) indicates the critical hosing value for the AMOC tipping under CBS (B&D). (A) The AMOC strength --computed as the maximum overturning strength at $26^\circ$N-- for varying hosing flux $F_H$, where the arrows indicate the direction of time during the hosing experiment. The difference in AMOC strength ($\Delta\,$AMOC) between the ON-states under CBS and OBS is shown in (B). (C) The average density $\rho_{surf}$ of the top $200$ m surface layer of the North Atlantic region between $50^\circ$N and $75^\circ$N, and the difference in average density $\Delta\rho_{\text{surf}}$ (purple) and average salinity $\Delta S_{\text{surf}}$ (brown) in this layer between the ON-states under CBS and OBS (D).
• Figure 3: The freshwater transports. Freshwater transports for the quasi-equilibrium simulations for an open Strait (blue), and a closed Strait (red), consisting of simulations where the hosing flux $F_H$ increases (A-D). The asterisks mark the estimated tipping points of the AMOC collapses (A-D), while the vertical line (dashed, black) indicates the critical hosing value for the AMOC tipping under CBS (E&F). (A-D) The freshwater transports through, respectively, the Bering Strait ($F_{\text{bering}}$), the lateral boundaries of the North Atlantic region ($F_{\nabla}$), the surface of the North Atlantic region ($F_S$), and the northern boundary of the North Atlantic region ($F_{\text{north}}$). (E) The difference between the ON-states under CBS and OBS in the freshwater transports through the the North Atlantic's boundaries ($\Delta\,F_{\nabla}$, purple), its zonal boundaries ($\Delta\,F_{\text{zonal}}$, brown), its northern boundary ($\Delta\,F_{\text{north}}$, citrus), and its southern boundary ($\Delta\,F_{\text{south}}$, cyan). (F) The difference between the ON-states under CBS and OBS in the freshwater transports through the North Atlantic's surface ($\Delta\,F_S$, purple), consisting mainly of the differences in precipitation-minus-evaporation ($\Delta\,F_{P-E}$, citrus) and in runoff ($\Delta\,F_R$, cyan). Inset (G) indicates the lateral freshwater transports and their direction (red arrows) into the North Atlantic region (light-blue, enclosed in orange). The black cells indicate grid cells with a zero ocean fraction on top of the current coastlines (black, solid).
• Figure 4: $1$$\%$/yr CO$_2$-forcing experiment. The safe carbon budget of the AMOC under OBS and CBS with a starting AMOC state under OBS at various fixed hosing values $F_H$. The CO$_2$-forcing is increased at a $1\%$ rate until the budget is reached, where either the Bering Strait is kept open (blue marks), or directly closed at the start of the simulation (orange marks). Case I and II are indicated with an additional red dot. The gray (white) region indicates hosing and carbon budget values under which the AMOC collapses (does not collapse) regardless whether the Strait is closed or open. The green (red) region indicates forcing values under which the AMOC only collapses if the Strait is open (closed). On the right axis the global mean temperature (GMT) increase corresponding to the carbon budget emitted is indicated, using $1.65\,^\circ$C/$1000$ PgC ipcc2021wg1, and on the top axis the approximate corresponding $F_{ov,S}$ value of the starting equilibrium state using the linear fit in Figure \ref{['fig:supfit']}.
• Figure 5: Case I & II. Case I with a $1$$\%$/yr CO$_2$ increase for $188$ yr and hosing $F_H = 0.05$ Sv with an open Strait (blue) and an immediate closure (red), and Case II with a $1$$\%$/yr CO$_2$ increase for $93$ yr and hosing $F_H = 0.15$ Sv with an open Strait (green) and an immediate closure (orange), showing the AMOC strength (A), and corresponding atmospheric CO$_2$-concentration (B). The horizontal dashed lines indicate the maximum attained CO$_2$-concentration, which is $1820$ ppm and $693$ ppm for case I and II respectively. Note that the CO$_2$-concentration drops faster if the AMOC has collapsed, since this affects the marine carbon uptake boot2024response.