Objective: Understand the signs and strengths of key feedbacks controlling AMOC hysteresis to better understand the likelihood of AMOC collapse in the future.


The Atlantic Meridional Overturning Circulation (AMOC) plays an important role in the climate system. There is evidence from abrupt changes in the paleo record, and from theories, simplified and more complex models, that the AMOC may be able to experience abrupt change and/or hysteresis. Abrupt change means that a small change in forcing (i.e. surface freshwater input) beyond a critical threshold could cause the AMOC to collapse to a weak state, while hysteresis means that the AMOC would not necessarily recover to its original strength when the forcing is reversed.

Until recently, this behaviour has only been seen in simplified models, Earth System Models of Intermediate Complexity (EMICS), and very low resolution climate models (Rahmstorf et al 2005; Hawkins et al 2011), however Mecking et al (2016) and Jackson and Wood (2018), have recently shown similar behaviour in a CMIP6-generation coupled climate model. The recent results from Jackson and Wood (2018) found a rather higher sensitivity of the AMOC to fresh water forcing than in previous AOGCM studies, with sustained hosing of 0.1 Sv being enough to substantially weaken the AMOC.  

There have been a few multi-model studies examining the response of the AMOC to increased freshwater input (hosing), however these have either been conducted with early generation climate models (Stouffer et al, 2006), or have focused on simulation of specific climate scenarios (Swingedouw et al, 2013; Bakker et al, 2016). Although some studies have examined processes, there has been limited analysis on identifying and quantifying the dynamical processes that control the AMOC across models.

Here we aim to understand the processes and feedbacks controlling the AMOC response in current generation GCMs, to understand which are the most important, and how they vary across models, and compare them with available observations. The outcome will be an improved ability to describe the circumstances under which an AMOC collapse might be triggered, with the potential to develop observable early warning signals for the approach of the AMOC towards a threshold. 

Project outline

We propose a set of idealised experiments with CMIP6 unfluxadjusted, coupled climate models to investigate the sensitivity of the AMOC to freshwater forcing. This project is called NAHosMIP (North Atlantic Hosing Model Intercomparison Project). Although we are using CMIP6 models this is not an official CMIP6 MIP. 

The aims of this project are:  

  • Understand whether other climate models exhibit a threshold in the AMOC (whether if it is weakened sufficiently by hosing, the AMOC stays in a weak state) 
  • Understand how different feedbacks contribute to different AMOC behaviour in different models, and what the implications of this are for the real world. 
  • Provide sources of data that can be used in additional studies (ie impacts of AMOC weakening). 
Proposed experimental design

A small number of idealised experiments based on the ensemble of Jackson and Wood (2018) were proposed, applying 0.3 Sv of hosing over the North Atlantic for 20 and 50 years (with some experiments applying hosing for up to 100 years). We also proposed parallel experiments with hosing of 0.1Sv around Greenland only, which is a more realistic extent and magnitude and may affect how the hosing affects deep water formation regions. In all experiments the addition of freshwater was compensated for throughout the volume of the ocean.  

Groups and models involved
Met OfficeUKHadGEM3-GC2 (pre CMIP6)
Additional studies planned

Potsdam University: Using HadGEM3-GC3.1MM experiments to look at impacts of AMOC collapse on W African climate and extremes 

Exeter University: Using HadGEM3-GC3.1MM experiments to look at impacts of AMOC collapse on precipitation and vegetation in the Amazon 

Exeter University:  Tipping points in overshoot scenarios. Comparisons with analytical models.  

ISAC-CNR: Impacts of AMOC weakening 

CCCma: Sea level response to AMOC weakening 


Bakker, P., Schmittner, A., Lenaerts, J. T. M., Abe-Ouchi, A., Bi, D., van den Broeke, M. R., Chan, W. L., Hu, A., Beadling, R. L., Marsland, S. J., Mernild, S. H., Saenko, O. A., Swingedouw, D., Sullivan, A. & Yin, J. (2016). Fate of the Atlantic Meridional Overturning Circulation: Strong decline under continued warming and Greenland melting. Geophys. Res. Lett., 43, 2016GL070457+.  

Hawkins, E., Smith, R. S., Allison, L. C., Gregory, J. M., Woollings, T. J., Pohlmann, H. & de Cuevas, B. (2011). Bistability of the Atlantic overturning circulation in a global climate model and links to ocean freshwater transport. Geophys. Res. Lett., 38. 

Jackson, L. C., & Wood, R. A. (2018). Hysteresis and resilience of the AMOC in an eddy‐permitting GCM. Geophysical Research Letters, 45, 8547–8556. 

Mecking, J. V., Drijfhout, S. S., Jackson, L. C., & Graham, T. (2016). Stable AMOC off state in an eddy‐permitting coupled climate model. Climate Dynamics, 47, 2455–2470. 

Rahmstorf, S., Crucifix, M., Ganopolski, A., Goosse, H., Kamenkovich, I., Knutti, R., Lohmann, G., Marsh, R., Mysak, L. A., Wang, Z. & Others (2005). Thermohaline circulation hysteresis: A model intercomparison. Geophys. Res. Lett, 32. 

Swingedouw, D., Rodehacke, C., Behrens, E., Menary, M., Olsen, S., Gao, Y., Mikolajewicz, U., Mignot, J. & Biastoch, A. (2013). Decadal fingerprints of freshwater discharge around Greenland in a multi-model ensemble. Climate Dynamics, 41, 695-720. 

For more information, please contact:

Laura Jackson

Met Office