This well written work proposes a conceptual model coupling ocean dynamics with sea-ice cover dynamics. Their model exhibits rate-induced bifurcations in the ocean dynamics and their results suggest that cascading bifurcations may be a general mechanism in inducing abrupt transitions in climate systems. The authors nicely illustrate the influence of noise on the bifurcation behaviour with convincing numerical simulations. The authors further show that the critical slowing down experienced during rate-induced bifurcations may be difficult to detect using the standard variance indicator (but is readily detected in the sea-ice dynamics which experiences a “classical” bifurcation) and propose a simple indicator based on estimating the elements of the associated Jacobian. 



I enjoyed reading the paper and believe that its results — both on the particular role of rate-induced bifurcations on cascading transitions and on warning indicators — are valuable for the community, and I recommend publication subject to some comments the authors may want to take on board.

This paper presents and analyzes a dynamical system where the Stommel model is coupled to a zero-dimensional sea-ice model. The sea-ice model exhibits a fold bifurcation. Since there is a separation of time scales between the ice dynamics and the ocean dynamics, as the ice model's tipping point is crossed, there is an almost instantaneous "parameter change" in the Stommel model. The change in ocean dynamics may hence be described as a rate-induced tipping point.   

The paper is well written, but I still find it difficult to read. I encourage the authors to present a clearer "take-home-message."  

I have a few suggestions and questions that the authors may use in their revision: 

1. Why not present and analyze the coupled model (Eq. 6) before discussing rate-induced tipping in the Stommel model? I feel that a "standard analysis" of the model in Eq. 6 is missing? You have a relatively simple dynamical system and one control parameter (R). Don't you have a simple saddle-node bifurcation in the three-dimensional system? 
2. As you go into more detail, is it possible to be more precise? I feel that it becomes very descriptive.  
3. How important is it that the Stommel model has a bistable regime? What would happen if you just coupled the sea-ice model to a simpler model with a smooth transition between "modes". 
4. I would like to see a discussion of how your proposed EW indicator would work in a "real-data setting". 

There are a few typos in the manuscript. You'll find them when you read through it carefully. I am looking forward to reading a revised version.