And We All Fall Down

The old dump truck struggled to make it up the grade. Long before it arrived I could hear it straining to maintain forward progress and not stall out on the slope, tires slipping on the gravel. The diesel engine growled, and the clutch complained, clanking from the pain of downshifting. The struggle was understandable: it carried six cubic yards of compost on its back. That’s about 4 tons. More or less.

As it reached the top and bumped around the corner towards the garden, the driver shot me a stare of accusation borne of his near-death experience. I managed a wordless “My bad” reply, hoping to smooth things over. The truck pulled up the drive, maneuvered around the shed, and backed into the customary pull-off, while I waved my arms in imitation of an airline ground-crew directing the docking of a Boeing 737 full of compost. “OK! You’re good!”

Ed hopped out of the cabin and unlatched the tailgate at the back of the truck bed, which dropped with a cacophony of clanking and rattling chains. Back up front, he engaged the hydraulics and the bed slowly began to lift. Squeaking and griping the whole way. I watched, expecting the compost to start streaming out the back. Except it didn’t! It just hung together in the bed as the front climbed higher and higher. When the slope of the bed reached about 30 degrees, all 4 tons was still comfortably at rest in the bed. Forty degrees, status quo. This is crazy. The bed kept rising. Forty-five degrees. Nothing. Then, just as I looked away, the pile gave way, all together and all at once.

Ssshhhhzzz...FWUMP!

The truck bed shook itself like a wet dog and, it seemed, the ground returned the meme. Four tons of compost in the back of the truck was now 4 tons of compost on the ground. And returning it from whence it came was now an impossibility. The truck bed had passed the tipping point.

tipping point noun
: The first screen on a point-of-sale terminal where you are asked to increase the price of your purchase by 25% in order to help the store owner continue to pay starvation wages

Well, yes. But that’s not exactly the kind of tipping point I had in mind.

Tipping point

Everyone instinctively understands the physics of a tipping point. We were all intimately introduced to the concept as toddlers trying to move about. Or as adults, getting too far out over our skis. Everything is going great, Granddad’s outstretched arms are right there across the room, when suddenly you are flat on your butt. Or experience the indignity of face-planting in the snow. And we all fall down.

For our purposes, a tipping point is when an incremental change in the variable of a system, a change previously tolerated, now suddenly causes a significant and irreversible alteration of state. Like 4 tons of compost suddenly going from the back of a truck to a big pile on the ground.

Here’s a trivial example:

The relevant variable, the “forcing”, is my finger pushing on the brick. I can push with constant force, and as long as the brick’s center of mass (white dot) remains to the left of the vertical red line, the brick is stable, can be controlled, and will return to its original standing state if forcing stops. However, once the center of mass crosses the vertical line – passes the tipping point – the brick falls irreversibly and assumes a new lower energy state. It falls by its own internal systems, its mass, completely independent of my finger. The previous forcing has no influence on, or control over, the outcome past the tipping point.

Environmental tipping points

Environmental scientists switched from talking about “hysteresis” to talking about “tipping points” somewhere around 2005, not long after Michael Mann’s “hockey stick” slapped the puck of popular consciousness. Not long after Malcolm Gladwell’s book hit the bestseller lists. (Though Gladwell’s book has little to do with climate, anyone writing about tipping points is legally required to mention his book. So there. I have discharged that obligation.)

By 2008, several “tipping point elements” were in the crosshairs, including arctic sea-ice, the Greenland ice sheet, the Atlantic thermohaline circulation, the west Antarctic ice sheet, the El Nino-Southern oscillation. Finished? No. Add the Indian summer monsoon, the Sahara/Sahel and West African monsoon, the Amazon rainforest, and boreal forest systems. Many arguments for additions, subtractions, and revisions have followed over the years.

Arctic sea-ice

Take the case of Arctic sea-ice. Tipping point models argue that as the sea-ice melts it exposes the darker surface of the water, which then absorbs more sunlight and, in turn, warms and melts the surrounding ice even more quickly. This non-linear positive loop is termed the “ice-albedo” feedback loop, “albedo” meaning reflectivity. Once past the tipping point, this loop is projected to result in the complete, irreversible loss of Arctic sea-ice with serious global repercussions for weather patterns (and geopolitics, I would add).

An early study of sea-ice dynamics suggested that we may have already passed that tipping point in 2008. In any case, this fall (2024) Arctic sea-ice again melted to near historic lows, according to NASA and NOAA. Zack Labe, an earth systems scientist at the NOAA Geophysical Fluid Dynamics Laboratory, shares outstanding data visualizations of Arctic sea-ice extent and other polar climate issues at his website. Zack produced the graphic I used as the header of this post. Really beautiful stuff! Scary too, if you are chasing tipping points.

Atlantic meridional overturning circulation

Not a big fan of Arctic sea-ice? Take the case of the Atlantic meridional overturning circulation (AMOC), a branch of the Atlantic thermohaline circulation. This is the ocean circulation (“conveyor belt”) that moves warm southern Atlantic surface waters up the east coast of the US, across the Atlantic to brush by the British Isles, and then on up towards the Arctic pole. By then, the surface waters are cold and longing for the Bahamas. So they sink to the ocean floor and make their way back south to pursue further worldwide adventures. I highly recommend an excellent two-part series on the AMOC by Bob Henson for the Yale Climate Connections newsletter.

Tipping point models predict that forcing by fresh water from melting Arctic and Greenland ice sheets, flowing into the North Atlantic, will blockade the AMOC and cause it to shut down. Indeed, the AMOC appears to have two quasi-stable states, “on” when it is circulating, and “off” when it is not. And it has oscillated between these two states before over geological timeframes. So it is not fantasy to think that further global warming could force a tipping point that drives the AMOC to shut down. The probability and timing of such a shutdown is hotly debated, both yes and no currently. But both sides agree the consequences would be calamitous: those living on the East Coast of the US would need to take up scuba diving, Hans Brinker would be able to ice-skate year round in the Netherlands, and the Amazon would likely dry out and need to order bottled water from Amazon.

And so that’s where we are with most of the climate tipping elements. They are plausible. The consequences are almost certain to be devastating for humanity. Current forcing trajectories are cause for significant concern. We may already be headed towards, or even be past, critical climate tipping points. But maybe systems are more resilient than we think. They are certainly complex. No one has caught a tipping point red-handed yet. Maybe these tipping points are in the thousand-year future, rather than our decades-near future. Theory fights with observation. Uncertainty abounds. And all the while the cost of false negatives is extremely high.

A fiction by way of Illustration

Several months ago I wrote about my attempts to find evidence for global warming locally, in my own backyard weather station data. In that post we bumped into recency bias, large variations in small numbers, and uncertainties with small data sets. In the end, we found convincing evidence of climate warming in official National Weather Service data for July, recorded at the airport 5 miles from my backyard. I am going to extend the use of that data set here to illustrate our current predicament.

In this graph, the mean temperature for July, in Charlottesville, is plotted as a function of year (green points). The red line is a non-linear least squares fit of the data to a model function, Func(x), where x is the year. I’ll come back to the actual function shortly. The dashed blue line is a linear extrapolation of the warming trend through the year 2100.

OK, that doesn’t look like such an unreasonable fit, eh? You could argue that future July temperatures are more likely to just keep bobbing along at about 77.5°F ad infinitum rather than steadily increase. I’d buy that. But if you accept the model fit, then we are currently in a slight upward trajectory (which we are). The trend is concerning but not alarming; average July temperatures of 83°F in 2050 would be uncomfortable. But if the world comes to grips with greenhouse gases soon, then maybe we can actually flatten out the curve and never reach mass heat stroke.

But what if we’re wrong? What if, instead, we are approaching a tipping point?

Time to talk about Func(x). That red line in the plot up there? That’s actually a tangent function. You may remember sines, cosines, and tangents from high school trig class. (If not, don’t freak out. It’s not important, so just go with the flow.) The tangent function describes the angle formed by two sides of a right triangle, the ratio of the opposite side length over the adjacent side. I picked it because it is just elementary math with a discontinuity; you don’t need to go all crazy to visualize the concept of a tipping point.

So if the tangent function represents the trend in our temperature data, then we don’t need to do the dashed blue line extrapolation out to the year 2100. Instead, we can just continue with the function calculations using the years in the future. Here’s what that looks like with the extended red line:

Yikes! By 2030 we are way beyond 83°F, and we have less than 20 years before the oceans boil. The year 2047 is pretty grim. The average July temperature in what is left of my backyard will be rapidly approaching positive infinity. Fahrenheit, centigrade, it doesn’t matter. And a few micro-milliseconds after that it will be recovering from negative infinity temperatures. (I’ve cropped the plot so you don’t have to scroll to infinity. I tried scrolling like that last year for politics, and it’s not fun.) Pretty hot, pretty cold, pretty undefined. That’s what discontinuities will do to you.

This is a fiction, of course. Not the tangent function; that’s about as solid a truth as you are going to find in this world. I mean there is no reason to think climate temperatures are following a tangent model. I made that up. But it serves to illustrate at least three points: (1) that’s how close you can be to a non-linear runaway freight train and not recognize it; (2) by the time you observe the transition it may be too late to do anything about it; and (3) if you can effectively model a system you may uncover advanced warning of potential tipping points.

Seeing the dangers ahead

Let’s spend a minute on that third point: early detection and advanced warning. This has been a top priority for environmental scientists from the beginning. Obviously, if we have an accurate model for a system, then that model can be exercised to find tipping points. This is what we did with my bogus tangent function for Charlottesville temperatures. And indeed, complex physical climate models are currently being used to examine potential transition behaviors. But physical models for complex systems are hard.

Alternatively, we might look to extract predictive parameters from historical data. For example, “spectral analysis” was applied to the Atlantic thermohaline circulation back in 2003. The basic idea is that fresh water flow into the North Atlantic varies naturally, and the response of AMOC to those changes can be evaluated by treating the data as a time series, sort of like a radio signal with characteristic frequencies. The time to a tipping point might be determined, at least theoretically, from the changes in these spectral properties. Numerous other approaches are currently being pursued. For example, as a system moves away from equilibrium, the variability in its measured parameters increases. So analysis of the statistical properties of a system might lead to predictions of tipping point behavior.

But all is not love and roses.

As early as 2010, Ditlevsen and Johnsen asserted that changes detected in ice core samples, like those used to support previous AMOC tipping points, were just stochastic noise and were useless in developing tipping point models. More recently, a team from Munich and Potsdam in Germany, and Exeter in the UK, argue that the variances in historical climate data are so large that they preclude making any reliable predictions of future behavior.

Finally, some even argue that “tipping points” have passed their own tipping point. In a paper published this month in the journal Nature Climate Change, titled “Tipping points’ confuse and can distract from urgent climate action,” a team led by Robert E. Kopp (Rutgers) wants to banish the term altogether.

Kopp is quoted by Kate Yoder in her article about the study in Grist:

“Tipping points are not, as a way of looking at the world, some inherent property of the world,” Kopp said. “It’s a choice to use that framing.”

Tell that to my 4 tons of compost that suddenly broke friction, tipping from the bed of a truck to a pile on the ground. Oh, that was just my framing. Kopp suggested to Yoder that we should call such transitions “potential surprises” rather than tipping points. Now that’s brilliant! Plus it gets around having to cite Malcolm Gladwell all the time.

Conclusion

There is nothing magic about explosive critical transitions in natural systems. In fact, such behavior may be universally expected in the presence of a forcing input. Like continuing emissions of greenhouse gases, for example. And the consequences of complacent miscalculation will be anything but just a “surprise."

Right now? Today? To be honest, I am less concerned about imminent tipping points than I am about imminent ignorance. We need to continue broad based climate research, both observational and theoretical. By themselves, the hands-on observational scientists could be off in a cloister preserving catalogs for a future in doubt. By themselves, the theoreticians might as well be smoking pot in their own breakout room. That’s not what’s happening right now. And it’s all good. The conflicts of interpretation, the push and shove, the drive for better models and better data is the way science progresses. But now this is all at risk.

The last time the incoming Wrecking Ball, Inc. company took control of the US government, it wiped “climate change” from the websites and daily operations of our national institutions. Massive spontaneous data downloads by scientists preserved critical data sets, just in case. Now, it is happening again because the peril has actually increased. After all, the CEO of Wrecking Ball, Inc. solicited bribes for $1 billion from the fossil fuel industry during his campaign, in exchange for the promise to gut environmental regulations. Efforts to privatize NOAA and NASA have been advertised in advance and are likely to be redoubled.

The political forcing trajectory is not good. Over the short term, we will need to apply all the blocking inputs we can manage. We will need to call out funding cutbacks, agency manglement, profligate destruction, and conflicts of interest. We must support what allies we can find in Congress. And we will need to support journalists and academic centers that report on science, climate, and environmental issues.

Or we all fall down.