There is a new study out that indicates that the rate at which climate change could occur is much higher than previously known or assumed.

Those of us who study actual (historical) evolution, looking at fossils and geological layers and such, have always known that the possible rates of change in earth systems and biological systems are much higher than what we can estimate by looking at the present day. There are two reasons for this. One is a glitch in the uniformitarianism principle, the idea that processes in the past must have been the same as processes we observe today. The glitch has to so with bias. Rare events, which likely include rapid change, are rarely observed, and the time range over which scientists have been observing and recording things is too short to have seen very many such events. So, in effect, processes in the past may have features that we do not observe today, not because they can’t happen today but because they are rare and we’ve only been closely watching for a few generations.

The second reason is that we happen to have been, until recently, living in a period of less rapid change because of the nature of the Earth’s climate. Consider living on the Nullarbor Plain of southern Australia during the end of the last glacial, when sea level rise was very rapid for a few thousand years. This is one of the flattest places on earth, and is divided into two parts. The upper part makes up part of modern Australia’s landscape, but the lower part rests beneath the sea. It is very likely that even moderately fast sea level rise on that lower plain, which was exposed during the last glacial maximum, would have been noticed by anyone living there (and there were people living there). The most rapid sea level rise across an essentially flat region may have even been catastrophic. The tide goes in. It does not go out. Take a trip inland to hunt some kangaroo because you are tire of seafood, and the sea follows you. Since we don’t see such events today, we have a hard time relating to them, and in particular, we have a hard time estimating just how fast they can happen.

Meanwhile, the fossil record of both species (newly evolved, newly extinct, or just moved in or out of a region) and ecological conditions tends to show a lot of abrupt change. We assume most of that abruptness is because the readable geological record is formed during certain periods of time, and during other times, things that happen are not recorded. We are less likely to spot a period of change in the geological record than a period of stasis. Add to this the fact that many geological columns are broken up by disconformities, periods where the record that may have been of some change or another is simply eroded away. From these effects we get a geological record that tends to show abrupt change but that only rarely means abrupt change actually happened.

There may be yet another factor. Rapid change may simply leave little evidence. Slower change may accumulate clearer evidence. The degree to which this happens may depend a great deal on the kind of change involved.

The upshot of the new research is to confirm that change in the past sometimes happens more quickly than we can know from our current experience. Also the research attempts to estimate the rate of some past changes, looking specifically at global surface temperature change. The logic is pretty straight forward. Imagine you want to know how many people per 10 minute time slice enter a shopping mall. You count the number of people in the mall every hour from 8:00 opening time to noon. You then estimate the rate by dividing the increase in number of people per hour by six. But what if a particular store opened at 10:30 and had a big sale that day? Well, your estimate for the 10:00 to 11:00 period would be higher than the rest of the hours. But isn’t it true that a lot of people would show up for the sale right around 10:30? To get a better estimate, you need to make more observations, say every 20 minutes. The 20 minute period including 10:30 would yield a number, divided by 2, that would be higher than the hour-long data divided by 6. From the paper:

The scaling relationship predicts that for every 10-fold increase in measurement timespan, there is an approximately 8-fold decrease in the recorded rate of temperature change. The logical explanation for this scaling is that climate change does not proceed in a linear, monotonic manner, but is instead characterized by transient stasis and reversals, even during episodes of extreme warming. Similar explanations have been put forward for observed timespan-dependent scaling in other Earth system processes, notably sedimentation rates16 and evolution. Geological temperature changes defined at typically centennial to multimillennial timespans cannot capture the full variance of the climate system operative at shorter timescales; aliasing variability that is readily apparent from higher resolution and more recent records.

The paper draws two conclusions that could be regarded as good news, though not really. First, when we look at modern rapid global warming, and say, “look, this is happening faster than anything in the past,” we ma be overstating the case. Past temperature change could have been faster than we were thinking. Second, when we look at future likely rapid climate (and related ecological) change, and say “this rapid change will be bad for species,” we may be overstating the case. Past rapid change probably happened, so species must be adapted to rapid change.

Unfortunately, it does not work that way. We know that pretty much every ecosystem we examine over long periods of time involves repeated and significant disturbance and turnover. Also, we know from evolutionary theory and observation that species do not adapt to rapid change that happens now and then. There is no known mechanism for that adaptation to occur. So, unfortunately, these implications of the research are invalid. Rapid change has always been bad, and in the future, nothing is going to change that.

This research does show us something very important. How fast global surface temperatures can rise is not well known. We see very rapid increases in temperature (and sea level rise) now and then, but there are also decreases now and then. The forces that increase temperature and decrease temperature are always playing against each other, with the trend for many decades now being that the increases outweigh the decreases. But just how fast can change happen if, say, the forces that increase surface temperature get a few decades without mitigation from the opposing forces? This recent paper may indicate that very rapid change, much more rapid than we see now, could happen. I personally think this difference in observed rate of change and possible rate of change is more important with ice sheet melting and deterioration than it is with temperature. We might be thinking of a few feet of see level rise over several decades as likely, but this research could indicate that while that might be the total final effect, the paleorecord does not rule out that a large amount of that change may happen in just short segment of that multi-decade horizon.

David B. Kemp, Kilian Eichenseer & Wolfgang Kiessling. 2015Maximum rates of climate change are systematically underestimated in the geological record. Nature Communications.