It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Tuesday, January 26, 2016

Is climate change the cause of irregular monsoon in India?

From Quora

Is climate change the cause of irregular monsoon in India?

My short answer is “I think so”, so you may want to skip the rest of this answer.

 This is a very interesting question, one of the larger class of questions “Is this particular phenomenon caused by climate change”, but one where the answer is particularly complicated, because monsoon variability lies right at the edge of “climate” and “weather”. Some would consider each individual monsoon onset a “climate variation” while others would consider it a “weather event”. The pat answers one usually hears about “the difference between weather and climate” rather founder on the rocks in addressing the question.

First of all, let’s make sure we agree what you mean by “climate change”. Both “climate change” and “global warming”, which are in some sections used interchangeably, are problematic. In a literal formal sense, “climate change” is a change in the statistical patterns of weather events. And you are trying to attribute a specific perceived change to “climate change”. So literally, you are asking “is this particular climate change because of climate change”, which is not really any more meaningful than asking “is Mars cause by planets”. The postulated increase in monsoon irregularity is by definitiion a type of climate change. So it cannot be caused by climate change.

To rephrase the question as “is irregular monsoon caused by global warming” is little help. Global warming is an observation of an increase in surface temperature. It is a consequence of a vast number of local responses to a global change in conditions. Among those conditions is the monsoon onset over India. This gets cause and effect backwards. The global change is an aggregation of local conditions - it is not directly the cause of local conditions.

I think the best literal phrasing for what is informally (and somewhat incorrectly) usually called “climate change” or “global warming” is “anthropogenically forced climate change”. That is, what you are probably asking is how much humans are responsible for an increased irregularity in monsoon behavior. A good shorthand for this is “climate disruption”. That is, we are discussing to what extent an event is attributable to human activity in changing the properties of the atmosphere.

There are two basic approaches to this problem, that yield dramatically different results.

The first is purely statistical. One looks at the historical record for a particular class of event - for example the onset of the monsoon in India. There are two steps to such an analysis - the first is to find a statistically significant trend, and the second is to test the hypothesis that the trend is correlated to some causal agent. In the case of climate disruption studies, in increasing order of appropriateness, one might use CO2 emissions, CO2 concentrations, CO2-equivalent concentrations, net top-of-atmosphere forcings (accounting for anthropogenic arerosols) or most difficult but theoretically best, all forcings as distributed geographically. (The use of emissions per se is strictly speaking incorrect, though it’s often seen in the less serious critiques of climate science.) Then, using “frequentist” argumentation, one tests the “null hypothesis”: is the observed correlation less than 5% likely due to chance.

It is quite possible, using this approach, to attribute global warming in the literal sense to climate disruption. But one cannot go much further with this approach, especially for rare events like typhoons in a given region, or for annual events like monsoon arrival dates. This is because the natural variability in these statistics is high.

 For example, the extraordinary late arrival of the monsoon in 2015 was not dramatically different from that of 1899, which caused even greater social devastation than last year’s event.

(Incidentally, this was a crucial event in the emergence of physical climatology as a scientific discipline. Sir Gilbert Walker, a diligent and creative scientist, was assigned by the British Empire to find causes for these events. He immediately concluded that “The variation of monsoon rainfall ... occur on so large a scale [that we can assume they are] preceded and followed by abnormal conditions at some distance” [“Floods, Famines and Emperors” by Brian Fagan, 1999], an insight which directly led to the discovery of the Walker Circulation and the Southern Oscillation, and indirectly to our current understanding of El Niño.)

 So while the events of 2015 were extraordinary (and I was very aware of them at the time because we were simultaneously having extraordinarily wet weather in Texas, which turned out to be by far the wettest month in our history) they are not, in isolation, without precedent.

There are only a few hundred observed monsoons, and only a handful of outliers. Finding trends among the outliers is by its nature doing statistical reasoning on very small data sets. And such reasoning is almost always inconclusive. This doesn’t mean human agency is absolved of influence, merely that it cannot be convicted beyond reasonable doubt using statistical methods.

But there’s a second way of looking at it. We can look at the physics of monsoons, and try to understand what makes them late, following in Walker’s footsteps with much richer understanding and tools. This amounts to studying the specific details of monsoon formation. I am sure papers are working their ways through the journals as I write this, but let me draw upon established knowledge here.

The monsoon is driven by differential heating between the continental landmass and the surrounding ocean. As the land heats, it creates a pool of hot air, which surrounded by relatively cooler, denser air is forced upward, causing condensation in rising air columns, causing intense rainfall. In 2015, however, the Indian Ocean was particularly warm. So this delayed the onset of the monsoon because the cool ocean air wasn’t actually cool enough to start the monsoon dynamic.

Next, we have to examine WHY the Indian Ocean was warm. This is complicated, and I’ll resist the temptation to speculate in detail. I am sure people are thinking about this in more detail than I have. But here we have a strong possible connection to anthropogenic forcing. Ocean surface temperatures have been extraordinarily high globally since 2014. So here is a very plausible connection to human activity, through radiative imbalance leading to global warming.

Can we prove this in a statistical sense? Absolutely not, as I explained above. The sample size is too small. But does it mean that there is no connection?

That is an absurd conclusion. If I punch you in the face and your jaw breaks, you do not need statistical significance to bring me to justice!

Furthermore, as I explain here , I think it is reasonable to expect larger seasonal excursions as climate change proceeds. This doesn’t mean that we can prove anything in court about any individual such anomaly. It’s important to understand that anomalies would have occurred in an undisturbed environment too. But it’s also important to understand that they would not be the same anomalies. Weather depends sensitively on the distribution of sea surface temperatures and their geographic gradients. We expect these to change with increasing vigor. Are we seeing the beginnings of these effects? Certainly so. Is there a causal chain? Clearly there is.

But are we worse off than we would have been in an undisturbed climate? This depends on the phenomenon in question, and in most cases it can’t be proved. That doesn’t mean we can’t say the evidence suggests that climate disruption is making matters worse. In the matter of the Indian monsoon, it seems more likely than not.


Oale said...

Indian monsoon has always been pretty erratic, so it's pretty hard to say even if there has been an increase in erraticity... The question might be a good subject for a study for such a time when Arctic is ice free, as this changes also the pressure systems involved with the cool season in india... This is by no means the only thing influencing the development of monsoon. We might f.e. ask are the ocean currents in indian ocean going to stay the same when their surface warms to the equilibrium levels set by greenhouse gases... I think the Pakistani floods were partly attributed to agw, and Yemen has had some more nasty storms nowadays, so a better question might be, is the indian ocean monsoon system changing its location due to Agw?

Michael Tobis said...

"Indian monsoon has always been pretty erratic, so it's pretty hard to say even if there has been an increase in erraticity... "

Yes, indeed. But on the other hand, it is extremely implausible to suggest that recent warming, in particular increased sea surface temperature in the Indian Ocean, is irrelevant.

That's the pickle we are in regarding numerous attribution questions. Physics says "inevitable" and statistics says "invisible", let's call the whole thing off.

"Is the indian ocean monsoon system changing its location due to Agw?"

Interesting approach. However, on the ground at a particular locality, this would appear to be a change in onset dates (the traditional measure of monsoon behavior) wouldn't it?

Oale said...

"it is extremely implausible to suggest that recent warming, in particular increased sea surface temperature in the Indian Ocean, is irrelevant."

This. Would it be the equatorial convergence zone and related currents that on Indian ocean stay put in their historical locations Spite the warmer surface waters? As the Ocean has warmed a bit and doesn't have an outlet to north like Atlantic and the Pacific, the expanded water might keep the itcz unmoved, though everything around it warms? The warmer waters on Sahul area (Indonesia) could possibly push the clouds forming the monsoon front a bit to the east? True also your second comment, this should indeed be seen more the more to the east your measurement location is... On El Nino years as there is slightly cooler waters in Indonesia, the old rules of weak monsoons would continue to be in effect... or maybe the west Indian coast will get some more rains due the overall warming?

I'm not acquainted at all of the more detailed monsoon histories in different parts of India so just throwing in some ideas.