Rapid attribution of PNW heatwave

Summary: It was almost impossible for the temperatures seen recently in the Pacific North West heatwave to have occurred without global warming. And only improbable with it.

It’s been clear for at least a decade that global warming has been in general increasing the intensity of heat waves, with clear trends in observed maximum temperatures that match what climate models have been predicting. For the specific situation in the Pacific NorthWest at the end of June, we now have the first attribution analysis from the World Weather Attribution group – a consortium of climate experts from around the world working on extreme event attribution. Their preprint (Philip et al.) is available here.

In the paper, they show that this event was truly exceptional in the temperature statistics in the region and specifically in Vancouver, Seattle and Portland, but that the geopotential height anomaly (one measure of the ‘heat dome’ or the ridging) was not that far off from the underlying trends. Using various methods to construct the return time for this event, they show that even in today’s climate this was a one in 400 yrs or longer event. Using only data from before 2021, an anomaly this large is estimated to never occur!

What happened?

Everyone is agreed that the specific synoptic situation is unusual. A large ‘omega’ pattern (so-called because of the resemblance to the Greek letter ) set up by a Rossby wave breaking event, intersecting with the topography and the warm temperatures to the south-west that were advected to the PNW. But the question is whether the temperature extremes are being made substantially more likely by underlying climate changes.

First, look the maximum annual daily temperature in the region (according to the ERA5 reanalysis for the historical data and the actual weather forecast analyses since June 1st). There is a real trend of about 4ºC over the last 70 years – roughly 3 to 4 times the trend in the global mean temperature. Nonetheless, the magnitude of the regional anomaly is more than 5ºC above the previous record. That is, literally, phenomenal.

The temperature anomaly at the local station level is similarly huge at SeaTac airport, Portland International Airport and New Westminster (nr. Vancouver). Note that while the cities might be affected by urban heat island effects that would exacerbate the temperature signal, that would not affect the regional analysis above, nor the situation in village of Lytton, BC which set a massive new all-time Canada-wide temperature record on Tuesday June 29th and promptly burnt down a day later.

However, the situation is a little different if you look at the geopotential height anomalies – these are affected by the synoptic situation as well as the integrated temperature anomalies. In that case, while still record-breaking, the anomaly is not totally beyond expectations. Indeed, the trend in z500 values is similar to the situation in Western Europe last year.

Together these analyses suggest a synoptic situation that is rare, but not inconceivable, but with temperature anomalies that are off the charts.

Attribution

The way that attribution for extreme events works (as discussed previously on RealClimate here and here etc.) is that you look at the situation with and without the anthropogenic global warming signal and calculate the ratio of probabilities. If an event is say, twice as common with the GW, then one can give a fractional attribution of 50% to anthropogenic forcing and the return time is half what it used to be. If it is five times more likely, the attribution is 80% = 100*(5-1)/5 and the return time is a fifth of what it used to be. In this case, we are seeing probability ratios of 150 to 1000s, suggesting that these, improbable, temperatures can be almost entirely attributed to global warming. Without the anthropogenic signal, temperatures this extreme wouldn’t have happened in thousands to tens of thousands of years.

Rainfall and soil moisture deficits as a precursor?

In many previous extreme heat events, such as the 2003 European heatwave, rainfall deficits and dry soils the prior spring were shown to have made an important contribution to the temperature extremes, and so it’s worth looking at the same phenomena here. The IMERG data which are mostly based on satellite rainfall amounts do show a moderate deficit in the area over the last four months, but not so much of an effect that it could explain the anomaly on it’s own. The magnitude of this effect will be examined further in the months to come.

All models are wrong?

This kind of attribution is of course only as good as the models being used. In such a rapid attribution study, that means that the authors depend on an existing database – in this case, from CMIP5 and CMIP6 – and while they screen the models for fidelity in matching this genre of event, it’s possible that there are systematic issues with this class of model for a specific aspect of the situation. For instance, Mann et al., (2018) find that the CMIP5 models have a poor representation of a quasi-resonant (QR) phenomena in jet stream waves that are associated with the ‘omega’ pattern blocking event seen here. [Update: the specific claim in the paper relates to oscillations with wavenumber 6-8, while this event was more of a wavenumber 4 phenomena – see comment #9 below]. The expected trends in QR suggest an increase of about 30% in such events today over the situation in the pre-industrial. If models don’t capture this behaviour, it will make the event seem more unlikely than it really is. This might be resolved in higher resolution modeling specific to this event, but doesn’t really affect the broader conclusions.

Maybe it was just really, really, really unlikely?

Some people still reject these lines of argument, typical of this is Cliff Mass in this recent blog post. For them, the trends in max temperatures are (literally) ignored, and the fact that this phenomenon is being seen around the world is just a series of increasing unlikely combinations of factors that for some inexplicable reason keep happening. But this is really just a case of synoptic myopia – paying too much close attention to the series of specific events that lead to the specific situation, and not seeing the wood for the (burning) trees.

p.s. (8 July): In 2012 we published the highly relevant post Extremely hot, starting like this:

One claim frequently heard regarding extreme heat waves goes something like this: ”Since this heat wave broke the previous record by 5 °C, global warming can’t have much to do with it since that has been only 1 °C over the 20th century”. Here we explain why we find this logic doubly flawed.

Pretty much exactly what happened! And it ends thus:

So in summary: even in the most simple, linear case of a shift in the normal distribution, the probability for “outlandish” heat records increases greatly due to global warming. But the more outlandish a record is, the more would we suspect that non-linear feedbacks are at play – which could increase their likelihood even more.

References

1. 
   M.E. Mann, S. Rahmstorf, K. Kornhuber, B.A. Steinman, S.K. Miller, S. Petri, and D. Coumou, “Projected changes in persistent extreme summer weather events: The role of quasi-resonant amplification”, Science Advances, vol. 4, pp. eaat3272, 2018. http://dx.doi.org/10.1126/sciadv.aat3272