[BostonCard]

My point is that you don’t need 4000 years of data to know that an event is a 1 in 4000 anomaly, if certain conditions are met (Normality, stability, independence, etc.).  The comment that the temperature in the relevant part of the ocean was 3.5 standard deviations above the fifty year average and that this would be expected to happen approximately 1 in 4000 years is fine.

A 3.5 SD temperature anomaly does not imply a 3.5 SD rainfall anomaly, or at least, it shouldn’t, since many factor other than the temperature of the ocean affect rainfall.  If you saw an article implying that we should see 1 in 4000 year rainfall, I think you are correct in skewering it.  However, the articles that I’ve seen have been careful to say that that the anomaly refers to temperature, and that typically, El Niños produce heavier rainfall in Southern California, and, at most, speculating that the rainfall might be very heavy.  But the articles in the Sacramento Bee that I linked earlier went out of its way to say that this doesn’t always happen, and one of the biggest El Niños in recent history, the “Godzilla” 2015-16 El Niño did not produce outsized rainfall.  Here’s the direct quote:

The 2015-16 El Niño, the strongest on record and dubbed as a “Godzilla” El Niño at the time, did not produce expected drought-busting rains for the state. Swain said this and the tiny sample size of large El Niño events also makes it difficult to communicate the situation.

Note that he specifically calls out the relatively small sample size of large El Niño events, which implies that correlating very large temperature anomalies to expected rainfall is fraught.

BC

[BostonCard]

My point is that you don’t need 4000 years of data to know that an event is a 1 in 4000 anomaly, if certain conditions are met (Normality, stability, independence, etc.).  The comment that the temperature in the relevant part of the ocean was 3.5 standard deviations above the fifty year average and that this would be expected to happen approximately 1 in 4000 years is fine.

A 3.5 SD temperature anomaly does not imply a 3.5 SD rainfall anomaly, or at least, it shouldn’t, since many factor other than the temperature of the ocean affect rainfall.  If you saw an article implying that we should see 1 in 4000 year rainfall, I think you are correct in skewering it.  However, the articles that I’ve seen have been careful to say that that the anomaly refers to temperature, and that typically, El Niños produce heavier rainfall in Southern California, and, at most, speculating that the rainfall might be very heavy.  But the articles in the Sacramento Bee that I linked earlier went out of its way to say that this doesn’t always happen, and one of the biggest El Niños in recent history, the “Godzilla” 2015-16 El Niño did not produce outsized rainfall.  Here’s the direct quote:

The 2015-16 El Niño, the strongest on record and dubbed as a “Godzilla” El Niño at the time, did not produce expected drought-busting rains for the state. Swain said this and the tiny sample size of large El Niño events also makes it difficult to communicate the situation.

Note that he specifically calls out the relatively small sample size of large El Niño events, which implies that correlating very large temperature anomalies to expected rainfall is fraught.

BC