It looks like you're new here. If you want to get involved, click one of these buttons!
The use of "climate networks" to predict things like El Niños is new, controversial, and frankly I'm not sure it works. So, we should expect pushback from climate scientists, and we should try not to come across as advocating these methods unti we feel sure that we should.
Over on G+, I had a strange exchange with a guy named Jim Carver, which spiralled out of control. It started out civilly enough when he wrote:
The definition in that paper you are using is not accurate. They say something to the effect that an ENSO event is when 5 consecutive months exceed thresholds by +- 0.5C.
That’s not the traditional definition. It is when 5 three month averages exceed the thresholds. That’s not the same as five months.
Other definitions do exist, such as conditions forecast in Nino 3.4 to be above thresholds for three months after a three month period has been validated but that’s only if atmospheric anomalies are present.
If you look at the real models output which take into account all parameters (not just the ones here) you will find the slope of trends are down and most forecasters are calling for a weak to moderate event by N. Hemisphere winter.
I am not. My outlook is for a 50% chance for it to remain stable below thresholds for the foreseeable future and no El Niño event occurring. If it does, it will be weak, and the strength of the event is more important than its actual occurrence.
Thanks for your helpful comments.
You’re right, Ludescher et al are not using the standard definition of El Niño – I don’t know why. Of course they’re allowed to predict whatever they want to predict. But in Part 6 of this series we’ll present serious criticisms of their ability to predict what they claim to predict.
Regarding current El Niño predictions, we should note that Ludescher’s paper was published in February and written before that: it’s claiming to do “long-range” El Niño forecasts, so it’s a different thing than the best forecasts we have now of what’ll happen this fall.
The National Weather Service came out with new predictions on July 10th:
They claim the chance of El Niño is about 70% during the Northern Hemisphere summer and is close to 80% during the fall and early winter. They expect it to peak at weak-to-moderate strength during the late fall and early winter (3-month averages of the Niño 3.4 index between 0.5°C and 1.4°C).
That's true John, and I am not going with that. I am a fully trained climatologist of 35 years and I'm going with my prediction. I may have to eat my words, but I doubt it. haha...I shouldn't be so cocky.
But you have a greater chance of being wrong than I do because you don't have much or maybe any experience in meteorological forecasting or modeling of climate systems for a full model interpretation. You just don't have the resources.
I thought it was weird for him to say I had "greater chance of being wrong" when I hadn't made any predictions yet. So, I wrote:
I haven't actually made any El Niño predictions. Until I do, my chance of being wrong is 0%. That's a trick I've learned for being right most of the time...
The Azimuth El Niño project certainly needs all the help it can get, especially from experts in meteorology and climatology. We've been lucky to get help from someone with fairly serious computer resources. But we're not planning to write and run atmospheric or ocean models - that wouldn't make sense! There are other niches to fill. Mainly, we want to figure out a bunch of stuff.
His response was helpful until the end, which took me by surprise:
That's great John. Maybe you should check your papers that you are using for errors.
Then you could maybe find what you are looking for. I would think that flawed definitions don't lead to very good results. I wouldn't know how I think that. Maybe I just know that from doing it.
Yeah, it's a three month average...and goes along for five cycles. Five consecutive three month cycles. Not five months. So we can explain this very easily, if you take an average of the sea surface anomalies in section Nino 3.4 for a three month average it could look like:
or whatever, I didn't put actual values in there but I'm pretty close because I work with these numbers everyday. You will notice that if any of the moving averages don't meet the threshold values then they don't count.
And they will never count. Obviously because they didn't reach the required value.
Now there are other things you should be aware of, and that is there are also people on the ground. Really? Actual people who are there and living it. Who knew?
Reports from Peru are the fish are coming back and millions of anchovies off the coast (or right near as a matter of fact) of California. Not empirical to be sure...BUT god damn significant and we can't rule that out in the forecast.
You can't sit there at your computer screen and tell me what is happening at any particular point. I can't tell you what your weather is at this particular time either. I can come close, but the best god damn analysis you will ever have is looking outside.
Weather and climate is about 50% statistics, and the rest is perspiration and intuition, and not necessarily in that order. Personally, at this point, I would give you about well, let's be generous: less than half of any of it.
I personally think it's a travesty because you are so lame and come off like you know something. You really don't know a damn thing. And to come off with that shitty paper as your basis shows how little you guys know.
Maybe he worked himself into a rage while writing this? Maybe I'm annoying to him I'm a "bigshot" and I'm blundering into his field of expertise?
Anyway, I feel my main mistake was to know Ludescher et al were using a nonstandard definition of El Niño but not say this in Part 3. I've corrected that. I don't think using a nonstandard definition completely negates the value of their work. However, it makes me wonder how much their work would change if they used the standard one.
Another mistake was not to make it clear that I know Ludescher's whole approach is controversial and perhaps misguided.