Ocean-wide Sensor Array Provides New Look at Global Ocean Current (nature.com)

An anonymous reader shares a Nature article: The North Atlantic Ocean is a major driver of the global currents that regulate Earth's climate, mix the oceans and sequester carbon from the atmosphere -- but researchers haven't been able to get a good look at its inner workings until now. The first results from an array of sensors strung across this region reveal that things are much more complicated than scientists previously believed. Researchers with the Overturning in the Subpolar North Atlantic Program (OSNAP) presented their findings this week at an ocean science meeting in Portland, Oregon. With nearly two years of data from late 2014 to 2016, the team found that the strength of the Atlantic Meridional Overturning Circulation -- which pumps warm surface water north and returns colder water at depth -- varies with the winds and the seasons, transporting an average of roughly 15.3 million cubic metres of water per second. The measurements are similar in magnitude to those from another array called RAPID, which has been operating between Florida and the Canary Islands since 2004. But scientists say they were surprised by how much the currents measured by the OSNAP array varied over the course of two years.

Comments

Re: OSNAP is an excellent name...

[rgbatduke]

And, if you read The Black Swan, by Nicholas Nassim Taleb, you will learn why even what you are doing -- predicting the market by assuming that it will behave tomorrow much like it behaved today (which is an excellent way to predict weather as well for up to three days) will one day cost you more money in a day than you've made in all the transactions up to date -- rare, large, expensive fluctuations in the market that do NOT conform to the usual Gaussian, linear regression, simple extrapolation models are a feature of chaotic systems and their kin.

One of the many things I dislike intensely about the way climate model results are presented and used -- and I am not making this up, you can read Chapter 9 in AR5 to verify -- is that they take a weather model, where weather models are where chaotic dynamics was discovered, tap it with a magic wand to call it a climate model instead, coarse grain it to where they can afford to run it (ignoring things like the actual Kolmogorov scale for the dynamics, the spatiotemporal scale where stepwise dynamics MIGHT actually integrate the problem you are trying to integrate), select model parameters -- many of them, the model space itself has a high dimensionality -- on heuristic grounds, making it simple to insert confirmation bias without even knowing it if you are building the model, select initial conditions that are more or less arbitrary because we do not KNOW the state of the Earth's climate system at a resolution anywhere close to that needed to initialize the model, then run it forward for as long as they want to/can afford to wait, tell themselves that they've reached some sort of "equilibrium" that means something relative to the Earth's climate state, make changes (like ramp up CO2) and run the model forward for as long as they can afford to.

Sometimes, of course, the Earth cools. Sometimes it warms. Sometimes it is in between. It's chaotic!

So then they AVERAGE all of those trajectories, and claim that the average is a prediction, projection, whatever, without ever actually acknowledging the width and variance of the range of outcomes.

This happens for ALL the many models in use. Many if not most of these models are not independent -- there are whole families of similar but not quite identical models all run by NASA GISS, for example. They then take ALL of the averages of ALL of the models -- without considering or eliminating the fact that multiply represented models get (in effect) more than one "vote" -- and superaverage them together and call that "the grand projection" because if they actually called it a prediction the gods of all science would smite them with lightning where they stand. Again, they ignore the considerable variance between all the model superaverages before they super-superaveraged them WITHOUT EVEN THINKING about how many actual RUNS contributed to the superaveraged results being super-superaveraged, so again a model with 10 runs counts as much as another model with 1000 in the statistical weighting.

The inclusion is also done without any reference to how successful the model(s) are. A model that hasn't come within three of its own standard deviations of the actual climate in its entire history is treated on the same basis as a model that has kept the actual climate within one standard deviation the entire time. This results all by itself in an enormous warming bias as the earth just hasn't warmed at anything like the rate the models overall have called for, and make it easy to then write a really scary summary for policy makers, leaving all of the actual warnings about the unbelievable travesty abuse of statistics that this is IN chapter 9 where nobody reads it or understands it unless they are in on the game.

I don't care for this because I actually do statistical analysis, statistical mechanics, predictive modeling, and so on, and this really, truly is horrific. Again, if you don't believe me, read chapter 9 in AR5. By the way, if anyone wants to argue, they can start by directing me to a paper wh