Remember the Nature sub-paper on September 14th, “Life on Venus”? Astronomers have observed signs of hydrogen phosphate in Venus’ “atmosphere”, strong evidence of life on venus. This is the big news that has rocked the astronomical world, and many people can’t believe it: Hydrogen phosphate and life don’t necessarily have a connection, do they?
However, after careful study of the paper, the scientists can not sit still: this paper, from the beginning of data processing problems ah!
Questions ensued, and several teams around the world, including NASA, questioned the paper.
NASA’s scientists studying life, planets, and more are “co-authors” Nature: Ask the authors to consider correcting or withdrawing the manuscript.
All questions point to a problem: the so-called hydrogen phosphate signal is overfitted.
The three teams questioned, from the 12th-order polynthic used to validate the conclusions to the way the data was viewed, all “bombarded”.
They said their results from the data analysis were not consistent with the paper.
In other words, the conclusion that hydrogen phosphate was found in Venus’ atmosphere may be fundamentally wrong.
What the hell is going on? Let’s have a look together.
How was hydrogen phosphate observed on Venus?
First, look at the sensational Nature Astronomy paper itself.
The international team, led by astronomer Jane Greaves, first discovered hydrogen phosphate in Venus’ atmosphere using the James Clerk Maxwell Telescope (JCMT) in Hawaii.
They have since confirmed the results at chile’s Atacama Mm/Sub millimeter Array (ALMA) radio telescope.
How did Earthlings hundreds of millions of kilometers apart know there was hydrogen phosphate on Venus?
It turns out that chemical molecules absorb electromagnetic waves at certain wavelengths, like the “ID card” of this substance. If we analyze the waves passing through the planet and find out which ones have been absorbed, we can speculate that there is some kind of gas on the planet.
Hydrogen phosphate, for example, absorbs electromagnetic waves at a frequency of 267GHz.
As a result, scientists found depressions at this frequency in signals received by the JCMT and ALMA, and concluded that hydrogen phosphate existed on Venus.
And based on the size of the depression, they calculated that the concentration of hydrogen phosphate in Venus’ atmosphere was two hundred millionths.
However, to draw conclusions from the observations, it is not as easy as it says.
Because of the Earth’s atmosphere, the structure of the telescope itself and so on, electromagnetic waves will inevitably be affected by noise, with its own jitter. If the jitter is too large, the noise may even drown the signal.
Martin Cordiner, an astrophysicist at NASA’s Goddard Space Flight Center, points out that the problem is particularly serious when using telescopes as powerful as ALMA to observe bright objects like Venus.
In response, the Greaves team first excluded all very long baseline interferographic results less than 33m in length before ALMA data imaging. Because the shorter the interference baseline, the greater the noise effect in the signal.
In addition, NASA found that Greaves’ team used polynthic equations to fit noise and then remove it from the data.
The simplest can be a first-order polynthic equation, i.e. y-mx-b. The second-order polynthic equation is in the form of y-m0x2-m1x-b, the n-order polynthic equation, and so on.
The original polynthic fit is a regular operation, nothing. But the Greaves team used the 8th-order polynthropomorly to fit the data from the JCMT telescope, and the ALMA telescope data, they used the 12th-order polynthropomorly!
Is the way the data fits too crazy?
To sum up the contents of the paper briefly, the inference of the possibility of life on Venus is as follows: (1) hydrogen phosphate was discovered through the JCMT telescope, and (2) the observations were confirmed through the ALMA telescope.
The first were four astronomers from Leiden University in the Netherlands who found problems with the way ALMA’s observations were processed.
Fire-eyed them, at a glance in the processing (spectrum band part) of the noise of the 12th-order polynthic.
12th-order polynthon, why is it so outrageous? This is about fitting the experimental curve.
In order for the curve to be as close to all experimental data as possible, it is easier to achieve by selecting a higher-order polynthic to fit. But the higher the order of the polynome, the more the curve oscillates, deviating from the real situation.
For example, the experimental data below, would have been near a straight line, if fitted with a high-order polynthic, there will be many peaks and valleys.
The famous mathematician von Neumann famously said:
Give me four parameters, I can fit an elephant, give me five parameters, I can make the elephant nose shake.
It may sound crazy, but it’s true, and it’s been done with papers (the picture below shows an elephant’s nose swinging with one parameter):
Four parameters have been able to achieve this effect, the 12th-order polynthic sounds outrageous… But astronomers did it again, in two main ways:
(1) Reapply the same noise reduction method to the Venus data collected by ALMA
(2) Using this noise reduction method, the other parts of the Venus spectrum are filtered for noise
From the results, the resulting spectral data not only does not conform to the Gaussus distribution, but also a little worse …
Also, if the baseline is selected differently, the statistical results may be different.
Astronomers at Leiden University have also processed data that show that hydrogen phosphate is absorbed at a frequency of 2 degrees, below a statistically significant universal threshold, and therefore does not indicate abnormal levels of hydrogen phosphate in Venus’ atmosphere.
In addition, hydrogen phosphate found in this paper has one prerequisite: it was found in Venus’ atmosphere.
The Paris Observatory’s Thérèse Encrenaz recorded observations from Venus for three years 2012-2015, which showed no sign of hydrogen phosphate in the atmosphere.
While this does not mean that there is no hydrogen phosphate in higher places, Encrenaz argues that the paper’s claims are debatable.
Moreover, ALMA is only the first step.
Soon astronomers discovered that JCMT’s data processing seemed to have a problem, too.
An astronomer from the UK, who reprocessed the JCMT data, found that, like ALMA, a similar situation occurred in the JCMT, known as “false positives”.
Previously, Villanueva’s paper suggested that the absorption line of the JCMT could also be sulfur dioxide, but the paper argued that Venus’ detection spectrum could not prove whether sulfur dioxide or hydrogen phosphate had been absorbed.
This matter is still fermenting.
The clearer the truth, the clearer it becomes
The day after NASA issued its “invitation to withdraw” statement, it suddenly removed the part of the paper that called for it to be withdrawn and changed it to a more sober “data we measured that did not match the data shown in the paper”.
However, the matter, netizens have been hotly debated.
Some netizens to tease this 12th-order polynthic said: When I do graduation thesis, the 5th-order has been confused…
Other netizens thought of NASA’s paper on “Arsenic can replace phosphorus in DNA” published in Nature, but it ended up being punched in the face.
But some netizens think that statistics is indeed a metaphysical problem.
Byrne, a member of NASA’s Venus research team, says the paper has once again sparked enthusiasm for Venus, and its existence suggests that there is still little understanding of venus. He believes that:
“The only way to get these answers is to go to Venus.”
If the earlier discussion was whether the presence of hydrogen phosphate in the atmosphere would prove the existence of life on Venus, now even the presence of hydrogen phosphate itself needs more detection to prove it.
However, there is also a view that reasonable questioning is a good thing, “this is what science really looks like”.
Constant exploration and discovery will make the truth more clear.
Isn’t your paper also gradually perfected in step-by-step questioning and revision?