Due to momentum conservation, you can speed up any free rotating object by pulling its mass closer to rotation axis. Now look at any webservice that shows air traffic live. There you can see a tens of thousands airplanes which are airborne continuously at any given moment. This number was significantly reduced at the beginning of COVID-19 pandemic. The mass of those airplanes in total may look insignificant compared to mass of oceans or various Earth's internal structures. However planes fly high, typically at 12 km of altitude, so the change of altitude at which their mass is located has more prominent effect than if they were objects that basically move around roughly same altitude. Due to pandemic, airplanes were grounded, so Earth sped up to maintain its momentum.
I decided to calculate this out to see if it is remotely plausible.
If there are 30k commercial aircraft in the world, and they're all Airbus A380s weighing 500 tons each fully loaded, and they all fly at the equator at an altitude of 12km, then the earth's rate of rotation would be slower by 1 part in 10^17, and thus the day would be ~1ps longer than if they were all on the ground (at the equator).
Not enough to explain it, and 10^-17 is just on the edge of what is measurable with state of the art atomic clocks.
Thank you for your calculation. So you say the effect is neglible.
To be more correct, I think the mass of fuel (comparable to mass of a "dry" aircraft) lifted by every flight and left up in the atmosphere should be accounted, too. But it'd require lots of estimations and I think it would not change the result significantly.
Thank you for your effort. I hope you had some fun with it. :-)
How many Russian & Chinese atomic sources are included in TAI? <tinfoil-hat> What if a nation-state had an interest in forcing a negative leap second to cause GPS problems?</tinfoil-hat>
I doubt that's possible both due to the sheer number of labs and how they come to consensus about TAI, but the latest circular about TAI from the BIPM lists out all the countries and labs contributing
Or similar: become a high tier NTP server in key carrier colos, vetted over years of reliable service, and then using a Janus set of colluding NTP servers, slowly drift new updates, thereby changing the clock time of various bank and power company servers relying on the Janus group. This may not have much practical value since most TLS apps and similar clocked protocols deal with Windows desktops from the 1990s, which already have significant consumer clock drift. Some tolerate minutes of clock differences. And critical infrastructure tend to run their own NTP systems for risk management, better reliability and to avoid throttling.
Still, these ideas might be useful for a very boring screen play, involving a multi-decade plot to shave a few seconds off key clocks in a casino, enlisting the help of nefarious nation states, all so the protagonist crew could double spend a groupon deal for a plate of shrimp at an off-strip casino.
Due to momentum conservation, you can speed up any free rotating object by pulling its mass closer to rotation axis. Now look at any webservice that shows air traffic live. There you can see a tens of thousands airplanes which are airborne continuously at any given moment. This number was significantly reduced at the beginning of COVID-19 pandemic. The mass of those airplanes in total may look insignificant compared to mass of oceans or various Earth's internal structures. However planes fly high, typically at 12 km of altitude, so the change of altitude at which their mass is located has more prominent effect than if they were objects that basically move around roughly same altitude. Due to pandemic, airplanes were grounded, so Earth sped up to maintain its momentum.
I decided to calculate this out to see if it is remotely plausible.
If there are 30k commercial aircraft in the world, and they're all Airbus A380s weighing 500 tons each fully loaded, and they all fly at the equator at an altitude of 12km, then the earth's rate of rotation would be slower by 1 part in 10^17, and thus the day would be ~1ps longer than if they were all on the ground (at the equator).
Not enough to explain it, and 10^-17 is just on the edge of what is measurable with state of the art atomic clocks.
Thank you for your calculation. So you say the effect is neglible.
To be more correct, I think the mass of fuel (comparable to mass of a "dry" aircraft) lifted by every flight and left up in the atmosphere should be accounted, too. But it'd require lots of estimations and I think it would not change the result significantly.
Thank you for your effort. I hope you had some fun with it. :-)
How many Russian & Chinese atomic sources are included in TAI? <tinfoil-hat> What if a nation-state had an interest in forcing a negative leap second to cause GPS problems?</tinfoil-hat>
I doubt that's possible both due to the sheer number of labs and how they come to consensus about TAI, but the latest circular about TAI from the BIPM lists out all the countries and labs contributing
https://webtai.bipm.org/ftp/pub/tai/Circular-T/cirt/cirt.408
Or similar: become a high tier NTP server in key carrier colos, vetted over years of reliable service, and then using a Janus set of colluding NTP servers, slowly drift new updates, thereby changing the clock time of various bank and power company servers relying on the Janus group. This may not have much practical value since most TLS apps and similar clocked protocols deal with Windows desktops from the 1990s, which already have significant consumer clock drift. Some tolerate minutes of clock differences. And critical infrastructure tend to run their own NTP systems for risk management, better reliability and to avoid throttling.
Still, these ideas might be useful for a very boring screen play, involving a multi-decade plot to shave a few seconds off key clocks in a casino, enlisting the help of nefarious nation states, all so the protagonist crew could double spend a groupon deal for a plate of shrimp at an off-strip casino.
I'd watch that. Call it Oceans 86400.
When Qanon meets low level sysadmin...