7 ms·

How does someone measure the average global temperature? And what does it means? The Wikipedia[0] has one interesting source[1] but what they show there isn't

by nzach 3y ago

How does someone measure the average global temperature? And what does it means? The Wikipedia[0] has one interesting source[1] but what they show there isn't an average. My guess would be that somehow satellites can measure the surface temperature and use that to make an average? I would be very happy if someone could some point me to some explanations about this process. [0] - https://en.wikipedia.org/wiki/Global_surface_temperature https://en.wikipedia.org/wiki/Global_surface_temperature [1] - https://www.ncei.noaa.gov/access/monitoring/monthly-report/global/202304/supplemental/page-3 https://www.ncei.noaa.gov/access/monitoring/monthly-report/g...

xvedejas 3y ago
Earth-observing satellites can measure surface temperature with so-called "thermal infrared" bands. I am unsure on the exact process typically used to go from that map to a global average, but that's a likely starting point.
raybb 3y ago
The source of the data is https://climatereanalyzer.org/clim/t2_daily/ https://climatereanalyzer.org/clim/t2_daily/ according to https://www.bloomberg.com/news/articles/2023-07-04/world-records-hottest-day-ever-on-july-3 https://www.bloomberg.com/news/articles/2023-07-04/world-rec... I wonder how a story like this gets picked up.
- tacker2000 3y ago
  Quote from that page : “ This page provides time series and map visualizations of daily mean 2-meter air temperature from the NCEP Climate Forecast System (CFS) version 2 (April 2011 – present) and CFS Reanalysis (January 1979 – March 2011). CFS/CFSR is a numerical climate/weather modeling framework that ingests surface, radiosonde, and satellite observations to estimate the state of the atmosphere at hourly time resolution onward from 1 January 1979. The horizontal gridcell resolution is 0.5°x0.5° (~ 55km at 45°N). Temperature anomalies are in reference to 1979–2000 climatology for each specific day of the year.”
amelius 3y ago
Next I want to learn how they measure Earth's entropy.
akiselev 3y ago
There are tens (if not hundreds) of thousands of land based weather stations, bouys and floats, weather balloons, radiosondes, observatories, and satellites around the world that form a lot of different sensor networks like Argo, ASOS, AWOS, etc. operated by a bunch of government agencies around the world. The World Meteorological Organization at the UN coordinates the sharing of all this data and there are too many players to list but the most important global datasets produced by the US are the GHCN v4 for land and ERSST v5 for sea. Each organization creates their own models and often compare with each other but most methods start by defining a grid that divides the planet, then use a model to estimate the average temperature for each cell given available sensor data. Many cells, especially in the middle of the ocean, don't have any direct measurements so have to be filled in using satellite based estimates or kriging (interpolation). They average over all of the grid cells in each hemisphere separately then apply some sort of formula to account for the fact that the northern hemisphere has significantly better coverage. It's not the most precise calculation and what most people are looking for is the rate of change over time using the same model and dataset (which are also versioned based on changes in methodology).
- fdsajkfdskl 3y ago
  [flagged]
  - comfypotato 3y ago
    Climate science has always acknowledged this. The rate of rise is unprecedented. This is based on non-man-made evidence we gather from the world around us. Concentration of pollen in layers of ice cores for example.
    - fdsajkfdskl 3y ago
      [flagged]
      - slater 3y ago
        > Then who are these kind of headlines (ie, "Hottest day ever!") for? Given they are factually incorrect how so?
jbotz 3y ago
The same way you take any temperature... by sampling and then averaging over the samples. Obviously you need to have enough well distributed samples to get a reasonable accuracy, but this is all an old hat. Btw., I suspect you get that "I don't know what average temperature even means" from some nonsense Ivar Giaever spouted at a presentation that's on YouTube. The simple answer to that is that average temperature means the same thing as temperature, because temperature is always an average that we can only arrive at by sampling (since it represents the average kinetic energy of a volume of particles and you can't meassure each particle's energy individually). Ivar Giavever knows that, but he also knows that people will believe almost anything that comes from him because he has a Nobel Price. After all, his autobiography is titled "I'm the smartest man I know".
wrycoder 3y ago
There are two sets of satellites that measure the temperature of the global troposphere, and have been doing so for over forty years. The data from one of them, UAH, is analyzed here.[0] It is increasing at 0.13C per decade, averaged over 44 years. [0] https://www.drroyspencer.com/2023/06/uah-global-temperature-update-for-may-2023-0-37-deg-c/ https://www.drroyspencer.com/2023/06/uah-global-temperature-...
- anonymousiam 3y ago
  Interesting that until Remote Sensing Systems revised their historical data six years ago, no warming trend was evident. How could they have been so wrong for such a long time? https://www.carbonbrief.org/major-correction-to-satellite-data-shows-140-faster-warming-since-1998/ https://www.carbonbrief.org/major-correction-to-satellite-da... https://www.drroyspencer.com/2017/07/comments-on-the-new-rss-lower-tropospheric-temperature-dataset/ https://www.drroyspencer.com/2017/07/comments-on-the-new-rss...
  - defrost 3y ago
    RSSv3 showed warming, it's there on the graph. RSSv4 corrected against ground data (calibration) and normalised for time drift and incident angle also showed warming .. more warming as prior post processing results had an incorrect transfer function. As a point of order, their raw historic data has not been altered - instruments rarely directly produce values of interest, they must be calibrated against other sources to provide a transfer function between raw instrument product and final values of interest. Bimetallic strips don't measure "tempreture" - they show the expansion of metals in tension. They must be calibrated to show a "tempreture" relative to amount of bend. Traditional aircraft altimeters don't directly measure height above ground, they are aneroid barometers that reflect the atmospheric pressure from a static port outside the aircraft. Calibration and barometric pressure on the ground are required to form a transfer function for height.
  - jfengel 3y ago
    It's so damn weird that Dr. Spencer -- whose job is producing the data that shows climate change -- is a climate denialist. The link in the grandparent comment, showing .12C/decade, is the very same site with the denialist claim you link to. It's not just him, either. His colleague John Christy is also responsible for the UAH lab, and also a climate denialist. They are different kinds of denialist. Spencer is in the "It's real but it's not our fault" camp, while Christy is in the "It's real and our fault but not important" camp. They are, however, very popular with the camp of "it's not real", and I've never seen them reject those claims, even though they are the ones responsible for the data that prove it is real.

by nzach 3y ago

xvedejas 3y ago

Earth-observing satellites can measure surface temperature with so-called "thermal infrared" bands. I am unsure on the exact process typically used to go from that map to a global average, but that's a likely starting point.

raybb 3y ago

The source of the data is https://climatereanalyzer.org/clim/t2_daily/ https://climatereanalyzer.org/clim/t2_daily/ according to https://www.bloomberg.com/news/articles/2023-07-04/world-records-hottest-day-ever-on-july-3 https://www.bloomberg.com/news/articles/2023-07-04/world-rec... I wonder how a story like this gets picked up.

tacker2000 3y ago
Quote from that page : “ This page provides time series and map visualizations of daily mean 2-meter air temperature from the NCEP Climate Forecast System (CFS) version 2 (April 2011 – present) and CFS Reanalysis (January 1979 – March 2011). CFS/CFSR is a numerical climate/weather modeling framework that ingests surface, radiosonde, and satellite observations to estimate the state of the atmosphere at hourly time resolution onward from 1 January 1979. The horizontal gridcell resolution is 0.5°x0.5° (~ 55km at 45°N). Temperature anomalies are in reference to 1979–2000 climatology for each specific day of the year.”

amelius 3y ago

Next I want to learn how they measure Earth's entropy.

akiselev 3y ago

There are tens (if not hundreds) of thousands of land based weather stations, bouys and floats, weather balloons, radiosondes, observatories, and satellites around the world that form a lot of different sensor networks like Argo, ASOS, AWOS, etc. operated by a bunch of government agencies around the world. The World Meteorological Organization at the UN coordinates the sharing of all this data and there are too many players to list but the most important global datasets produced by the US are the GHCN v4 for land and ERSST v5 for sea. Each organization creates their own models and often compare with each other but most methods start by defining a grid that divides the planet, then use a model to estimate the average temperature for each cell given available sensor data. Many cells, especially in the middle of the ocean, don't have any direct measurements so have to be filled in using satellite based estimates or kriging (interpolation). They average over all of the grid cells in each hemisphere separately then apply some sort of formula to account for the fact that the northern hemisphere has significantly better coverage. It's not the most precise calculation and what most people are looking for is the rate of change over time using the same model and dataset (which are also versioned based on changes in methodology).

fdsajkfdskl 3y ago
[flagged]
- comfypotato 3y ago
  Climate science has always acknowledged this. The rate of rise is unprecedented. This is based on non-man-made evidence we gather from the world around us. Concentration of pollen in layers of ice cores for example.
  - fdsajkfdskl 3y ago
    [flagged]
    - slater 3y ago
      > Then who are these kind of headlines (ie, "Hottest day ever!") for? Given they are factually incorrect how so?

jbotz 3y ago

The same way you take any temperature... by sampling and then averaging over the samples. Obviously you need to have enough well distributed samples to get a reasonable accuracy, but this is all an old hat. Btw., I suspect you get that "I don't know what average temperature even means" from some nonsense Ivar Giaever spouted at a presentation that's on YouTube. The simple answer to that is that average temperature means the same thing as temperature, because temperature is always an average that we can only arrive at by sampling (since it represents the average kinetic energy of a volume of particles and you can't meassure each particle's energy individually). Ivar Giavever knows that, but he also knows that people will believe almost anything that comes from him because he has a Nobel Price. After all, his autobiography is titled "I'm the smartest man I know".

wrycoder 3y ago

There are two sets of satellites that measure the temperature of the global troposphere, and have been doing so for over forty years. The data from one of them, UAH, is analyzed here.[0] It is increasing at 0.13C per decade, averaged over 44 years. [0] https://www.drroyspencer.com/2023/06/uah-global-temperature-update-for-may-2023-0-37-deg-c/ https://www.drroyspencer.com/2023/06/uah-global-temperature-...

anonymousiam 3y ago
Interesting that until Remote Sensing Systems revised their historical data six years ago, no warming trend was evident. How could they have been so wrong for such a long time? https://www.carbonbrief.org/major-correction-to-satellite-data-shows-140-faster-warming-since-1998/ https://www.carbonbrief.org/major-correction-to-satellite-da... https://www.drroyspencer.com/2017/07/comments-on-the-new-rss-lower-tropospheric-temperature-dataset/ https://www.drroyspencer.com/2017/07/comments-on-the-new-rss...
- defrost 3y ago
  RSSv3 showed warming, it's there on the graph. RSSv4 corrected against ground data (calibration) and normalised for time drift and incident angle also showed warming .. more warming as prior post processing results had an incorrect transfer function. As a point of order, their raw historic data has not been altered - instruments rarely directly produce values of interest, they must be calibrated against other sources to provide a transfer function between raw instrument product and final values of interest. Bimetallic strips don't measure "tempreture" - they show the expansion of metals in tension. They must be calibrated to show a "tempreture" relative to amount of bend. Traditional aircraft altimeters don't directly measure height above ground, they are aneroid barometers that reflect the atmospheric pressure from a static port outside the aircraft. Calibration and barometric pressure on the ground are required to form a transfer function for height.
- jfengel 3y ago
  It's so damn weird that Dr. Spencer -- whose job is producing the data that shows climate change -- is a climate denialist. The link in the grandparent comment, showing .12C/decade, is the very same site with the denialist claim you link to. It's not just him, either. His colleague John Christy is also responsible for the UAH lab, and also a climate denialist. They are different kinds of denialist. Spencer is in the "It's real but it's not our fault" camp, while Christy is in the "It's real and our fault but not important" camp. They are, however, very popular with the camp of "it's not real", and I've never seen them reject those claims, even though they are the ones responsible for the data that prove it is real.