Showing posts with label alert. Show all posts
Showing posts with label alert. Show all posts

Tuesday, March 19, 2019

Stronger Extinction Alert

The February 2019 temperature is in line with an earlier analysis that 2019 could be 1.85°C above preindustrial and that a rapid temperature rise may take place over the next few years, as illustrated by the image on the right.

Let's walk through the steps once more.

Baseline adjustment

The combination image below shows that the February 2019 temperature was 0.93°C above a 1951-1980 baseline (left) and 1.21°C above a 1885-1915 baseline (right), a difference of 0.28°C.

In other words, when using a baseline that is centered around 1900, the data should be adjusted by 0.28°C. In the image below, the gold graph uses 1951-1980 as baseline and two linear trend are added, one using data starting in 1880 (gold) and one using data starting in 1900 (blue).

Both linear trends are out of line with the recent temperature rise, the gold trend even more so than the blue trend, illustrating that starting a linear trend from an earlier year can make an analysis worse.

As said, if we want to use a baseline that is centered around 1900, the data should be adjusted by 0.28°C, and this is what the green graph does. A 4th-order polynomial trend is added that lines up perfectly with zero at the year 1900.

Further adjustment is needed for a 1750 baseline, which better reflects preindustrial as in the Paris Agreement. As discussed in an earlier post, this could result in an additional adjustment of 0.3°C.

Higher Arctic temperature

Furthermore, have another look at above maps. Much of the extreme anomalies are in line with changes to the Jet Stream, as also illustrated by the insert. More cold air escaping the Arctic and more warm air entering the Arctic are both speeding up Arctic warming. In the map on the right, much of the Arctic is left grey, since no data are available for the Arctic around 1900, but the Arctic should not be left out of the picture and adding a further 0.1°C adjustment seems appropriate to better include the Arctic.

Air temperature over oceans

Finally, the NASA temperatures for oceans are the surface temperatures of the water, but it makes more sense to use air temperatures close to the water, which likely adds a further 0.1°C. This adds up a total adjustment of 0.78°C as applied in the red graph, which also has an 8th-order polynomial trend added.

Which trendline works best?

How appropriate is it to apply an 8th-order polynomial trend to climate data? Have another look at above graphs and consider that in the gold graph, R²=0.687 for the gold linear trend (1880-Feb 2019 data) and R²=0.752 for blue linear trend (1900-Feb 2019 data), while in the green graph, R²=0.812 for the dark green 4th-order polynomial trend, and in the red graph, R²=0.828 for the pink 8th-order polynomial trend. In other words, the pink trend better follows the ups and downs of the data than the lower-order polynomial trend, and it does so much better than the linear trends that both are clearly unrealistic in an analysis of warming acceleration.

Selecting the axes

Is warming accelerating? Trend analysis that uses data going back many years can only be part of the picture; it's also important to anticipate changes that loom in the near future. When taking the many feedbacks, tipping points and further warming elements more fully into account, warming could accelerate even more strongly than depicted in the red trend in the graph at the top.

In the 'Extinction Alert' graph at the top,  the vertical axis is cut off at 5°C, since life on Earth will already have disappeared by then (see box on the right), but when looking at near-term human extinction, 3°C will likely suffice.

How soon could 3°C warming be reached? The 'Extreme Alert' image below looks at data over the past decade, and a fifth-order polynomial trend (red) shows how warming could cross 3°C as early as next year.

How could such a scenario eventuate?

In such a rapid warming scenario:
  1. a stronger-than-expected El Niño would contribute to
  2. early demise of the Arctic sea ice, i.e. latent heat tipping point + 
  3. associated loss of sea ice albedo, 
  4. destabilization of seafloor methane hydrates, causing eruption of vast amounts of methane that further speed up Arctic warming and cause 
  5. terrestrial permafrost to melt as well, resulting in even more emissions, 
  6. while the Jet Stream gets even more deformed, resulting in more extreme weather events
  7. causing forest fires, at first in Siberia and Canada and
  8. eventually also in the peat fields and tropical rain forests of the Amazon, in Africa and South-east Asia, resulting in 
  9. rapid melting on the Himalayas, temporarily causing huge flooding, 
  10. followed by drought, famine, heat waves and mass starvation, and
  11. collapse of the Greenland Ice Sheet.

Even when adding a rather inappropriate linear trend (as done in the 'Extreme Alert' image, in blue),  warming still looks set to cross 2°C by 2026 in the Extreme Alert image, but as the chart below shows, there could be a rise of as much as 18°C by 2026.

[ from an earlier post ]
The situation is dire and calls for comprehensive and effective action, as described at the Climate Plan.


• Co-extinctions annihilate planetary life during extreme environmental change, by Giovanni Strona and Corey Bradshaw (2018)

• How much warming have humans caused?

• Extinction

• A rise of 18°C or 32.4°F by 2026?

• Climate Plan

Thursday, February 7, 2019

Extinction Alert

Above image confirms an earlier analysis that it was 1.73°C (or 3.11°F) warmer than preindustrial in 2018. The image also shows that it could become 1.85°C (or 3.33°F) warmer in 2019.

This according to the non-linear trend (red line) that follows from the data and also follows the data better than the blue linear trend, which also follows from the data, but is out of line with the recent temperature rise.

Data are adjusted for a number of reasons. The first reason is a baseline issue. At the Paris Agreement, nations pledged to ensure that the temperature rise would not cross 1.5°C above preindustrial. Accordingly, data should reflect a 1750 baseline. The default baseline for the NASA Land+Ocean Temperature index (L-OTI) is 1951-1980. The above image features two maps, one showing the 2018 temperature rise compared to 1951-1980 (left) and another map showing the 2018 temperature rise compared to 1885-1915 (right). The difference is 0.25°C. In other words, using 1900 as a baseline would require a 0.25°C adjustment.

That figure of 0.25°C is conservative, firstly because 2018 was a La Niña year. Furthermore, as above image illustrates, the period from 1900 to 1920 was almost 0.3°C below 1951-1980. Anyway, this conservative figure of 0.25°C is used in this analysis. Additional adjustment of the data is needed, in order to reflect a 1750 baseline. The total baseline adjustment could add up to as much as 0.55°C, as discussed in an earlier post.

Furthermore, the large grey area in the Arctic on above map on the right reflects a lack of measurements in the Arctic that go back to 1900. Simply excluding those data would downplay the temperature rise, since temperatures have been rising faster in the Arctic than in the rest of the world. An additional adjustment of 0.1°C therefore seems appropriate.

Finally, NASA L-OTI data are for air temperatures over land and for sea surface water temperatures for oceans. To get an idea how much the temperature of the atmosphere has risen close to the surface, it makes more sense to use air surface temperature over oceans, rather than sea surface water temperatures, resulting in another additional adjustment of 0.1°C.

The total adjustment adds up to 0.75°C, resulting in the graph below.

The final step in this analysis is a projection into the future. In the image at the top, the trend is extended to the year 2033, but the vertical axis doesn't go beyond 5°C warming. Why 5°C? A recent study looked at plant temperature tolerances and concluded that extinction will already occur far earlier than when upper tolerance levels were reached for individual species, since "loss of one species can make more species disappear (a process known as ‘co-extinction’), and possibly bring entire systems to an unexpected, sudden regime shift, or even total collapse. There was a small group of species with large tolerance limits and remarkable resistance to environmental change, but even they could not survive co-extinctions. In fact, their extinction was abrupt and happened far from their tolerance limits and close to global biodiversity collapse at around 5°C of heating."

Importantly, the image at the top doesn't even depict the worst-case scenario, in the sense that the non-linear trend merely follows from the data, i.e. it doesn't take into account tipping points such as abrupt disappearance of the Arctic sea ice or sudden eruptions of methane from the seafloor of the Arctic Ocean.

A rapid 5°C rise could occur if an influx of warm salty water triggered methane eruptions from the seafloor of the Arctic Ocean. Combined with snow and ice loss, it could rapidly raise temperatures by 1.5°C, which increases water vapor. If cloud feedback is strongly positive, water vapor feedback can lead to 3.5 times as much warming, so these warming elements alone could cause 5°C warming within years. And then, of course, there are further warming elements.

The situation is dire and calls for comprehensive and effective action, as described in the Climate Plan.


• Co-extinctions annihilate planetary life during extreme environmental change, by Giovanni Strona and Corey Bradshaw (2018)

• National Aeronautics and Space Administration (NASA), Goddard Institute for Space Studies (GISS), Surface Temperature Analysis, Land+Ocean Temperature index (L-OTI)

• As El Niño sets in, will global biodiversity collapse in 2019?

• How much warmer is it now?

• How much warming have humans caused?

• IPCC seeks to downplay global warming

• Climate Plan

• Extinction