August 2016 was the warmest August in 136 years of modern record-keeping, according to a monthly analysis of global temperatures by scientists at NASA’s Goddard Institute for Space Studies (GISS).
Although the seasonal temperature cycle typically peaks in July, August 2016 wound up tied with July 2016 for the warmest month ever recorded. August 2016’s temperature was 0.16 degrees Celsius warmer than the previous warmest August (2014). The month also was 0.98 degrees Celsius warmer than the mean August temperature from 1951-1980.
“Monthly rankings, which vary by only a few hundredths of a degree, are inherently fragile,” said GISS Director Gavin Schmidt. “We stress that the long-term trends are the most important for understanding the ongoing changes that are affecting our planet.” Those long-term trends are apparent in the plot of temperature anomalies above.
The record warm August continued a streak of 11 consecutive months (dating to October 2015) that have set new monthly temperature records. The analysis by the GISS team is assembled from publicly available data acquired by about 6,300 meteorological stations around the world, ship- and buoy-based instruments measuring sea surface temperature, and Antarctic research stations. The modern global temperature record begins around 1880 because previous observations didn’t cover enough of the planet.
+ For more information on NASA GISS’s monthly temperature analysis, visit: data.giss.nasa.gov/gistemp.
+ For more information about how the GISS analysis compares to other global analysis of global temperatures, visit:
+ To learn more about climate change and global warming, visit:
Related Reading in the News
+ Mashable: Earth sets record for hottest August, extending warm streak another month
+ XKCD: A Timeline of Earth’s Average Temperature
+ Climate Central: August Ties July as Hottest Month Ever on Record
Is the specific data table you used available? (I would like to see if I can create a similar graph.)
I glanced at the tables in http://data.giss.nasa.gov/gistemp/, and the values in the tables there don’t seem to have that seasonal curve.
The seasonal adjustments, which come from MERRA-2, can be found here: http://data.giss.nasa.gov/gistemp/faq/merra2_seas_anom.txt
So if I take data from either:
Then subtract seas_anom from http://data.giss.nasa.gov/gistemp/faq/merra2_seas_anom.txt, I get data that peak at 3, not at 2.
What am I doing wrong?
The standard GISTEMP baseline is 1951-1980. Since the MERRA-2 seasonal data this chart used is based on a 1980-2015 baseline, the GISTEMP data needs to be updated to match. The resulting difference for each month is given below:
Month Baseline update
Is it possible to get the data table with the seasonal adjustments already factored into the data? I would love to use this data in a musical composition to create awareness of global warming…
Great work! Thanks for that and the explanations. However, although a bit familiar with the recalculation of those base periods, I am not sure how you got to the baseline updates. As far as I understood I need to calculate the average for the years 1980-2015 of the original data (http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts+dSST.csv), which gets me these for Jan to Dec: 0.48 0.49 0.51 0.47 0.45 0.43 0.43 0.44 0.45 0.46 0.46 0.45
Now, I need to subtract these obtained values from the original data, for each year and month. (Not quite sure if this needs to be subtracted or added, though.)
Now, there is some difference between my baseline updates and yours. Would you be able to explain?
Thanks for your advice. I would very much appreciate your help!
In 1990 NASA,GISS reported that the entire year was 0.8°C above the long term mean of 59*F (1951-1980), which translates to 15.44*C Presumably, July and or August were higher in 1990? How warm was this past August? They say it was 0.98*C warmer.
“The month also was 0.98 degrees Celsius warmer than the mean August temperature from 1951-1980.”
seems relevant and appropriate.
One question I have is about the accuracy and precision of the older data from the set. There’s often a question (sometimes legitimately asked) about how much weight can be given to the precision of a temperature reading from 100 years ago. How reliable and reliably precise are these measurements, particularly the older ones?
I get the recent focus on warmer months due to global warming theory, but Why is it that the winter months have been colder than the mean?
I have the exact same question… How can that graph be showing below the average from Jan to Mar, and from late Sep to Dec – even in recent years…
In fact why is it always below the average even during the referenced years. It just doesn’t make sense. Say… If we are soaring above the average months after months since October 2015… THen why is January 2016 below the average. Or you did not pic the right graph to present what you wanted to present.
If i take the average of the last 136 january.. pretty sure Jan 2016 would show above the 0 line isn’t ? And the same since october 2015.
Because of the uneven distribution of landmasses between the northern and southern hemispheres, global average temperature is higher during the northern hemisphere summer than during the southern hemisphere summer. Hence, months should only be compared to the same months of different years to understand long term trends.
It appears the 0 line is the global mean for all data, and isn’t corrected to be a mean for each month. That would be an interesting graph to see.
If the dataset contains data that was collected between 1980 and today by the MERRA-2 method, where did the before 1980 data come from? Did the 1880 data come from Sattelites as well?
I’ve seen these plots a few times now and it’d be nice to have someone explain why the pattern follows a northern hemisphere temperature signature, i.e., hot in July and cold in January. Shouldn’t these lines be flatter as the southern hemisphere experiences a signature opposite of the north?
Also, how do the bounds on error change as we go further back? And shouldn’t the anomaly be monthly as opposed to annual?
If I recall correctly, this is due to the Northern Hemisphere containing the bulk of the world’s land mass. Vast oceans moderate temperature changes, so looking at the global average temperature skews toward the Northern Hemisphere’s seasonality.
I’ve seen these plots a few times now and it’d be nice to have someone explain why the pattern follows a northern hemisphere temperature signature, i.e., hot in July and cold in January. Shouldn’t these lines be flatter as the southern hemisphere experiences the inverse pattern?
Also, how do the bounds on error change as we go further back?
And shouldn’t the anomaly be a sown as a monthly anomaly?
The southern hemispheres inverse pattern is already included, and already flattens the data. If you look at the magnitude of change between summer and winter, it’s only about 4 degrees. The difference between winter and summer temperatures in any latitude outside the tropics is much higher than 4 degrees, hence we can see that the data is already flattened.
By inference, you can assume that southern hemisphere summers and winters are 4 degrees less extreme than their northern hemisphere inverse counterparts. This is primarily because of the uneven distribution of landmasses between the two hemispheres and how that affects global climate patterns.
I would also like to see a chart showing the difference from the monthly means.
Has there been any published study that has compared the measurement methods over time, the number of data points over time, or how the area around the measurement sites has changed over time. Over time many rural areas have changed into urban areas. For example there was a headline about a year ago that the temperature at Regan Airport in DC had exceeded the record temperature in 1850. In 1850, large areas of asphalt, concrete, and jet exhaust did not exist. A few years ago I drove from Atlanta to where I live, a distance of about 160 miles, observing the Bradford Pear blooms in the areas where there was a lot of asphalt but this same species were bare of blooms in the rural areas.
I have often wondered the same thing Wilson. The official Melbourne weather station had to be moved recently for this very reason. From a news article at the time….
“Mr Missen said when it opened in 1908, moving from the Observatory at the Royal Botanic Gardens, the site was surrounded by gardens, low-rise buildings and horses and carts on the roads.
But the growing number of skyscrapers and volume of traffic have been deemed not conducive to reflecting the weather conditions of greater Melbourne. Because of blockage from skyscrapers, from 2009 to 2013 wind readings were taken from Essendon Airport.”
Would be interesting to know if this happens with other stations.
“…the Warmest August in 136 Years”
gives me pause for wonder:
“In 136 years of modern record-keeping, July 2016 was the warmest July
according to a monthly analysis of global temperatures by scientists at
NASA’s Goddard Institute for Space Studies (GISS).”
“The combined average temperature over global land and ocean surfaces
for May 2016 was the highest for May in the 137-year period of
Yet, in central New York (Syracuse area)
each of these months would seem to have been
cooler this year than in any previous year
in my memory which begins
with winter 1948-49.
What is the explanation for this condition?
The unfortunate part of your visualization lies in the use of global temperature data versus a more targeted sample area. The global approach greatly skews the data coming from areas closer to the poles toward the mean. A more interesting and I would argue telling graph would be one that tracks temperature data for the areas lying north and south of the 40 deg parallel. The temperature change in those areas for the period of record is orders of magnitude greater than it is for the global mean, and offers a much better illustration of what is actually happening on the ground with respect to global warming.
The GIF image says 1980-2015. Shouldn’t it be 1880-2015?
The 1980-2015 note is in regards to the baseline for calculating the anomalies. Each of the monthly measurements from 1880 onward are compared to that baseline.
I think the most important question is the bias of variation in the meteorological stations. If one station is a periphery of a city, it will has increasing temperatures across the years. What is the ratio between isolated stations and (now) urban ones?
Graph in Comic form. Pay close attention to the tool-tip when you hover over it.
xkcd: Earth Temperature Timeline: http://xkcd.com/1732/
The problem is that there are too many people on this earth. This causes a lot of problems, not just global warming. For example, by 2050 we won´t have enough food worldwide. Nice graph to world population: https://upload.wikimedia.org/wikipedia/en/3/3a/Human_population_growth_from_1800_to_2000.png
I am ready to help.
I am in communication with a community group engaged in exploring Climate Change, and would like to email certain graphs, including the animated one of successive years of global temperatures to the e-list. Is this possible? I don’t see the “email to a friend” option.
Of course, October, November, and December 2015 had the highest monthly temperatures on record, as did January-May and July and August of 2016.
Basis: GISS-NASA database
Hi.. Thank you for the nice info. I was wondering what kind of tools / APIs did you use to render the graph data and into this image?
I´m living in Brasil and make inside the Luteran Church “lectures”/presentations about Climat Change. I would like to use this aplication with the temperatures of agust, but I don´t know how to copy THE ANIMATION in a power-point presentation!
Could you help me?