Guest Post by Willis Eschenbach
I haven’t posted here at WUWT in a bit. I’ve been preoccupied writing for my own blog, Skating Under The Ice. It’s a work in progress. However, my climate work continues. There’s a paper out about a year old, unfortunately paywalled, regarding precipitation with some interesting conclusions. [UPDATE: available here, h/t to commenter Taphonomic]. The paper is called Changes in annual precipitation over the Earth’s land mass excluding Antarctica from the 18th century to 2013, by W.A. van Wijngaarden and A. Syed. Here is what ScienceDirect lists as the key points of the paper:
Highlights
• Over 1½ million monthly precipitation totals observed at 1000 stations in 114 countries analysed.
• Data record much longer than 3 recent conflicting studies that analysed a few decades of data.
• No substantial difference found for stations located at northern, tropical and southern latitudes.
• No substantial difference found for stations experiencing dry, moderate and wet climates.
• No significant global precipitation change from 1850 to present.
The Abstract of the paper reads:
Precipitation measurements made at nearly 1000 stations located in 114 countries were studied. Each station had at least 100 years of observations resulting in a dataset comprising over 1½ million monthly precipitation amounts. Data for some stations extend back to the 1700s although most of the data exist for the period after 1850. The total annual precipitation was found if all monthly data in a given year were present.
The percentage annual precipitation change relative to 1961-90 was plotted for 6 continents; as well as for stations at different latitudes and those experiencing low, moderate and high annual precipitation totals. The trends for precipitation change together with their 95% confidence intervals were found for various periods of time. Most trends exhibited no clear precipitation change.
The global changes in precipitation over the Earth’s land mass excluding Antarctica relative to 1961-90 were estimated to be: -1.2 ± 1.7, 2.6 ± 2.5 and -5.4 ± 8.1% per century for the periods 1850-2000, 1900-2000 and 1950-2000, respectively. A change of 1% per century corresponds to a precipitation change of 0.09 mm/year.
Being a visual kind of guy, I converted the one sigma errors to 95% confidence intervals, and graphed up the trends:
As you can see, there is no statistically significant change in global rainfall, whether we look back fifty, a hundred, or a hundred and fifty years.
How are we to understand this? Well, to start with it is another example of the amazing stability of the global climate system.
However, I have to ask myself, given that in general a warmer world is a wetter world .. why so little change? This is often a valuable exercise, trying to understand a negative result.
One thing that comes to mind is that the effect may be there, but not be big enough to see. After all, the land has warmed up something like one degree Celsius or so since 1850, from something like 14°C to 15°C.
So … IF we assume that atmospheric water goes up linearly with increasing absolute temperature T, the temperature has gone up from about 287 Kelvin to 288 Kelvin, which is about a third of a percent … so we wouldn’t expect to see that.
On the other hand, IF water vapor goes up by T^4, then the increase is a bit larger. But it’s still only a percent and a half, which still would not be visible.
Finally, IF atmosphere water goes up from approximately zero at 0°C to 100% at 100°C … well, even then it’s only one degree, so that would still only be 1%.
So one possibility is that warmer is wetter, but the temperature change over the last century and a half is not large enough to cause an observable change in rainfall.
Another possibility is more subtle. As I write this, the wind outside is pushing the rain sideways at something like 20° from the vertical. This means that my rain gauge shown below, complete with a chambered nautilus shell in the window that we picked up on some beach “long ago and in another country”, that gauge is not recording the actual rainfall amount.
So … how much is the error? Well, at the simplest level it’s equal to the cosine of the angle of the vertical. If the rain is falling at 45° the gauge will read about 30% less than the true value. At 20° from the vertical like today, it only reads about 6% low. However, in reality of course it’s much more complicated. For example, for a given wind, smaller droplets are more likely to be blown away from the gauge. So light rains will have a greater error than heavy rains.
The total analysis is quite complex, depending on the exact shape of the rain gauge and how it is affected by the winds. There’s a fascinating analysis here.
Setting the details aside, it is clear that a slow secular change in either global average wind speed or global average raindrop size could easily make a 1% change in the long-term rainfall records …
In any case, it seems that the rumored droughts and floods due to rising global temperature are not visible in either the global or regional analyses, nor in the short, medium, or long-term analyses, nor in the tropic, subtropic, or polar analyses, nor in the dry, medium or wet locations. Here is their final conclusion:
Stations experiencing low, moderate and heavy annual precipitation did not show very different precipitation trends. This indicates deserts/jungles are neither expanding nor shrinking due to changes in precipitation patterns. It is therefore reasonable to conclude that some caution is warranted about claiming that large changes to global precipitation have occurred during the last 150 years.
Another alarmist myth run aground on an inconvenient reef of facts …
Best to all,
w.
PLEASE: In your comments, be so kind as to QUOTE THE EXACT WORDS YOU ARE REFERRING TO. This prevents misunderstandings regarding your subject.
REGARDING THE POST TITLE:
WHO’LL STOP THE RAIN—Creedence Clearwater Revival Band
Long as I remember the rain been comin’ down
Clouds of mystery pourin’ confusion on the ground.
Good men through the ages tryin’ to find the sun.
And I wonder, still I wonder, who’ll stop the rain.
I went down Virginia seekin’ shelter from the storm
Caught up in the fable I watched the tower grow
Five year plans and new deals wrapped in golden chains.
And I wonder, still I wonder, who’ll stop the rain.
Heard the singers playin’, how we cheered for more.
The crowd had rushed together tryin’ to keep warm.
Still the rain kept pourin’, fallin’ on my ears
And I wonder, still I wonder, who’ll stop the rain.
Willis:
Very nice post.
Here is the bit I question:
“So … how much is the error? Well, at the simplest level it’s equal to the cosine of the angle of the vertical. If the rain is falling at 45° the gauge will read about 30% less than the true value. At 20° from the vertical like today, it only reads about 6% low.”
If you make an adjustment for wind – are you not comparing apples (adjusted rain gauges) to oranges (older rain gauges but not adjusted for wind).
You either have to adjust all the precipitation data or none – but not just going forward (in my opinion).
If you adjust it all – it makes no difference, so why bother?
Here is the money-graph from the paper:
http://www.leif.org/research/RainFall-Global-Since-1700.png
Here is a synopsis of the above W. A. van Wijngaarden study from Journal of Hydrology (2015).
https://rclutz.wordpress.com/2016/05/07/data-vs-models-2-droughts-and-floods/
In my glorious days as a soil research scientist in Indonesia, I had access to daily rainfall data going back over 100 years for about 120 weather stations on the island of Java. I performed Markov Chain analysis on these data and found consistent rainfall patterns over the time period.
I am intrigued by the statement that wind would cause a lower amount of rainfall measured. Need to think about this. Walking through the rain does not make you less wet when there is wind and the amount falling per meter square should not change.
I am too lazy to look for the source but I recall that one of the assumptions about CAGW is constant relative humidity. ie. as the temperature increases more water is evaporated into the atmosphere. If that didn’t happen, the relative humidity would decrease.
The foundation of CAGW is that CO2 will cause warming which will, in turn, cause more water in the atmosphere which will, in turn, trap more heat. (ie. positive feedback)
If we don’t have more precipitation, that means we don’t have more evaporation because, over time they have to equal. True, the atmosphere would hold more water but that is a small amount compared with the amount of precipitation and evaporation.
I think we have here another proof that CAGW isn’t happening.
Thank you sir. Interesting. I was thinking of that same CCR song this morning on my way to work, as I often try to imagine new illustrations for folks who might be persuaded to question the climate “consensus” based on models. Consider this: no one is measuring and capturing a record of the rain that evaporates before it hits the ground. Who’ll stop THAT?
I’m confused by one thing though. The visual satellite data show a noticeable greening of some desert areas (Antarctica is a notable exception). Maybe there is no data gathered there (too remote?), or maybe the greening requires no excess water, simply the increase in CO2. Or perhaps there’s some other reason? I just would have thought that the greening would’ve required more than a simple increase in CO2.
“If the rain is falling at 45° the gauge will read about 30% less than the true value.” Consider a “gauge” of a square mile. Why would wind have such an effect? Why would a square inch gauge behave differently?
Hi Willis!
A 1 Celsius degree rise in temperature increases saturated water vapor pressure by about 7%, because vapor pressure is roughly an exponential function of temperature. See the section “Meteorology and climatology” at https://en.wikipedia.org/wiki/Clausius-Clapeyron_relation . Assuming relative humidity of 50%, a 1 degree rise in global temperatures would also increase water vapor by 7%. This is one reason for the importance of positive water vapor feedback in estimating climate sensitivity. OTOH, doubling CO2 from 300 to 600 ppmv produces about 0.6 degree warming (not counting feedbacks) because the accepted radiative forcing of 3.7 W/m^2 comes from computer calculations of increased absorption for a cloud-free troposphere. Clouds decrease the path length and the number of vibrationally excited CO2 molecules which are responsible for the increased absorption centered at 618 and 721 cm^-1 (see the Section “Spectral transitions”, Diagram 3, at http://www.barrettbellamyclimate.com/ and note the increased absorption between the green and blue curves in the MODTRAN computed spectra at https://en.wikipedia.org/wiki/Radiative_forcing ), so the accepted value of 1 degree warming (not counting feedbacks) is too high. A 0.6 degree rise in temperature increases water vapor by only 4%, not 7% (see tables of water vapor pressure in the Handbook of Chemistry and Physics). The reduced water vapor feedback corresponding to a reduced climate sensitivity (before feedbacks) means that positive feedback is no more than 50% (100% if one uses the formula for an infinite geometric series), and not the 200% that triples the accepted value of 1 degree to 3 degrees after feedbacks. When negative cloud feedbacks are included, the net feedback is going to be small, so the overall climate sensitivity (including water vapor and cloud feedbacks) is around 0.6-0.7 degrees, not 3 degrees. This explains why all the computer projections of CO2-driven warming are too high when compared to actual temperature data. I enjoy reading your articles, Willis.
This reminds me of “weather not climate”.
If the effect on rainfall of a change in temperature is too small for us to discern day-by-day then… look at the effect over decades.
So, we can’t see a change in measured rainfall on daily measurements. That doesn’t mean the change isn’t there. Absence of evidence isn’t evidence of absence.
Every observation looks like a nail through a hammer-o-scope.
Are deserts growing or shrinking?
Are we getting less return on fertiliser or not?
These are climate questions.
The link between measured weather and the effect of the average climate is not simple.
I’m still a sceptic. But this does not convince.
Willis, why do the error bars get larger as you move forward in time? Do we really know less about rainfall in 1950 than 1850? What am I missing?
It’s raining, it’s pouring
The climate is snoring.
Willis,
I think if you were to surround the top of the gauge with a tube say ~ 5x dia of collecting dish & same ht above rim of dish, that would remove turbulence from top of collecting dish & allow gravity to take charge of direction of water droplets.
You already have one side shielded by the house so rain coming from that direction will tend to fall verticaly-ish.
Full text of article available at: http://wvanwijngaarden.info.yorku.ca/files/2015/11/WvW-J.-Hydrology-2015.pdf
Besides the turbulence effect, the authors in the link provided pointed out that the streamlines move up and over the top of the rain gauge, carrying smaller raindrops with them. They therefore stated that the error increases with the amount of blockage. However, the total error for windspeeds of 2-3 mps for the Hellman gauge ranged only from 3-5%, and for windspeeds of 3-4 mps from 3-10%, with the high-end errors associated with the lowest rate of rainfall. I conclude first that rain gauge errors are relatively small and second that they would not affect trends unless windspeeds and/or rainfall rates are themselves trending.
Nice post Willis. I guess too, it’s you we have to thank for this latest news about Dr Judith Curry’s resignation. I remember you asking her many times to look at what the real world had to say about IPCC nonsense, and many ensuing debates.
It shows what integrity looks like in the scientific community. Let’s hope the dam breaks quickly in the era of President Trump!
https://pindanpost.com/2017/01/04/what-honesty-and-integrity-look-like-in-science/
Would evaporation rates also go up in a warmer world? Just sayin, yin yang……
The new cry will be, “Sure, it’s not changing on an annual average…but it’s being concentrated so that we get more extreme rainfall events and longer dry periods as opposed to the natural/perfectly flat trend,” lol.
So… we need to add a weather vane to rain collection cups – then the ambient wind will push the cup so the rain enters a ‘normal-to-wind’ surface area that remains constant no matter how much wind there is… 😀
Rain gauges meant to provide the most accurate measurements have a wind screen around them. Again, it’s because an *isolated* gauge has wind flow around it that tends to diverge and reduces the catch by the gauge….but has nothing to do with the angle at which the rain is coming down, which has no effect because the fall speed of the drops is still the same (avg. of 10 m/s, depending on drop size). This is Meteorology 101 stuff, folks.
How does the rain gage data compare to flood events? To stream flow data? Droughts? Some other rain dependent measurable thing not corrupted by other man made changes?
Just to link to this. This is the only CCR song I can play on the piano:
Another source of info, from 2016:
“Trends in Extreme Weather Events since 1900 – An Enduring Conundrum for Wise Policy Advice”
https://www.omicsgroup.org/journals/trends-in-extreme-weather-events-since-1900–an-enduring-conundrum-for-wise-policy-advice-2167-0587-1000155.pdf
Not gonna post details here. Rather a general reminiscence. My father was posted as commander of the 409th typhoon chasers off Guam starting 1948. He was a commnd pilot on those 24-36 hour flights in modified B-29s. Worst intercept resulted in a tail 18degrees off true. Bent by turbulence. He landed the plane anyway. It never flew again. Some of us been there, done that, at least by proxy of true heros.
Their error ranges are as large or larger than the results. Seems pretty meaningless to this layman. I don’t think this paper gives us any useful conclusion.