Saturday, 4 April 2015

Irrigation and paint as reasons for a cooling bias

Irrigation pump in India 1944

In previous posts on reasons why raw temperature data may show too little global warming I have examined improvements in the siting of stations, improvements in the protection of thermometers against the sun, and moves of urban stations to better locations, in particular to airports. This post will be about the influence of irrigation and watering, as well as improvements in the paints used for thermometer screens.

Irrigation and watering

Irrigation can decrease air temperature by up to 5 degrees and typically decreases the temperature by about 1°C (Cook et al., 2014). Because of irrigation more solar energy is used for evaporation and for transpiration by the plants, rather than for warming of the soil and air.

Over the last century we have seen a large 5 to 6 fold global increase in irrigation; see graph below.

The warming by the Urban Heat Island (UHI) is real. The reason we speak of a possible trend bias due to increases in the UHI is that an urban area has a higher probability of siting a weather station than rural areas. If only for the simple reason that that is where people live and want information on the weather.

The cooling due to increases in irrigation are also real. It seems to be a reasonable assumption that an irrigated area again has a higher probability of siting a weather station. People are more likely to live in irrigated areas and many weather stations are deployed to serve agriculture. While urbanization is a reason for stations to move to better locations, irrigation is no reason for a station to move away. On the contrary maybe even.

The author of the above dataset showing increases in irrigation, Stefan Siebert, writes: "Small irrigation areas are spread across almost all populated areas of the world." You can see this strong relation between irrigation and population on a large scale in the map below. It seems likely that this is also true on local scales.

Many stations are also in suburbs and these are likely watered more than they were in the past when water (energy) was more expensive or people even had to use hand pumps. In the same way as irrigation, watering could produce a cool bias due to more evaporation. Suburbs may thus be even cooler than the surrounding rural areas if there is no irrigation. Does anyone know of any literature about this?

I know of one station in Spain where the ground is watered to comply with WMO guidelines that weather stations should be installed on grass. The surrounding is dry and bare, but the station is lush and green. This could also cause a temperature trend bias under the reasonable assumption that this is a new idea. If anyone knows more about such stations, please let me know.

From whitewash to latex paint

Also the maintenance of the weather station can be important. Over the years better materials and paints may have been used for thermometer screens. If this makes the screens more white, they heat up less and they heat up the air flowing through the Louvres less. More regular cleaning and painting would have the same effect. It is possible that this has improved when climate change made weather services aware that high measurement accuracies are important. Unfortunately, it is also possible that good maintenance is nowadays seen as inefficient.

The mitigation skeptics somehow thought that the effect would go into the other direction. That the bad paints used in the past would be a cooling bias, rather than a warming bias. Something with infra-red albedo. Although most materials used have about the same infra-red albedo and the infra-red radiation fluxes are much smaller than the solar fluxes.

Anthony Watts started a paint experiment in his back garden in July 2007. The first picture below shows three Stevenson screens, a bare one, a screen with modern latex paint and one with whitewash, a chalk paint that quickly fades.

Already 5 months later in December 2007, the whitewash had deteriorated considerably; see below. This should lead to a warm bias for the whitewash screen, especially in summer.

Anthony Watts:
Compare the photo of the whitewash paint screen on 7/13/07 when it was new with one taken today on 12/27/07. No wonder the NWS dumped whitewash as the spec in the 70’s in favor of latex paint. Notice that the Latex painted shelter still looks good today while the Whitewashed shelter is already deteriorating.

In any event the statement of Patrick Michaels “Weather equipment is very high-maintenance. The standard temperature shelter is painted white. If the paint wears or discolors, the shelter absorbs more of the sun’s heat and the thermometer inside will read artificially high.” seems like a realistic statement in light of the photos above.
I have not seen any data from this experiment beyond a plot with one day of temperatures, which was a day one month after the start, showing no clear differences between the Stevenson screens. They were all up to 1°C warmer than the modern ventilated automatic weather station when the sun was shining. (That the most modern ventilated measurement had a cool bias was not emphasized in the article, as you can imagine.) Given that Anthony Watts maintains a stealth political blog against mitigation of climate change, I guess we can conclude that he probably did not like the results, that the old white wash screen was warmer and he did not want to publish that.

We may be able to make a rough estimate the size of the effect by looking at another experiment with a bad screen. In sunny Italy Giuseppina Lopardo and colleagues compared two old aged, yellowed and cracked screens of unventilated automatic weather stations that should have been replaced long ago with a good new screen. The picture to the right shows the screen after 3 years. They found a difference of 0.25°C after 3 years and 0.32°C after 5 years.

The main caveat is that the information on the whitewash comes from Anthony Watts. It may thus well misinformation that the American Weather Bureau used whitewash in the past. Lacquer paints are probably as old as 8000 years and I see no reason to use whitewash for a small and important weather screen. If anyone has a reliable source about paints used in the past, either inside or outside the USA, I would be very grateful.

Related posts

Changes in screen design leading to temperature trend biases

Temperature bias from the village heat island

Temperature trend biases due to urbanization and siting quality changes

Climatologists have manipulated data to REDUCE global warming

Homogenisation of monthly and annual data from surface stations


Cook, B.I., S.P. Shukla, M.J. Puma, L.S. Nazarenko, 2014: Irrigation as an historical climate forcing. Climate Dynamics, 10.1007/s00382-014-2204-7.

Siebert, Stefan, Jippe Hoogeveen, Petra Döll, Jean-Marc Faurès, Sebastian Feick and Karen Frenken, 2006: The Digital Global Map of Irrigation Areas – Development and Validation of Map Version 4. Conference on International Agricultural Research for Development. Tropentag 2006, University of Bonn, October 11-13, 2006.

Siebert, S., Kummu, M., Porkka, M., Döll, P., Ramankutty, N., and Scanlon, B.R., 2015: A global data set of the extent of irrigated land from 1900 to 2005. Hydrology and Earth System Sciences, 19, pp. 1521-1545, doi: 10.5194/hess-19-1521-2015.

See also: Zhou, D., D. Li, G. Sun, L. Zhang, Y. Liu, and L. Hao (2016), Contrasting effects of urbanization and agriculture on surface temperature in eastern China, J. Geophys. Res. Atmos., 121, doi: 10.1002/2016JD025359.


  1. Thanks for the post. Irritating off-topic but vaguely related question: wiki's pan-evaporation article ( is maybe a bit out of date (see the section). What do you think?

  2. Yes, pan-evaporation is intriguing. You would expect it to increase (like the ocean), but the data shows decreases. I guess it is easy to introduce non-climatic changes due to changes in the instrument, how well people correct for precipitation, and irrigation in the surrounding.

    It seems hard to find non-climatic changes by comparison with neighbouring stations, the network is likely not too dense.

    It is a topic I would love to know more about. If someone is interested in studying instrumental biases in POST, I am interested.

  3. So, I think that at least one contrarian point was not necessarily that we've changed the kind of paint we've used over time, leading to a bias, but rather that every time a station is painted, it will cool down in a step function. The automated algorithms will detect that step function and correct for it. But they won't correct for the slow warming over time... until the next time it is painted. Thereby leading to a "ratchet" effect.

    I don't think we see nearly enough micro-corrections in the automated algorithms for this to be a big thing, but I think that's their argument. Well, one of their arguments. Which conveniently seem to change every time one looks too closely. Maybe it is some kind of Schrodinger wavefunction, where the act of observation causes the argument to change...


  4. MMM, I have not seen this argument with respect to painting, but who knows. The same argument is sometimes used with respect to urbanization. Here the temperature also slowly goes up and then abruptly down when the station is moved to a better location.

    Only correcting the breaks would in this case make the series worse. Mitigation skeptics like to cite James Hansen explaining that that is a bad idea, without realizing that this potential problem has thus been recognized long ago and that that is thus not how climatology removes non-climatic changes. That kind of change point detection and correction is sometimes used in other sciences, for example economics, econometrics, but not in climatology.

    The homogenization methods used in climatology compare the urban station to its neighbors. (In economics you normally do not have a second series of basically the same variable, but only with different measurement errors.) If you would only correct the breaks and not the gradual warming, the urban station would have a much stronger trend than its neighbors and you would thus see in the comparison that it is wrong. In fact, this is not even what is happening, by comparing with a neighboring station you detect and remove any kind of non-climatic change, gradual and abrupt.

    In case of paint degradation and repainting an additional reason is that such small changes over short periods can normally not be detected by homogenization methods. (The variance of the deviation is too small relative to the variance of the noise of the difference between stations.) Thus degradation and repainting only add to the (measurement and weather) noise, but will typically not corrected by homogenization methods. In some exceptional cases, there might be a very nearby neighbor observing almost the same weather (where the difference between the stations would have almost no noise) and you might be able to see such small and short term deviations. I am not sure if that actually happens, but it could be a theoretical rare occurrence.

  5. The term mitigation denier/skeptic is new to me. I like it as a step in the right direction. It moves the message toward a focus on the actual goal of replacing fossil fuels. It leads to better questions.

    One benefit of a mitigation focus is that climate is one of several valid reasons to favor replacing fossil fuels, which greatly widens the pool of support.

    "Do you think it is a good idea to replace fossil fuel use over the next 40 to 75 years, starting now?" is the salient question to ask. The endless arguing over the science is counterproductive, and ultimately irrelevant to the task at hand.

    Now, I understand much of the science at WUWT is drivel. I also know Watts is an early adopter of carbon free technology. Has anyone ever asked his opinion on energy transformation?

  6. There are numerous series in economics that measure the same things but slightly differently, with different strengths and errors.

  7. Paul Kelly, maybe it would be good to do a post on the term "mitigation skeptic". It emphasizes that they are not skeptical when it comes to science, but when it comes to mitigation policies. They accept any "science", no matter how shoddy, if they can make it look to fit their political agenda. And they normally are not against adaptation to protect their own, but against mitigation to also help others. They cannot act as if they are being compared to holocaust deniers. I have not met mitigation skeptic that claimed that he is not skeptical of mitigation.

    There is at least the story of Anthony Watts going to a scientific conference (AGU) and asking as only question a political question on nuclear power.

  8. Eli, that could be, I am no expert for economics. The different series should really measure the same. I have seen people homogenize insolation also using cloud cover as reference because we have only a few insolation measurements. However, then you remove the influence of the optical thickness of the clouds when they are there. That may also change.

    Whatever this case, relative homogenization is quite typical for climate and statistician thus unfortunately often only study absolute homogenization (of single series) and when they do often the case with only one change point, whereas in climate we normally have several in one series. It would be nice if statisticians would work more on multiple breakpoints and relative homogenization.

  9. I'm more interested in those people who, although in some small or great way are skeptical of climate science, nevertheless favor energy transformation.

    I wonder if there are degrees of mitigation skepticism. Let's take Watts and his question about nuclear power at the AGU conference as an indication of where he stands. AKAIK Watts favors nuclear power. Is nuclear power, the one currently available source of emissions free base load power, a mitigating technology? Why yes.

    Solar power is another mitigating technology. Watts apparently uses solar energy in his home and has promoted it his community. Electric cars are an important mitigating technology. Watts reportedly drives an electric car. Can someone who favors and even adopts mitigating technologies be properly called a mitigation skeptic? I am assuming that you know more about Watts than I do.

  10. Paul Kelly, do you know of any statement where Anthony Watts advocates mitigation to solve the climate problem? I am assuming that you know more about Watts than I do.


Comments are welcome, but comments without arguments may be deleted. Please try to remain on topic. (See also moderation page.)

I read every comment before publishing it. Spam comments are useless.

This comment box can be stretched for more space.