Tuesday, November 08, 2011

Analysis of Australian Temperature - Part 7 - Adjusting temperature for cloud cover

Previously I analysed the relationship between cloud cover and temperature in Australia at certain times of the day. Not surprisingly the results found that increases in cloud cover were related to lower temperatures during the day and higher temperatures at night. Similarly, lower levels of cloud cover resulted in higher day time temperatures and lower night time temperatures.

I have also shown previously that the cloud cover levels have decreased substantially since 1950.

So what happens if we eradicate the variable cloud cover from the database? In other words, if we keep the cloud cover level and adjust the temperatures, we should be able to see what happens to Australian temperatures at certain times of the day irrespective of any increase or decrease in cloud cover levels.

The following are the standard 8 graphs of temperature anomalies with the trend shown by the light blue line, and also the cloud adjusted temperature anomalies as represented by the regressed red line.

















The graphs above might be confusing to some but let me explain.

Basically it shows that when we adjust for cloud cover during that particular year, the temperatures increases at Midnight, 3am and 6am have little difference. The slope of those lines is very similar. However at 9am a massive change occurs. When adjusting for clouds, instead of an increased temperature at a rate of 0.92 degrees per 100 years, it only increases by 0.68 degrees per 100 years, or a 26% less.

This increases even more so as you get to the 3pm time slot. Where previously we recorded a 1.09 degree per 100 years increase, when we account for cloud cover during that particular year, we only see a 0.44 degree increase per 100 years, or a 60% decrease in temperature. Interestingly, this is generally part of the day where the maximum temperature often occurs.

The trend then decreases, and then by 9pm, there is no difference between normal temperature anomalies and cloud adjusted temperature anomalies.

Quite clearly, the only times where clouds make a difference in the long term Australian temperature averages is during the day. The more clouds there are, the less the sun can get through and the cooler it is. Similarly for the opposite.

The graph below shows the difference between the average of the maximum and minimum temperatures in Australia and the average of the cloud adjusted temperature anomalies.



What the above graph shows, is that by using the normal method of calculating the average long term temperature, we have an increase of 1.17 degrees per 100 years, whilst the average of the cloud adjusted temperatures record an average of just 0.54 degrees per 100 years, or 54% less.

Perhaps it is true that we are overestimating how much Australia is warming up by more than double. But there is more to come on this, and in the next post I will look at some more long term trends in cloud cover and what effects they have on temperature, which will have very interesting results indeed.

11 comments:

Phil said...

Jonathan,

Do you know how the BoM measures cloud cover?

The BoM website suggests it is still done by the traditional method of an observer estimating the 8ths cloud cover.

Philip

Jonathan Lowe said...

yep that is correct Philip, it is recorded in 8ths

Jonathan Lowe said...

from pure observation

Phil said...

The reason I ask, is I am looking for a measure that distinquishing changes in solar isolation from clouds and aerosols.

So far without success.

Phil said...

Jonathan,

This is of interest.

http://wattsupwiththat.com/2011/11/14/linked-aerosol-pollutants-and-rainfall-patterns/#comment-797059

Jonathan Lowe said...

yeah that's really interesting Phil. Not sure if you are pointing to a specific comment or just the article. Either way, this could well be something, that surprisingly, has not largely been studied

Phil said...

Jonathan,

Did you get the email I sent you about the international 6 hourly temperature dataset?

Anonymous said...

So by assuming that the cloud cover has not changed, we can deny that the temperature is increasing as fast as the temperature data shows?

Isn't this 'Hide the incline'

Mick

Dominios Colombia said...

It looks really good! Very creative post about weather measurement.

Technical translations said...

The measurement is made as a combination of human and laser measurements if I am correct.

Doug Proctor said...

Tallbloke's referenced me here when I asked this question.

I was thinking that the lack of WIND was a factor: air movement from the oceans across the continent, as a co-factor to cloud cover.

On Tallblokes two years ago I did a study on Central England Max Temperatures vs Bright Sunshine 1930 to 2010. I felt I correlated very well to a portion of the temperature increase with sunshine hours. If there are bright sunshine hours for Australia, a similar comparison could be done here. The remnant warming had a very similar pattern to the warming attributable to AMO-PDO. I felt that there was very little left to be either CO2 or land use.

But I didn't consider wind, but England is so narrow it wouldn't matter. For Australia, I'd consider the wind. It is related that on very warm days windpower is close to nil. So wind could be a co-factor.