Tuesday, October 09, 2007

Differences between minimum and overnight temperatures

Previously we discovered that minimm temperatures in Australia have been on the increase since world war 2 at a rate of 0.522 degrees Celsius in the last 64 years.

However we also found that temperatures at Midnight, 3am and 6am have not seen any significant increases or decreases in temperature over the same time period. This is quite strange, as it has long been thought that the minimum temperature is a good representation of overnight temperatures. Obviously with different trends occurring in each case, we suggest that it is not. However, because no scientific literature has even bothered to look at time based temperature, this type of discovery goes unnoticed in the science world, and we still see dozens of studies looking at the consequences of night time warming.

So why is this occurring? Well lets have a look at the differences between the minimum temperature and temperature anomalies at 3am and 6am, which is shown in the graph below.

The graph above shows an obvious increasing, and statistically significant trend. Minimum temperature anomalies from the second world war up until now have been increasing at a greater rate than temperatures at 3am and 6am. If the minimum temperature were to be a good representation of overnight temperatures then we would expect no trends in the data, and the average "bar" to be around zero.

I've included two parabolic trend lines to highlight the differences between the two variables. It would seem that 6am anomalies closer resemble minimum temperature anomalies. This makes sense, as the minimum temperature generally occurs closer to 6am than 3am. In fact, in about the last 10 years, there has been no major trend between 6am and minimum temperature anomalies.

It is no surprise then that we have seen in the last 10 years, and indeed since 1973, no significant increase in minimum temperature.

Hence in more recent times, the minimum has been a reasonable variable to measure overnight temperature (despite being about 0.05 degrees Celsius above temperatures at 3am and 6am), but was a poor measure of overnight temperature before 1975.

Even the ABOM agrees with us here (aside from other irregularities), in that minimum temperatures have not seen any increase since 1973.

As shown on Andrew Bolts forum, Bob Forster, suggested the following:

If you ignore the distraction of the bars for individual years, and just look at the 5-year running mean, you will discover that most of the warming was in a single step-change in the latter half of the 1970s.

That jump in Australian average temperature correlates with the Great Pacific Climate Shift of 1976/1977. The Shift marks a change from Pacific Decadal Oscillation cool phase to warm phase – thus reversing a cool shift in the early 1940s. This 76/7 climatic step-change correlates with an abrupt reduction in the upwelling quantity of cold deep water in the equatorial eastern Pacific. Put another way – before the shift there was a preponderance of La Niña conditions, and after it El Niño dominated. Change in upwelling quantity on this scale (it varies a lot - but say, from about 26 Sverdrups before the Shift to about 18 Sv after) is an inertial event of huge magnitude. I don’t see how human-caused CO­2 emissions could have done that.

An interesting theory, one that I believe will be heavily scrutinized no doubt. But whatever the reason we can conclude that the weather and temperature has been changing since the second world war, and it has not been just a gradual increase in temperature as global warming alarmists cause.

Temperatures at night have not been increasing, and the minimum temperature has not increased in the last 30 years, a period where is more closely resembles overnight temperatures.

And what of differences in temperature anomalies between other times? We'll look at them next.


Philip said...

It's difficult to see how PDO changes impact incoming and outgoing radiation rates. I still think clouds are the explanation. I recently saw a study on decreasing cirrus cloud formation in recent years, which appears to explain increased in/outgoing radiation. I'll dig out the link.

BTW, nice graphic. I haven't seen that kind of trend analysis before.

Ralph said...

Forgive me for asking a daft question, but how can the min - 3am anomaly ever be positive? Surely the daily minimum temperature is by definition no higher than any other temperature recorded on that day?

-- Ralph

Jonathan Lowe said...

its actually the minimum anomaly minus the 3am anomaly. not the minimum temperature minus the 3am temperature.