Wednesday, July 18, 2007

Colder at night, warmer during the day in South NSW/North Vic

Putting a couple of close weather stations together we have Rutherglen and Wagga Wagga.

Unfortunately no midnight temperatures to speak of, but all other temperature records go back to 1945 with max and min since near the start of the 20th century. To the analysis:

Maximum temperatures show no significant increase, however the past 6 or so years (like most places in Australia) have seemed to be above normal. Minimum temperatures however were a stones throw from significantly decreasing (p = 0.066). An obvious downtrend is occurring, however it must be noted that, although close, the trend is insignificant.

Temperatures at 3am show no significant increase or decrease (p = 0.85). Interestingly temperatures at 6am and 9am show insignificant increases or decreases, however they to are very close to a significant decrease (6am: p = 0.052, 9am: p = 0.075). The 9am graph is less obvious of a trend than at 6am as it shows colder than normal temperatures during the 80s and early 90s, and warmer than average temperatures this decade.

Despite this temperatures at Noon and 3pm (during the heat of the day), showed significant increases in temperature (Noon: p = 0.04, 3pm: p = 0.03). Of obvious note is the last 5-6 years of higher than normal temperatures at this time, especially in 2006, which reached almost 2 degrees hotter per day at 3pm.

When the sun loses it's impact, we don't see the temperature raise, with 6pm and 9pm providing insignificant trends (6pm: p = 0.1, 9pm: p = 0.8)

When analysing temperatures at 3am and 6am compared to the minimum, we didn't see the normal results of an increasing trend of minimum temperatures as compared to overnight temperatures. Hence in this case, the minimum is a good indication of overnight temperatures (which were found not to significantly increase or decrease).
I am not quite sure of the reason for this, as it has not occurred in other places in Australia. It could well be random variation or a certain attribute for the area.

With analysis into neighboring times, temperature anomalies at 6am were significantly decreasing to temperature anomalies at 3am. The sudden decrease at around 1973 should raise many peoples eyes in regards to the reliability of the data taken from here. Noon temperatures anomalies were significantly increasing at a greater rate than at 9am, and 9pm temperature anomalies were decreasing significantly as compared to 6pm.

Whilst the reasons for the trends is not so clear cut as previous. This graph, which plots the t value (positive t values indicate increasing trend, negative indicate decreasing trend), highlights the situation in the area.

As one can see, overnight this area has had a decreasing temperature trend, but during the day, when the sun plays an influence, the temperature of the years has shown an increasing trend.

More evidence of the suns effect on temperature in Australia.


Peter Ruthven said...

Have you considered that maximum minimum temperatures are related to variations in precipitation?
I have observed that during dry weather cycles, such as the last 10 years, maximum temps tend to be higher and minimum temps tend lower than average.
Whilst during wet weather cycles the reverse is true. This is probably related to variations in cloud cover although soil moisture is also a factor.
It is interesting that 1914, the hottest year ever recorded in Victoria is also the dryest year.

philip said...

Jonathan, what's the y axis on that graph?

Jonathan Lowe said...

Peter it most likely has a correlationn with rain, however perhaps only because that has a correlation with cloud, and that has a correlation with no sun. I have been trying to get data for each area on cloud for Solar Radiation, Sunshine Durartion and Total Cloud cover - but on last talking to the BOM they said they were not giving that out - despite the fact that they have it well advertised on their website.

Philip, as mentioned above the graph: the yvalue on the graph is: the t value of the relationship between temperature at time(positive t values indicate increasing temperature over time, negative indicate decreasing temperature over time)

philip said...

This graph, which plots the t value (positive t values indicate increasing trend, negative indicate decreasing trend)

I interpreted this to mean there was an increasing trend (warming) during the day and a decreasing trend (cooling) during the night.

Which would be very much contrary to your earlier findings. Although as predicted by GHG theory.

I now understand the graph to mean increased daytime warming and almost identical increased nighttime cooling, consistent with your previous findings and directly contrary to GHG theory.

Jonathan Lowe said...

yep previously I have found no increase or decrease in temperature over time at night, but only an increasing trend during the day, which increases the closer you get to the maximum, which goes against the green house gas theory and for a solar theory,.

peter ruthven said...

This is my point, that the maximum temp anomalies seen over the past few years, and the slight decrease in overnight temps during the same period, are related to the decreased cloud cover associated with the below average rainfall conditions of this period.
This is about to change and we should see a reversal of conditions over the next few years. Nothing to do with GW and everything to do with the relationship between solar activity and the amount of low level cloud cover.
The Southern Victoria weather pattern does show a variable cycle lasting for approx 17 +-4 years, ie approximating the solar cycle, there is also a 50 year cycle.
It's interesting how temperatures in Victoria (and Antartica) fail to agree with GW theory. Could it be because this part of the world, swept by the Great Southern Ocean is relatively uncontaminated by the heat from human activity? I do believe so.
Arthropogenic CO2 in my view is probably no more important as a source of warming than is variation in solar activity, both affect our climate.