AUSTRALIAN BUREAU OF METEOROLOGY – EVIDENCE OF REVISION OF SOLAR EXPOSURE DATA FOR 2013

Bruiser Murray writes:

Several times I have commented that the BOM has falsified Solar Exposure data for 2013.  Recently “Griff” challenged me to substantiate the claim.  I have not had access to my old computer where some of the files were stored so, until recently, could not provide the evidence.  That changed last week and the attached file provides the basis if my assertion.

1. On several occasions, information has been provided to your blog that the Australian Bureau of Meteorology (BOM) has comprehensively revised solar exposure data recorded in 2013. This was done so that the revised data would support the BOM assertion that 2013 was the hottest year on record due to anthropogenic drivers.

Background

2. In Jan 2014, the alarmist web site “Sceptical Science” ran an article: AUSTRALIAS HOTTEST YEAR – HUMANS CAUSED IT. The complete article and comments can be viewed here:

https://www.skepticalscience.com/australias-hottest-year-humans-caused-it.html

Selected excerpts are included below. After reading the article, I (Bruiser) posted the following comments:

bruiser at 20:16 PM on 20 January, 2014

2013 is a poor example to choose as the poster child for AGW.  In fact, quite the opposite. Taking Sydney as an example, new records for solar radiation were set for May, July, August, September, October and December.  The average monthly maximum temperatures track the solar radiation trends. The really hot days seem to correlate closest to low humidity; suggesting that the absence of H2O allows transmission of more solar radiation.   All three climate systems, IOD, ENSO and SAM favour dryer conditions across Australia.

The comment resulted in several individuals responding, with one posting the BOM solar exposure chart for 2013.

barry at 21:37 PM on 20 January, 2014

bruiser @ 13,

Eyeballing the BOM average and anomaly maps, the amount of solar exposure for NSW over 2013 was about average.

1990 – 2011 average

2013 average

Are you able to furnish data links to corroborate what you’re saying?

Lower humidity should in general equal cooler minimums in the dark hours. You might want to check 2013 minimum temps to get more insight. If minima were still anomalously high in low humidity periods, then that’s going to put a hole in your theory.

bruiser at 19:18 PM on 21 January, 2014

@Barry – Hi Barry, Your map of NSW solar radiation in 2013 would appear to be within the “normal” range however the map itself is something of an anomaly if compared to almost any weather station across Australia.  Take the capitals, Alice Springs and fill in the blank spaces at your leisure.  They mostly show 6 months of record exposure. 2-4 months above average and 2-4 slightly below average.  On balance, they are well above average on an annual basis.

barry at 12:48 PM on 22 January, 2014

Composer, I posted links to BOM solar exposure maps @ 14. The authors of the article state,

…we calculated the probability of hot Australian temperatures in model experiments. These incorporated human (changes in greenhouse gases, aerosols and ozone) and natural (solar radiation changes and volcanic) factors. We compared these probabilities to those calculated for a parallel set of experiments that include only natural factors. In this way, natural and human climate influences can be separated.

If Bruiser has lit upon observations that call that study (and BOM results) into question, then Australia-wide data would be an improvement on unverified claims about two locations. I hope he/she obliges us, because I don’t know where he/she is getting the data from.

Reference please, Bruiser?

bruiser at 08:08 AM on 23 January, 2014

@Barry – Hi Barry, 4th try.  The BOM solar radiation data can be accessed at:

http://www.bom.gov.au/climate/data/ Just select solar exposure from the drop-down, enter a town and follow the bouncing ball.  When the data comes up you can select “Plot statistics for this year”.  Have fun.

bruiser at 08:21 AM on 23 January, 2014

@Skywatcher – The data is very similar for any town.  The particularly hot days coincide with low humidity.  No surprise there; it was a very dry year.  You can plot humidity against max, min, temp difference, cloud cover and solar exposure.  My point remains, solar radiation was at record or above average levels for most of 2013.  Particularly hot days had low humidity. Whilst a new national average record was set, no new absolute maximum temperature record was set.  That record is 50.7 degrees C at Oodnadatta airport 2 Jan 19

barry at 00:24 AM on 25 January, 2014

Bruiser, thanks for the link.

I can confirm that all the capitals and about 20 other stations I checked all had record-breaking solar exposure. I think it’s safe to assume I didn’t just chance upon the outliers. So, I checked the colour maps against the values and they are a good fit. The problem, I realize, is that the resolution of the colour maps is too coarse. The difference between average and record-breakers is 1-2 Mj/sq/m, but the colour maps are in increments of 2 Mj/sq/m (annual), and 3 Mj/sq/m (long-term average). So, I withdraw the earlier comparison. I could find no national average data, so can’t crunch the numbers.

bruiser at 21:03 PM on 25 January, 2014

Hi Barry, my check of values against the map found only one where the actual record was within the correct colour band; some were over 2MJ above the displayed range.  2Mj/sq M/day is an enormous amount of additional energy if you accept the calculation that just 0.4 W/sq M is sufficient to cause the decline in arctic ice. (energy required to melt 290 cubic Km of ice).

Diurnal variation has interested me for a few years.  For example, if you plot the data for Melbourne, you get a steady decline. However, the Melbourne weather station is situated at a busy city intersection.  I plotted the data for Laverton on the same graph and there is a great correlation until about 1964 when there is a steady roll-off in the Melbourne values.  No such decline occurs for Laverton – suggesting that the Melbourne site was affected by some form of development.

barry at 00:06 AM on 26 January, 2014

Bruiser, I rechecked the 2013 solar exposure map against annual values for various locations again and you are quite right. The colour bands do not match the local values. I will email BOM about it.

2Mj/sq M/day is an enormous amount of additional energy if you accept the calculation that just 0.4 W/sq M is sufficient to cause the decline in arctic ice. (energy required to melt 290 cubic Km of ice).

How then would you explain that 4 out of 7 states had warmest years when before 2013, when solar exposure was not highest (in the capitals)?

Following are the warmest years by state, and solar exposure data value (Mj/sq M/day) for that year, ranked from highest, per single locational in the capitals, then record solar exposure (2013) in bold. Period is 1990 – 2013 incl. which is all the solar exposure data we have.

barry at 00:06 AM on 26 January, 2014

Bruiser, I rechecked the 2013 solar exposure map against annual values for various locations again and you are quite right. The colour bands do not match the local values. I will email BOM about it.

SUPPORTING EVIDENCE

3. I have been tracking solar exposure data on the BOM web site throughout 2013. It appears from about August onwards that the BOM was tracking for record levels across the country. The records only go back to 1990 (not a large sample). Table 1. displays the monthly totals for 2013 and the mean for all capital cities, Alice Springs, Derby and Cairns.

Table 1. Solar exposure MJ/M2/Day

4. The BOM chart graphs for Sydney provide evidence of the BOM revisions. The first graph was downloaded in Jan 14 whilst the second is the same graph downloaded in May 14:

Product Code: IDCJAC0016

Jan 2014

Product Code: IDCJAC0016

Page created: Thu 08 May 2014 14:01:04 PM EST

5. The difference in the two charts provides insight into the revisions. The first chart clearly shows record levels of solar exposure for May, July, August, September, October and December. The second chart does not have any record months. The extent and degree of the data revisions in the Sydney chart, the data for all the cities is displayed in Table 1. Table 2 has been conditionally formatted to show where the BOM has increased the levels of solar exposure (coded pink) and decreased solar exposure (coded blue). The conditional formatting clearly highlights that solar exposure data was increased slightly for January, February and March and decreased dramatically for the rest of the year.

Table 2:

6. In addition, a comparison of the Sydney complete data set indicates that the revision of data by the BOM extends back to 1 July 2011. Clearly, the BOM altered the data some time between Jan 14 when Barry undertook to e-mail them about the discrepancy and May 2014 when I revisited the issue.

7. For your considered response.