Tennant Creek NT [BoM ID 015135 (airport), 015087 (post office); joined 1 November 1970] Temperature data from February 1910 to December 2022.

Is homogenisation of Australian temperature data any good?

Dr Bill Johnston, former research scientist, NSW Department of Natural Resources

www.bomwatch.com.au

Protocols, which are essentially a description of the research question (hypothesis) and the means by which it would be addressed, lie at the heart of the scientific method. This report outlines the basis for, and steps involved in undertaking unbiased analysis of trend and change in Australian maximum temperature (Tmax) datasets using Tennant Creek as the case study.

Based on the First Law of Thermodynamics, BomWatch protocols are transparent, objective and replicable and provide a firm baseis for assessing trend and change in maximum temperature datasets and for resolving issues related to the Bureau of Meteorology’s homogenisation of the same data.

The Tennant Creek Telegraph station in 1872, showing a wind vane and Stevenson screen on the far right of the photograph (National Archives of Australia Item ID: 11774998). As the site was close to buildings, it was probably watered by bucket or watering-can during the seasonal ‘dry’.

Summary

The case study concluded that step-changes related to site changes caused Tmax to increase 1.51oC between 1910 and 2022 independently of rainfall, and that site-changes and rainfall simultaneously explained 73.8% of Tmax variation. Post hoc analysis found no residual trend or change was attributable to any other factor including CO2, coalmining, electricity generation or anything else.

Homogenisation cooled pre-1935 and pre-1963 Tmax to varying extents and achieved trends approaching 0.15oC/decade. Australian Climate Observations Reference Network – Surface Air Temperature V.2.4 also adjusted data higher from 1999 to 2015 by 0.52oC thereby smoothing the effect of post-2013 site changes so they appeared to be due to the climate. However, while trend became more significant and goodness-of-fit improved, variation in Tmax explained by rainfall declined from 46.3% initially, to 16.5% for ACORN-SATv.2.4.

Maps, plans, aerial photographs and satellite images show unequivocally that trend in Tmax data was mostly due to spraying out the grass and replacing the former 230-litre Stevenson screen in 2012, and installing a wind-profiler array within 45m of the screen before March 2013.

Bomwatch protocols

BomWatch protocols comprise four elements, namely:

· The overall relationship between Tmax and rainfall partitions total variation into that due to rainfall, and the residual non-rainfall part. Linear regression also derives the overall Tmax/rainfall coefficient, and significance (P) and goodness of fit (R2adj) statistics that indicate conformity with the First Law of Thermodynamics.

· Homogeneity analysis of rescaled residuals identifies non-climate impacts on data, which are categorised as step-change or (Sh)ift scenarios

· Segment-by-segment analysis with rainfall detects outliers, lack of fit, and other potential problems, and,

· Categorical multiple linear regression (and interaction analysis) finalises and verifies outcomes.

Segmented trend and graphical analysis confirm and verify that relationships are linear, residuals are normally distributed, independent, with constant variance, and that they are timewise homogeneous.

Based on the First Law Theorem that maximum temperature depends on rainfall, BomWatch protocols provide an unequivocable basis for understanding the effect of non-climate impacts on data, and for objectively assessing the BoM’s homogenisation methods.

Click here to download the full paper with photos, graphs and data

Bill Johnston

6 February 2024

Is homogenisation of Australian temperature data any good?

Is homogenisation of Australian temperature data any good?

Part 7b. Timber Creek, Northern Territory, Australia

Bureau of Meteorology ID 14850, Latitude -15.6614 Longitude 130.4808.

Dr Bill Johnston

scientist@bomwatch.com.au

Summary

Data quality is poor. The use of faulty data that are not homogeneous, to adjust faults in ACORN-SAT data is unscientific and likely to result in biased outcomes. As the ACORN-SAT project does not ensure comparator datasets are homogeneous, it is deeply flawed and should be abandoned. Read on …

The quality of Tmax data for Timber Creek is exceptionally poor. Its use for adjusting faults in ACORN-SAT data is highly questionable. Data are affected by missing daily data, poor site-control and the probable replacement of a 230-litre Stevenson screen with a 60-litre one around 1996. Although data lack precision, accounting for rainfall and the up-step in 1996 left no residual variation attributable to CO2, coalmining, electricity generation or anything else.

Background

The isolated town of Timber Creek is located beside the Victoria River, on the Victoria Highway, 600km south of Darwin and 280m southwest of Katherine in the Northern Territory (Figure 1). The Victoria River flows west to the Joseph Bonaparte Gulf close to the Western Australia – Northern Territory border. Daily maximum temperature (Tmax) data is available from the Bureau of Meteorology (BoM) from 23 January 1961, to when the station closed on 5 September 2014.

Figure 1. Timber Creek and other weather stations having more than 10-years of records (grey points), and Victoria River Downs (blue) and other ACORN-SAT sites in northern Australia (red buttons).

Tmax data for Timber Creek was used to adjust ACORN-SAT sites at Victoria River Downs, Kalumburu, and Rabbit Flat (Table 1). So, is Tmax data for Timber Creek any good?

Table 1. ACORN-SAT sites adjusted at the dates indicated using Tmax data for Timber Creek.

ACORN-SAT site: (Distance to Timber Ck.)	Vic. River Downs (ID 14825) (106km)	Kalumburu (ID 1019) (434km)	Rabbit Flat (ID 15666) (509km)
Adjust dates AcV1 (2009)	1 Aug 1987	1 Jan 1986 23 Aug 1991	Nil
Adjust dates AcV2.1 (2020)	Nil	1 Jan 1986 23 Aug 1991	1 March 1997
Adjust Dates AcV2.3 (2023)	Nil	1 Jan 1986 23 Aug 1991	1 March 1997

Methods

Daily temperature and monthly rainfall were downloaded from the Bureau of Meteorology, Climate Data Online facility (Climate Data Online – Map search (bom.gov.au)). Monthly rainfall was infilled as necessary using data for the nearest available site and flagged for reference. Data were summarised into an annual dataset (Timber Creek.xlsx) and analysed using the same BomWatch protocols described previously in the Parafield case study1 and subsequent reports, including for Victoria River Downs2 of which this report is a subset.

Only maximum temperature (Tmax) data were analysed.

Results

Missing daily observations may have affected mean annual Tmax in 1987, 2008 and 2013 (N <346 observations/year); however, influence plots did not suggest those years were outliers. Outlier years which were identified (but also not omitted) were 1998, 2005 and 2013. Observations appeared to become lackadaisical after 2006. No other attributes of the data were remarkable.

Rainfall, which is the deterministic portion of the Tmax signal explained 27.0% of Tmax variation (Figure 2(a)), which is low relative to the BomWatch benchmark of R2adj = 0.50. Either data quality is exceedingly poor, or an influential variable has not been accounted for by the naïve Tmax ~ rainfall case (Table 1(i)).

(Note: R2adj calculated by the statistical package R, adjusts variation explained for the number of terms in the linear model, as well as for the number of observations3. It is therefore more robust (less biased) than unadjusted R2 calculated by spreadsheet programs such as Excel.)

Figure 2. Composite analysis of Timber Creek Tmax.

Inhomogeneities in rescaled Tmax ~ rainfall residuals were evaluated using STARS, which objectively tests whether the mean of subsequent values is significantly different (P<0.05) to that before, using a running t-test of the difference.

Indicated by the horizontal line in Figure 2(b), STARS detected an up-step of 0.47oC in 1996 (P = 0.003). Segments defined by the step-change were examined separately in Figure 2(c) and Figure 2(d) and Table 1(ii).

Categorical multiple linear regression (Table 1(iii)) showed rainfall reduces Tmax 0.175oC/100mm, and that segmented regressions were offset by a rainfall-adjusted difference of 0.58oC (Figure 2(e)), which is within range of that detected by STARS.

Post hoc tests (Table 1(iv)), confirmed that data consisted of two non-trending segments interrupted by the step-change in 1996.

Although data lack precision (R2adj <0.50, Table 1(iii)), accounting for rainfall and the step change in 1996, left no residual variation attributable to CO2, coalmining, electricity generation or anything else.

Table 1. Statistical summary. RSS refers to residual sum of squares. Partial R-square (R2partial) estimates the proportion of variation explained by the Sh(ift)factor that is not explained by rainfall alone (calculated as: [(RSSfull – RSSrain)/ RSSfull)*100].

Model	Coef. (oC/100mm)	P	R2adj	Segment	RainAdj (SE) (oC)(1)	RSS (R2partial)
(i) Tmax ~ rain (all)	-0.151	0.001	0.270			7.64
(ii) Tmax ~ rain(2) 1961-1995 1996-2013	-0.192 -0.240	0.005 0.001	0.420 0.464
(iii) Tmax ~ Shres + rain(2)	-0.175	<0.001	0.473	1961-1995 1996-2013 Delta(1 vs 2)	34.6(a) (0.120) 35.2(b) (0.109) 0.58 (0.173)	5.56 (37.5%)	Interaction Tmax ~ Shres * rain ns
(iv) Tmax ~ Year(2) 1961-2013 1961-1995 1996-2013	(oC/decade) 0.106 0.617 -0.196	0.325 0.329 0.528	ns ns ns
(1) Letters in parenthesis indicate differences between means (2) No outliers

Discussion

Relationships between Tmax and rainfall (Table 1) show shows the quality of Tmax data for Timber Creek is exceptionally poor, which raises the question why would it be used to adjust ACORN-SAT sites at Victoria River Downs, Kalumburu, and Rabbit Flat? Furthermore, the step-change in 1996 is likely due to the former 230-litre Stevenson screen being replaced by a 60-litre one during the previous year.

Despite problems with watering and missing observations, Victoria River Downs is one of only several closed or open sites within 200km with reasonable runs of Tmax data (Kununurra town and airport and Kimberley Research Station being others). Although Timber Creek closed in September 2014, it seems from available Google Earth Pro satellite images the site was watered (Figure 3).

Figure 3. The satellite image of the alleged Timber Creek weather station site in August 2022 at Latitude -15.6614o, Longitude 130.4808o. However, the site looks nothing like the site diagram shown in site-summary metadata. There is no sign of the ‘chook shed’, while the ‘house’, appears to be the Timber Creek Travellers Rest Motel and caravan park. Successive images also show the grass surrounding the site is regularly watered. Either the coordinates provided in metadata are incorrect or the diagram is inaccurate.

The problem is that site-summary metadata is faulty.

Close inspection of properties around the town, located the Stevenson screen inside a fenced area that matched the site diagram, at Latitude -15.6600o, Longitude 130.4819o, approximately 190m NE of the location shown in Figure 3.

While regular manual observations may have ceased, the most recent satellite image (August 2023) shows the Stevenson screen is still located within the enclosure. Images from August 2002, show land-use changes, possibly irrigation and the use of herbicide between seasons and years contributed variation that impacted observations, and therefore that lack of site control was the main factor affecting the quality of Tmax data.

Given the paucity of nearby sites, and the lack of useful long-term data, Timber Creek is not such a big overcrowded place that the BoM could not have installed an automatic weather station on vacant land somewhere in the town where site conditions were consistent with the local climate – high monsoonal rainfall from November to March (summer) and low rainfall during the ‘dry’. Unreliable metadata, including routinely failing to document when Stevenson screens are replaced, and lack of site control is a major problem across Australia’s weather station network.

Conclusion

It is concluded that the quality of Tmax data for Timber Creek is exceptionally poor and its use for adjusting faults in ACORN-SAT data at Victoria River Downs, Kalumburu and Rabbit Flat is highly questionable. Data are affected by missing daily data, poor site-control and the probable replacement of a 230-litre Stevenson screen with a 60-litre one around 1996. Although data lack precision, accounting for rainfall and the step change in 1996 left no residual variation attributable to CO2, coalmining, electricity generation or anything else.

Bill Johnston

18 February 2024

Preferred citation:

Johnston, Bill 2024. Is homogenisation of Australian temperature data any good? Part 7b. Timber Creek, Northern Territory, Australia http://www.bomwatch.com.au/ 4 pp.

Disclaimer

Unethical scientific practices including the homogenisation of data to support political narratives undermines trust in science. While the persons mentioned or critiqued may be upstanding citizens, which is not in question, the problem lies with their approach to data, use of poor data or their portrayal of data in their cited and referenceable publications as representing facts that are unsubstantiated, statistically questionable or not true. The debate is therefore a scientific one, not a personal one.

Acknowledgements

David Mason-Jones is gratefully acknowledged for providing invaluable editorial assistance. Research

Is homogenisation of Australian temperature data any good?

Part 7a. Larrimah, Northern Territory, Australia

Bureau of Meteorology ID 014612. Located 430 km SE of Darwin at Latitude -15.5748 Longitude 133.2137. Maximum temperature data from 1 January 1965 to 29 June 2012.

Dr Bill Johnston[1]

scientist@bomwatch.com.au

Data quality is poor. The use of faulty data that are not homogeneous, to adjust faults in ACORN-SAT data is unscientific and likely to result in biased outcomes. As the ACORN-SAT project does not ensure comparator datasets are homogeneous, it is deeply flawed and should be abandoned. Read on …   

Summary

While data quality is poor, accounting for rainfall and the step-change simultaneously using multiple linear regression left no additional trend or change that could be attributed to the climate, CO2, coalmining or anything else.

1.    Background

Situated on the Stuart Highway 430 km southeast of Darwin (Figure 1) maximum temperature (Tmax) observations at Larrimah ceased on 29 June 2012. However, Larrimah Tmax was used by ACORN-SAT (the Australian Climate Observations Reference Network – Surface Air Temperature project) to homogenise Tmax data at Victoria River Downs (1/1/1976 and 1/8/1987, AcV1 and 1/1/2007, AcV2.x) and Burketown airport (1/1/2002, AcV1 and 1/1/1986, AcV2.x).

Figure 1. Larimah Tmax was used to homogenise Tmax data for Victoria River Downs (256km away), and Burketown airport (762 km distant). ACORN-SAT sites are indicated by red buttons, and sites having more than 10-years of data, by grey circles.

This raises the question: is Tmax data for Larrimah any good?

2.    Methods

Daily temperature and monthly rainfall were downloaded from the Bureau of Meteorology, Climate Data Online facility (Climate Data Online – Map search (bom.gov.au)). Data were summarised into an annual dataset and analysed using the same BomWatch protocols described previously in the Parafield case study[2] and subsequent reports, including for Victoria River Downs[3] of which this report is a subset.

Only maximum temperature (Tmax) data are used in the study.

3.    Results

Ignoring outliers (red squares in Figure 2(a)), rainfall, which is the deterministic portion of the Tmax signal explains 39.5% of Tmax variation (R_adj = 0.395). As this is less than the R²_adj BomWatch benchmark of 0.50, either data quality is exceedingly poor, or an influential variable has not been accounted for by the naïve Tmax ~ rainfall case (Table 1(i)).

(Note: R²_adj calculated by the statistical package R, adjusts variation explained for the number of terms in the linear model, as well as for the number of observations[4]. It is therefore more robust (less biased) than unadjusted R² calculated by spreadsheet programs such as Excel.)  

Rescaled so values are comparable, Tmax ~ rainfall residuals were evaluated for inhomogeneities using STARS, which objectively tests whether the mean of subsequent data is significantly different (P<0.05) to that before, using a t-test of the difference. Indicated by the horizontal line in Figure 2(b), STARS detected an up-step of 0.41^oC in 1986 (P = 0.015). Segments defined by the step-change were examined separately in Figure 2(c) and Figure 2(d).

The spread of points about each line, and influence plots (not shown) confirmed that data for 1982 and 2011 were likely outliers. Also, as R²_adj was <0.50 data quality was generally poor.

Categorical multiple linear regression of the form Tmax ~ Sh(ift)_factor + Rainfall showed rainfall reduces Tmax 0.174^oC/100mm, and that segmented regressions were offset by a rainfall adjusted difference of 0.41^oC (Figure 2(e)), which is the same as that detected by STARS. Post hoc tests confirmed that data consisted of two non-trending segments interrupted by the step-change in 1986 (Table 1(iv)).

Figure 2. Composite analysis of Larrimah Tmax. A statistical summary for each phase of the analysis is provide in Table 1.

4.     Discussion and conclusions

Diagrams in site-summary metadata for 1 September 1964 and 7 December 1968, show the site was originally at the rear of the post office on the western side of the Stuart Highway. However, by 8 June 1992 it had moved to the disused WWII rail terminus of the North Australian Railway, which ceased operations and closed in February 1981.

While the alleged move within the confines of the terminus in November 1998 was not influential, the 1986 step-change was probably due to relocating the site there and replacing the original 230-litre Stevenson screen with a 60-litre one. It seems the move to the trucking-yard, allegedly in 1998, was erroneously reported. 

Google Earth Pro satellite images from October 2004 show surrounds of the trucking-yard site were generally bereft of ground cover, which would be sufficient to cause Tmax data to be warmer than at the previous site behind the post office.

Although categorical multiple liner regression confirmed that the up-step in the mean occurred in 1986 (Table 1(iii)), low R²_adj and apparent overlap in scatter between the series (Figure 2(e)) show data were of exceptionally poor quality, particularly after 1986. This may have been due to lackadaisical observing practices, including excessive numbers of missing data/year. (Fewer than 340 observations were noted in 1970, 1975 and 1976, and from 1994 to 1998). Although given its role in ACORN-SAT and that the site is isolated, it is surprising that the BoM did not install an automatic weather station at Larrimah. The current site probably closed due to a lack of local interest in making observations and undertaking maintenance.  

It was concluded that while data quality is poor, accounting for rainfall and the step-change simultaneously using categorical multiple linear regression, left no additional trend or change that could be attributed to the climate, CO2, coalmining or anything else.

Table 1. Statistical summary. RSS refers to residual sum of squares. Partial R-square (R²_partial) estimates the proportion of variation explained by the Sh(ift)_factor that is not explained by rainfall alone (calculated as: [(RSS­_full – RSS_rain)/ RSS_full)*100].

Model	Coef. (oC/100mm)	P	R2adj	Segment	RainAdj (SE) (oC)(1)	RSS (R2partial)
(i) Tmax ~ rain (all)      Not 1982, 2011(2)	-0.176 -0.171	<0.001 <0.001	0.379 0.395			10.649
(ii) Tmax ~ rain(2) 1965-1985 1986-2011	-0.175 -0.175	<0.001 0.004	0.482 0.400
(iii) Tmax ~ Shres + rain(2)	-0.175	<0.001	0.473	1965-1985 1986-2011 Delta(1 vs 2)	33.7(a) (0.113) 34.2(b) (0.101) 0.41 (0.151)	12.507 (14.9%)	Interaction Tmax ~ Shres * rain ns
(iv) Tmax ~ Year(2) 1965-2011 1965-1985 1986-2011	(oC/decade) 0.028 0.240 -0.426(3)	0.720 0.329 0.025	ns ns 0.165
(1) Letters in parenthesis indicate differences between means (2) Outliers omitted from the analysis (3) Affected by a single extreme value

Bill Johnston

17 February 2024

Click here for the full Larrimah data

Preferred citation:

Johnston, Bill 2024. Is homogenisation of Australian temperature data any good? Part 7a. Larrimah, Northern Territory, Australia http://www.bomwatch.com.au/ 3 pp.

Disclaimer

Unethical scientific practices including the homogenisation of data to support political narratives undermines trust in science. While the persons mentioned or critiqued may be upstanding citizens, which is not in question, the problem lies with their approach to data, use of poor data or their portrayal of data in their cited and referenceable publications as representing facts that are unsubstantiated, statistically questionable or not true. The debate is therefore a scientific one, not a personal one.

Acknowledgements

David Mason-Jones is gratefully acknowledged for providing invaluable editorial assistance. Research includes intellectual property that is copyright (©).

---

[1] Former NSW Department of Natural Resources research scientist and weather observer.

[2] Parafield Ref Welcome to homogenisation – arguably the greatest scam of 20th century science (bomwatch.com.au)

[3] Victoria River Downs Ref VictoriaRiverDowns-16-Feb-2024-1.pdf (bomwatch.com.au)

Is homogenization of Australian weather data any good?

Dr Bill Johnston[1]

scientist@bomwatch.com.au

Situated 443 km south of Darwin, 410 km NE of Halls Creek and 433 km north of Rabbit Flat in the Kimberley Region of the Northern Territory, the iconic Victoria River Downs Station was once the largest pastoral holding in the world. The station homestead beside the Ord River (Figure 1), is the location of an Australian Climate Observations Reference Network – Surface Air Temperature dataset (ACORN-SAT) weather station, one of 112 such sites used to monitor climate warming in Australia. Due to the sparseness of the Bureau of Meteorology’s (BoM) network in northern Australia, data for Victoria River Downs (BoM ID 14825) is weighted by ACORN-SAT to be representative of some 3.3% of Australia’s land area.

Figure 1. The Victoria River Downs homestead and outbuildings south of the Ord River in the Kimberley Region of the Northern Territory in 1954 (National Library of Australia, copyright expired: https://nla.gov.au:443/tarkine/nla.obj-137684286).  

From when observations commenced in 1965 until about 1979 data were plagued by runs of missing observations, or low data counts per month. Observations were also mostly reported in whole and ½^oC from 1975 to 1982. The frequency of whole degrees was higher than other decimal-fractions from when the automatic weather station (AWS) was installed in May 1997 until about 2006. Site-surrounds and the lawn beyond were also irregularly watered. Maximum temperature data (Tmax) could not therefore be judged as high-quality. As data were unavailable from June to September 1973 (N=210), 1973 was omitted from the analysis.

The overall Tmax trend of 0.125^oC/decade (P = 0.055) was spuriously due to an abrupt Tmax up-step of 1.12^oC in 2013, which was not related to a change in the weather or climate. Accounting for that and the effect of rainfall on observations, left no additional signals that could be attributed to CO₂, coalmining, electricity generation or anything else.

The change may have been due to cessation of watering (which was said to have ceased in 2007), replacement of a former 230-litre Stevenson screen with a 60-litre one, or replacement of a wooden 60-litre screen with a plastic one. According to site-summary metadata the Tmax thermometer was removed in July 2012 (thus backup manual observations ceased), but curiously, it was re-installed in March 2016 and replaced again in September 2019. Like many ACORN-SAT sites, metadata (data about the data) is vague and unreliable and not a basis for correcting data for site-change effects.

Consistent with the First Law of Thermodynamics, analysis of trend and change was undertaken using BomWatch protocols that are transparent, objective, and replicable, and cannot be ‘fiddled’ to achieve pre-determined outcomes.  

With the First Law onside, nothing can possibly go wrong.

Tmax depends on rainfall such that the drier it is the hotter it gets. If the relationship between Tmax and rainfall is not significant, weak or positive, something is wrong with the data, not the First Law Theorem. Further, if variation explained (R²_adj) is less than a benchmark of 0.50 (or 50%) data may either embed a ‘missing variable’ (one that is not explained by the naïve Tmax ~ rainfall case), or the quality of data is arguably too poor for determining trend and change in the climate.

Site related inhomogeneities occur in parallel with observations and are therefore confounded with the Tmax signal. However, as the Tmax ~ rainfall relationship implicitly accounts for the rainfall effect, non-rainfall residuals embed all other sources of variation, including underlying systematic changes related to site-changes. The second step in the Protocol investigates residuals for significant shifts or step-changes in re-scaled residuals indicative of such factors. Importantly, changepoints are detected objectively and cannot be specified in advance.

Identified using a factor variable, final-round analysis verifies that segmented responses to rainfall are the same (slopes are parallel), and that rainfall-adjusted segment means are different (that individual relationships are offset). Each of the three BomWatch protocol steps is transparent, objective and replicable, and supported by subsidiary investigations including analysis of residuals and post hoc tests.

Homogenisation is near the limit of plausibility

The same protocols used to analyse Tmax were used to evaluate homogenisation of the same data by ACORN-SAT.

Three iterations of ACORN-SAT applied adjustments at different times such that past-data were cooled and/or warmed to varying extents toward the present, which is a trick that affects trend. However, while adjustments may stabilise and improve statistical significance (and possibly moderate data quality problems), segmented relationships with rainfall become less clear-cut and less precise overall. Lack of statistical control and the absence of post hoc evaluations is a major weakness of ACORN-SAT.

The political narrative supported by ACORN-SAT will eventually be shown to be false, either as more ACORN-SAT sites are analysed using rigorous BomWatch protocols; or, as time passes and BoM scientists run out of options for making adjustments that seem plausible; or, adjustments cause fundamental Tmax ~ rainfall relationships to break-down. The question then arises: just how many more analyses are needed? or, how many years need to pass, before ACORN-SAT is shown irrevocably to be too unsound, unscientific and unbelievable to continue? As the ACORN-SAT project is unsalvageable, for the sake of those involved it should be abandoned without exception, in its entirety. 

Implications

The practical implication is that most messaging related to climate warming in Australia, including the indoctrination of vulnerable schoolchildren and young adults, and State of the Climate reports published by CSIRO, is demonstrably fake. The whole warming agenda has been made-up and carried forward since the call went-out from the World Meteorological Organisation in about 1989 to find and supply data that supported future Intergovernmental Panel on Climate Change (IPCC) reports. Temperature data homogenisation resulted from that.

The scientific implication is that along with models used to predict future climates, scores, possibly thousands of scientific papers and reports that depend on the warming narrative, are worthless.

Studies related to the effect of ‘a warming world’ on health, on agriculture, tourism, urban planning, the Murray-Darling Basin, the Great Barrier Reef, urban water supplies, species extinctions, and for converse reasons, on mining and resource use, are based on a premise that has been fabricated by consensus from the beginning. 

The political implication is that the billions of dollars that have been spent on, or are intended to be spent, in order to limit warming to the mythical value of 1.5^oC sometime in the future, is entirely wasted. With the national debit spiralling out of control, crippling electricity prices, subsidies and carbon taxes flowing from diminishing numbers of primary producers and workers to the elites, will eventually cripple Australia’s ability to remain sovereign, democratic and free. 

As The Science is underpinned by data that has been fabricated to support it, and as the manipulations of past-data will become increasingly implausible going forward, the whole edifice must eventually collapse. Collapse will probably occur within a decade, possibly sooner than 2030.

Finally, as Tmax depends on rainfall, which in Australia is stochastic (unpredictable) and episodic (occurs in episodes), without knowing rainfall in advance, it is impossible to predict the trajectory of Tmax into the future. 

Dr Bill Johnston

15 February 2024

Preferred citation:

Johnston, Bill 2024. Is homogenisation of Australian temperature data any good? Part 7. Victoria River Downs, Northern Territory, Australia http://www.bomwatch.com.au/ 15 pp.

Click here to download the full paper with photos, graphs and data

Click here for the full Victoria River data

---

[1] Former NSW Department of Natural Resources research scientist and weather observer.