Climate of the Great Barrier Reef, Queensland: Climate change at Gladstone – abstract and case study

Dr Bill Johnston [1]

Dr. Bill Johnston’s scientific interests include agronomy, soil science, hydrology and climatology. With colleagues, he undertook daily weather observations from 1971 to 1979.

Abstract

Main Points

  • The weather station at Gladstone Radar marks the approximate southern extremity of the Great Barrier Reef.
  • Temperature and rainfall data are used to case study an objective method of analysing trend and changes in temperature data.
  • The 3-stage approach combines covariance and step-change analysis to resolve site change and covariable effects simultaneously and is widely applicable across Australia’s climate-monitoring network.
  • Accounting for site and instrument changes leaves no residual trend or change in Gladstone’s climate.

Background

In Part 1 of this series, temperature and rainfall data for Gladstone Radar (Bureau of Meteorology (BoM) site 39326) are used to case-study a covariate approach to analysing temperature data that does not rely on comparisons with neighbouring sites whose data may be faulty.

Advantages of the method are:

  • The approach is based on physical principles and is transparent, objective and reproducible across sites.
  • Temperature data are not analysed as time-series in the first instance, which side steps the problem of confounding between serial site changes and the signal of interest.
  • Changes in data that are unrelated to the causal covariate are identified statistically and cross-referenced where possible to independent sources such as aerial photographs and archived plans and documents. Thus the process can’t be manipulated to achieve per-determined trends.
  • The effect of site-changes and other inhomogeneties are verified statistically in the covariate domain. Thus the approach is objective and reproducible.
  • Covariate-adjusted data are tested for trend and other systematic signals in the time-domain.

Further, statistical parameters such as significance of the overall fit (Preg), variation explained R2adj and significances of coefficients provide an independent overview of data quality.

Click here to download the full case study including photographs and tables of data used.

[1] Dr. Bill Johnston’s scientific interests include agronomy, soil science, hydrology and climatology. With colleagues, he undertook daily weather observations from 1971 to 1979. 

3 thoughts on “Climate of the Great Barrier Reef, Queensland: Climate change at Gladstone – abstract and case study

  1. Bill, can I suggest that you apply your methods to some synthetic (but realistic) data with a known amount of gradual 20th century warming, and see if you can deduce the size of the warming. That exercise would give greater confidence in your deduction of zero warming in NE Australia.

    1. Thanks Michael,

      Adding trend to data is not difficult; STARS also can create an artificial dataset with steps. The challenge has been to separate effects due to the weather from those caused by site changes – so-called inhomogeneties.

      The climate (accumulated ‘weather’) can also not be homogeneous – there was a shift in rainfall in about 1947 for instance. Accounting for dependencies (covariance) between maximum temperature and rainfall removes that effect, which allows underlying ‘non-climate’ changes to be investigated and attributed using independent sources of information – photographs and archived reports etc.

      Cheers,

      Bill

  2. Bill take a look at the plots that I display at my website. http://www.nvtech.com.au/Climate/AGW.html

    The energy in the atmosphere is described by Bernoulli’s Equation, ie, it is the algebraic sum of the thermal energy, ie, temperature*specific heat of the system + kinetic energy of the system + potential energy of the system (the result of pressure). In an adiabatic system, or one where the enthalpy does not change, when there is an increase in kinetic energy (eg wind velocity) then the potential energy (exhibited by the pressure) will drop as could the temperature of the system. If one part of a system is cool, chances are another part will be warm such that the total enthalpy of the system is unchanged. That is why it is necessary to consider all temperature stations within a network and then note the number of times over a certain period a particular temperature threshold is exceeded by any sensor in the network.

    The result of such an analysis shows that both for the US climatology network and for the Australian BoM network the climate is cooling, not warming.

    You will notice that only in recent years has the BoM graph started to diverge from the US graph and that is because the BoM has been fiddling with not only the method of data collection but the data itself.

    Hope you enjoy the website and find what I have presented there thought provoking and interesting.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.