Warming in Wellington

In this guest post, first published last week in the Dominion Post, Jim Salinger looks at the long term temperature record for Wellington, and how it has been constructed. Jim’s currently the Lorry Lokey Visiting Professor in the Program in Human Biology at the Woods Institute for the Environment at Stanford University in California.

Climate scientists want to monitor how climate is changing and global warming progressing. This is particularly pertinent as this week the New Zealand Climate Science Education Trust are currently being heard in the Auckland High Court to try to persuade a judge to invalidate New Zealand’s temperature records which have been compiled and collected by the National Institute of Water and Atmospheric Research and the former government agencies. The coalition asserts the only way NIWA can claim a warming trend of 1°C over the past century is by cooking the books.

This century climate scientists are very interested in tracking climate as human factors are going to be the dominant influence on climate this century, save a meteor crashing in to the planet. They are interested in adjusting the readings as though they are taken from one location in an area. Wellington has one of the longest and best climate records of any region in New Zealand. This is why climate scientists carefully adjust temperature records.

When Sir James Hector, Director of the Colonial Museum in Wellington in the 1860s established a network to monitor New Zealand’s weather and climate, the primary stations were established for weather forecasting, so the priority on permanency of location of a climate monitoring site for climate change was lower. However we are indebted to Sir James’s Scottish heritage as in setting up the network he purchased precision thermometers which were housed in Stevenson screens to ensure consistency. Observations were taken under standard conditions, in his words ‘rigorous….’. This has given us a legacy of climate monitoring under rigorously enforced methods with very reliable observations from the 19th century, the envy of many countries.

Climate scientists, in constructing a climate record over the last 150 years for Wellington have to adjust the measurements taken for a number of reasons. The climate record has been taken from not one, but five sites. The various sites are at different locations in the area of Wellington City. These sites have different temperature characteristics and may be cooler or warmer than other sites because of factors such as different elevation above sea level, or be in city centres. By adjusting temperature series changes that are caused by climate, rather than the changes in site or environment can be monitored.

The diagram shows the record of mean temperature at the five sites: Knowles Observatory, halfway between the harbour and Tinakori hills; the Government Astronomical Observatory on a hill in the Bolton Street cemetery; Buckle Street; the Thorndon Esplanade; and then the current one at the top of the Botanical Gardens at Kelburn, 125 metres above sea level. This is the highest, and coolest of all the Wellington City long-term climate recording sites.


Figure 1. Unadjusted climate observations in Wellington City 1863-2011 from five different monitoring sites. The mean annual temperature is in °C. Wellington Airport is also shown for comparison.

To obtain a consistent record to monitor global warming locally adjustments are made so that the readings reflect those at the current site. To make these changes climate data are rigorously checked for any obvious errors. Adjustments are made by comparing a period of overlap between the old and new site, and comparing climate data before and after the site change with other neighbouring climate stations.

By these means a number of adjustments to the mean temperature have been calculated for the Wellington sites. All earlier sites are warmer than the current well ventilated Kelburn site by as much as 1.0°C because they are much lower in altitude.

It is from Wellington’s adjusted long-term record that true climate trends can be obtained. The long term record from Wellington City thus calculated shows that there is year to year variability which can be as much as 1.5°C between years. However there is an overall clear trend with Wellington mean annual temperatures now 1.3°C warmer by 2011 than in the early 1860s. This has been noticeable in the ability of gardeners to now grow frost sensitive plants in warmer parts of Wellington. Both the shorter climate series from the single sites at Palmerston North and Westport show extremely similar trends and variations, with mean annual temperature increases of about 0.8°C from the 1930s to 2011.


Figure 2. Long term climate series from Palmerston North, Westport and the combined record for Wellington City. Arrows indicate site changes at Wellington City. Annual mean temperature °C.

Long term monitoring is essential to detect small, but significant changes in climate. As records are taken from a number of different sites in a locality these require adjustments to reflect the true climate trends. Wellington’s temperature readings have been carefully adjusted with each site change to provide a consistent and reliable long-term record. The excellent climate record from Wellington shows a clear warming trend over the past 150 years.

5 thoughts on “Warming in Wellington”

  1. Fine. But what is the method used for calculating the adjustments? If we had 10 analysts do the work would we get 10 different results? Should the results of adjustments be random, or do they all point to a warming trend?

    Is there any agreed best practice technique? Did you use it when you were at NIWA? Did NIWA use it in its 2010 review? Are the techniques the same as BOM use in Australia? Or the same as CRU use in the UK?

    How do you make adjustments for gradual changes in exposure? What is the urban heat island effect in Auckland? Have these numbers been peer-reviewed?

    What assurance can you offer the public that the adjustments to the New Zealand record are “scientific” and robust, rather than a reflection of the analyst’s subjective opinions?

    These are all important questions, if we are to have confidence in people fiddling with the historic data from a century or more ago.

    1. Australis: armed with nothing else than high school physics or a meteorology chapter from a flight school manual you could satisfy yourself within minutes that the adjustments are within the ballpark range that you would expect with a very simple calculation on the back of an envelope. If you shift a measuring station upwards by 125m you will see a drop in measured temperature roughly equivalent to the atmospheric lapse rate. This depends on temperature and moisture levels but is normally close to of 1 Deg C / 100 m altitude change. Kelburn is 125m above the city elevations and will measure about 1.25 Deg C less than those. So when comparing the older close to sea level measurements with Kelburn you must raise the Kelburn measurements by about that much or drop the others, depending which of the two you want to make your fixed series. Assuming we take Kelburn as the new fixed series we drop the others by the adiabatic lapse rate and get a first order result that allows us to compare the historic with the new temperatures.
      If you do this the warming trend you get is very close to Salingers results already. Of cause for the scientific evaluation a more sophisticated calculation will improve on this.
      But what you get from this simple argument alone is a good deal of confidence that the reported warming trend is about where you would expect it to be.
      And honestly, who cares if Wellingtons trend was 1.3 or 1.1 Degree. There is no material difference in the realization and further, little NZ’s national measurements are just a very small element in the big picture. And how this looks must be clear to you.
      So your denier speak of calling the methodology of adjusting for station elevation changes etc. “fiddling” is simply entirely unfounded.

  2. Thanks Jim for the presentation.

    The graphs show quite compellingly why temperatures need to be adjusted. The fact that two locations can show virtually identical signature peaks and troughs but vary by a constant level of temperature seems to me the key point. With enough datapoints it shouldn’t be too hard to calculate an appropriate adjustment.

    As for the lawsuit, it seems to me that even if you could show a trend of 10 degrees in the last 150 years, that would still not galvanise the government to action. So I wonder why the bother. The denialists need not make fools of themselves since they are winning in the public apathy stakes already.

  3. Australis’ post shows that after hundreds of previous ones on this subject, he has learnt absolutely nothing – or far more likely, he is using deliberate troll-ignorance tactics, talking a load of malarkey. Go and get youself educated – it’s not rocket science!!

  4. Australis

    For a description the method used for calculating the adjustments for Wellington in the NIWA 2010 revision of the temperature series, you should read this:


    As to your question “Should the results of adjustments be random…?” do you mean “Should the adjustments be randomly distributed about zero?” If so, then my answer is “No, why on earth should they?”

Leave a Reply