What it means to explain adjustments

After the New Zealand Climate Science Coalition released it's paper, "Are we feeling warmer yet?" there was much discussion around the necessity of adjusting data. Focus was on the Wellington data though, as I pointed out, there is no overlap between Thorndon and other sites given. So adjusting to a nearby site of the same elevation, while possibly reasonable, can be questioned: busy international airport, large amounts of asphalt (especially when considering the multi-decadal time difference), urbanisation effects. Even disregarding that Anthony Watts gives Kelburn a rating of CRN=4.

This is an issue of the legitimacy of specific adjustments. The paper above acknowledges the figures are adjusted, they question why,

At a minimum, the adjustments made to the official NZ temperature record must be made public.

Treadgold had also said,

The real issue is that adjustments were made by NIWA and not acknowledged publicly on their website....

NIWA responded initially to the paper with this article and referencing the same Wellington data. They also stated concerning the temperature adjustments (corrections) (as I mentioned in my earlier post),

NIWA climate scientists have previously explained to members of the Coalition why such corrections must be made.

NIWA have also expanded their explanation displaying the Wellington graphs previously shown by Gareth Renowden (as per my earlier post).

NIWA have also released their own graph from raw data. They used sites that have not been shifted since 1930.

I am not certain why they used a trend line without also including a smoothed line like they had for their earlier graph.

NIWA give more justification that New Zealand temperature is rising here. They mention several confirmatory records that show the temperature rise over the decades. And they refute several claims by the NZ Climate Science Coalition, specifically NIWA claims that the Coalition have had access to:

1. the raw data
2. the adjusted data (anomalies)
3. information needed to identify the adjustments made by Dr Salinger
4. information needed to develop their own adjustments.

The Coalition confirm that the raw data is available and they stated that in the original paper. The other 3 claims are not so clear. An email sent to the Coalition said that the adjusted (corrected) data was held by Salinger,

Dr Jim Salinger maintains the "corrected" dataset and is the best person to talk to you about it.

But the Coalition deny receiving all the corrected data despite requests for it. Further, the 7 sites labelled on the data they received differed from the 7 sites that were specified in other correspondence from NIWA. And even if they had the information to identify the adjustments (item 3 above), they do not have the reasoning as to why this was done. This also makes it difficult to do item 4. When the Coalition analyses the Hokitika data they see no good reason as to why it is adjusted as it has been. Wellington is the example that is repeatedly mentioned. What of Hokitika and the other 5 stations?

Which raises the issue of the post title. What the Coalition want, and what NIWA should have had available on their site before any requests were made, is both the raw data and the adjusted data, and the explanation of the adjustment. NIWA offer as their explanation this paper (doi:10.1002/joc.3370130807).

I will return to this paper tomorrow but the response from NIWA misses the point. It is not a methodology or technique of temperature adjustment the Coalition is asking for. It is why the adjustments were made for every adjustment.

• Adjustment A was made in this data set because of a site change. And the reason for this amount of change is because we took sites 1, 2 and 3 into account.
• Adjustment B was made because of a change in thermometer.
• Adjustment C was made to take into account the urban heat island effect.
  

One can mention the methodology as well if helpful,

• Adjustment... following the method of Smith for isolated stations.

It is then one can reproduce the adjusted data and discuss the merits of each adjustment.

• What about sites 4 and 5, they would attenuate the data by 0.1 °C.
• You haven't allowed for wind effects which are relevant before 1965 because...
• Smith was developed for continents and has not been validated for temperate islands, the method of Jones is more appropriate.

The when and why is needed so that this discussion can be had.