Not exact matches
The time series land based thermometer records are hopeless (not simply because of question adjustments and
homogenisation, and instrument error bounds) but also because that throughout the time series the
stations used with which the data is being compiled, at any one moment of time, is continually changing, so too their spatial coverage, such that at no time is like with like ever comparable.
The satellite has the best coverage and suffers least from UHI and errors in TOB
homogenisation,
station drop outs etc, and is verified independently against radiosonde temperature measurements, but it is only of short duration.
Each organisation in turn then subject sthe «raw» or «unadjusted» data to a series of «adjustments» /»
homogenisations» supposed to allow for
station movements, instrument / equipment changes and in some cases Urban Heat Island (UHI) correction.
Homogenisation may legitimately be used in climate science to correct for anomalies in data when
stations are relocated from one site to another.
This included taking in submissions and presentations from the scientists developing ACORN - SAT, as well as an examination of the Bureau's observations practices,
station selection methodology, data
homogenisation, data analysis methods and communication.
Homogenisation refers to the method of adjusting temperature records to remove artificial biases, such as the impact of a weather
station moving from one location to another.
Yet inland
stations are included in the
homogenisation process for Darwin.
Professor Muller said the warming was not the result of data bias caused by selective elimination of some weather
stations from the analysis, or the practice of «
homogenisation» to take into account changes in weather
station positions or instrumentation.
The point I am making has nothing to do with resiting
stations, I am addressing the idea that one can expect
stations sited within a few miles to produce similar readings, and that
homogenisation between such
stations could be a valid technique for improving data quality.