This post shows the manual detection/confirmation of suspect monthly Tmax data recorded at Queenscliff, just South of Melbourne. This station is important, as it has data back to 1898, and is one of only 3 Melbourne-area stations with data in the 19th century. Probably of greatest significance is that most of the data is free of major glitches and has a respectable level of consistency with its near neighbours.
The following figures show anomalies (offset for display purposes) of RAW monthly Tmax data from the 3 Melbourne-area stations with data in the 19th century (BoM ids in brackets): Queenscliff (87054), Cape Schanck (86017) and Melbourne Regional Office (86071).
JULY TO DECEMBER:
JANUARY TO JUNE:
There are clear examples of data spikes in the months of July, August, September and October in the data from Queenscliff, otherwise the data looks pretty good and consistent, both in terms of trend and year-to-year fluctuations. The trend differences that there are may be due to differences in the amount of sea influence, highest at Queenscliff and Cape Schanck, lowest at Melbourne.
In reconstructing the past climate it helps to first correct spikes such as the ones shown here, but only after checking data from nearby stations. Many would-be spikes or values that look unusually low for a particular month are often found to be genuine weather when checked against other data.
I estimate the sizes of deviations from the local norm visually from graphs such as the ones shown above, but usually with fewer months per graph in order to expand the temperature scale of the vertical axis. All such corrections will be published here later on a dedicated page.