This post provides both a visual detection and an analysis of the step down in RAW Tmax data at Cape Otway in 1908.
The analysis makes use of a subset of the data in the area, in particular the following stations (BoM ids in brackets): Melbourne Regional Office (86071), Wilsons Promontory Lighthouse (85096), Cape Schanck Lighthouse (86017) and Cape Otway Lighthouse (90015), see the map below for locations.
Data from other stations in the area (for example Lorne and Port Fairy) will be shown in later posts.
The following figure shows 12-month moving averages of RAW Tmax data from the four stations listed above:
The figure above acts as both a justification for correcting the step change, and as an analysis tool. The justification relies on the notion that it is not credible for Cape Otway to have the relative-to-neighbours warmth indicated for decades after 1908, but not before that date. The step down in 1908 is regarded as the consequence of a system change, probably the introduction of a Stevenson screen, providing better shielding of radiation and/or air circulation relative to whatever was there before. Later posts will cover the documentary evidence for system changes.
The analysis provided by the above figure are estimates of the date and size (annual average) of the step down. A subsequent figure will show the seasonal profile of the step.
The date of the step can be estimated visually from the 3 lines shown at 1908/9/10. It takes a year for a step to move through the 12-month average, so one expects to see a transition that lasts exactly one year. The centre of the transition period (the date of the step) can be determined visually by finding the start and end times, followed by averaging.
The result of this visual examination is a step date of “late” 1908.
The figure above can also be used to estimate the annual average size of the step, which comes out at around 1.2C.
The following figure, showing seasonal average anomalies, can be used to determine the seasonal profile of the step change. To clarify the process I’ve constructed a reference regional climate by averaging Cape Schanck and Wilsons Promontory:
It can be seen that there is a small but significant seasonal variation of the step down, a bit higher in summer months, a bit lower in winter. Finer detail can be obtained from similar plots of single months and pairs of months. In rough handfuls the maximum step down was in summer months and was around 1.5C. The minimum step down was at around 0.75C in winter months.
The modest but distinct seasonal variation of the step down suggests that there was some level of radiation screening and/or air circulation from the system in use prior to 1908, but not as good as that provided by a Stevenson screen. Later posts will check this conclusion against documentary evidence of the station and/or regional system history.