Development & Analysis Of A UK Storm Indicator (part 5)

Today I take a different approach to analysing the source data and wield the big spanner that is autoregressive integrated moving average (ARIMA) time series modelling

Let’s look at mean wind speed in a different way before we move along the bus.

What I’m going to do is slice the data period up into 30-year chunks, this interval being the standard for a climatological normal. I’ll start with the period 1844 – 1873 so I can end squarely with 1994 – 2023, for data records are still incomplete for 2024. I can then produce an error bar plot of monthly means by data period so we can get a feel for the seasonal dynamic over time. Try this colourful chappie:

I rather like this, and I hope you like my rainbow crayoning! We can now eyeball the red hot current era (1994 – 2023) and see how those red blobs fare against the historic cooler colours on a month-by-month basis. If we start with January we can see that 1964 – 1993 was a particularly calm period for these islands and this contrasts greatly with 1994 – 2023. Anybody doing analysis using such data is going to come to the conclusion that wind speeds are up… ‘coz climate change. Sensible folk enjoying a pi…