Create calc_anomaly() to work on all time series, and then use the existing anomaly-type plots.

[andrew-edwards]

• Want the function to work on any time series, such as average temperature over a region (from bccm or OISST), hake biomass, etc.
• Probably requires adding a standard Date() column to each data set [except the spatial ones], while keeping the year and month columns; will probably pick 1st of month for months, but emphasise in the help that the values are for the whole month (e.g. averaged over the whole month).
• Adapt plotting functions to use these. Maybe just need to add the class pacea_anomaly and then create plot.pacea_anomaly().

[andrew-edwards]

• Above ideas might be too hard. We can write calc_anomaly.buoy_sst() type functions to do explicit anomaly calculations for each class we have. Will be easier than trying to standardise everything, especially given the second point above.

[travistai2]

I wrote a climatology and anomaly function that handles the data objects with class: pacea_st, pacea_buoy, and pacea_oi.

However, the plot functions will need to be adapted for the BCCM and OISST outputs as the format/column names change. Acutally I still have to write a plot function for OISST.

[andrew-edwards]

For consistency we should use the anomaly definitions in Angelica's BCCM paper 2019 paper, plus it has shelf, slope and offshore regions (which she's sending you - hopefully all three regions).

I also read a few papers about marine heatwaves and have some definitions in those we can incorporate at some point.

[travistai2]

The only thing we haven't done here is create a plot function for buoy sst anomaly outputs.