The seasonal phenology dynamics requires estimates of conditions within the growing season of a site, so we need a routine to identify the growing season within a time series of observations across years.

This basically just takes a year or longer time series of temperature values and their timestamps and the growing season is "just" where the temperatures within years are above 0°C, but needs:

• Arguments to tweak treatment of cold snaps: how long a snap breaks a season?
• We might also need hemisphere to handle array inputs that include northern and southern summers and hence where the season should be measured around winter not summer equinox.
• This is going to differ for models on daily data versus subdaily data: is it mean daily temperature or mean daytime temperature.