Seasonal update

As I'm back in New Zealand and haven't been here since the beginning of the growing season, one of the things I did was catch up on growing degree day (GDD) accumulation for Lincoln, Canterbury.

The concept of heat accumulation (one way of measuring which is GDD) and how plant growth corresponds with it is an important tenet of modelling the phenology, or seasonal development of plants, and in this case, grapevines.

It goes a little something like this:

Plants need heat to grow. At its most basic, this is because the enzymes that do the chemical work in plants can't function when it gets cold (in fact, this is true for animals as well -it's one of the reasons cold blooded animals hibernate in the winter. Warm blooded animals use energy to generate heat, which keeps the enzymes working). So below a certain temperature threshold, plants won't grow. As the temperature rises above that threshold the enzymes work faster and faster, up until the temperature gets so high as to prevent the enzymes from working again (around the mid 30s in celcius).

So measuring the accumulation of heat during the growing season results in a pretty good match of how far along the vines have come, or what stage they have gotten to. One question, though, is what do you set the temperature threshold at?

Based pretty much on enzyme activity in relation to temperature, 10C is the most commonly chosen base temperature (the temperature below which there is no plant-active heat accumulation). However, this may differ for different types of plants, and even for different times of the season (e.g. it seems that for the process of budbreak, a base temperature of 4C is omre appropriate, and for the first leaf appearances, 7C, Moncur et al. 1989).

Setting aside those special circumstances, a pretty straightforward way to quantify heat accumulation is to take the average temperature for a month (which falls within the growing season) and if it is greater than 10C, subtract 10 from it. That result is then multiplied by the number of days in the month, which gives the number of growing degree days for that month:

[ (average temperature for the month-10) * (number of days in the month) ] = GDD

In other words, on average each day was X amount over 10 degrees, and over the whole month, X times the number of days equals the amount of plant-useful heat that was experienced. Note that negative GDDs are not counted (though it's cold, the plants don't regress - they just sit there until it warms up again)

So to put some numbers in there, if the average temperature of November was 12.6C (as it was at Lincoln in 2009), 2.6C times the 30 days in the month equals 77 GDD accumulated.

If this value is calculated for each month of the year, we can follow the heat accumulation in a useful way, especially when comparing one year to another, or comparing on location to another.

In the case of the former, the graph for the 2009-2010 season looks like this, given the data collected up through November:

Growing degree calculations for the 2009-2010 season (up through November, the orange line) and for the long term average (LTA, the blue line), which is the average over the last 40 years.

The Long Term Average (LTA, blue) is the smooth sigmoidal curve, which signifies that heat accumuation is slower (the line is more horizontal) in the spring and autumn, and quicker (an more vertical line) during summer. The number in parentheses is the LTA GDD accumulation for Lincoln - a paltry 924GDD.

You can see that up to this point in the season, the orange line (current season) is below the LTA line, which means that the season has been cooler than average. If you squint just right, you can make out that the last orange dot corresponds to 77GDD, which is what we calculated above. This also means that there was no accumulation of plant-useful heat in October, or earlier in the season (at least, when based on monthly averages - more on that in another post!).

If you want to compare seasons, this way of looking at the data is fine, but if you want to see the differences more clearly, you can plot the current season's GDD relative to the LTA, which looks like this:

Growing degree calculations for the 2009-2010 season (orange) and for the LTA (the blue line, which is the X-axis). This figure is showing the same data as the previous one, but in a slightly different way.

Now it is pretty plain to see that we're veering away from the LTA. If the line is below zero (the LTA blue line), then there has been less heat accumulation than the LTA, and if it's above, there has been more, and it's been a warmer season.

You can also see more clearly what's happening for each month. If the line moves down, then the month has been cooler than LTA; if it slopes up, then it's been warmer. If the line is parallel to the LTA, then the average temperature for the month has been the same as the LTA.

So what we'd like to see is the line above the LTA - sadly, up through most of December, this has not been the case!