The thought recently struck me that the data I’ve been collecting in Hillhouse Wood for the TrackaTree phenology project might enable me to look for any noteworthy differences between years. (Phenology is the study of how the timings of the seasons are affected by variations in weather. The TrackaTree project requires fieldworkers to record data on the dates when certain specific things happen. These ‘things’ include bud break and leaf burst on particular trees, and the numbers of certain wild flowers that are out within a set radius of these trees.)
I thought it might be interesting to see what charts this data would enable me to create. Looking back at my records I found that (to my surprise) I started in 2014 (doesn’t time fly?), and that the data that would be best to chart would be the numbers of flowers I had recorded as being out. The data was recorded as individual dates, and the actual numbers were recorded as a band: 1-5, 6-10, 11-25, 26-50, 51-100, and 100<. To complicate things, the recordings were not on the same day of each month in each year. I had useable data on three species of flower – Wood anemones, Lesser celandine and Bluebells. Of these, Bluebells are by far the latest to flower.
To be able to draw a line chart (which I thought would be the best type of chart to show changes between years), I decided to:
- Convert each band into a single number, using the median of each band, thus:
100< 125 (arbitrary and overly simplistic, but practical. However it means that the shape of the lines on the charts are ‘flat-topped’, and potentially useful data has been lost.)
- For the days in between those on which data had been recorded I used the same number of flowers as in the previous readings. This meant that the chart lines would go up and down in steps.
- I found that in some years I had stopped recording Bluebell numbers before they had stopped flowering. So for these years I added ‘plausible’ made up numbers to achieve an end to the chart lines.
- Of my four TrackaTree trees, the mix of species flowering under them varied in some years. So, to keep things simple, I simply added together all the flowers of each species that flowered under at least one tree, to achieve a total for all four trees.
Having decided on all this, I started work. The only place where my records exist is in my account on the TrackaTree website. This page was not designed for someone to do what I wanted to, so the task of finding each record and entering into a useable spreadsheet was a laborious pain. Doing it took two evenings and some good beer. Once the data was in the spreadsheet, manipulating and creating the charts progressed quickly. I will show you what I’ve found in the next post – aren’t I a tease?