Click Picture to Enlarge
As the bottles are far from rigid there was a tendency for the top of the gauge to deviate from the round. Towards the end of 1999 I overcome this problem by finding a very rigid yoghurt beaker (Rachel 500 gram) that had exactly the same internal diameter as that of the drinks bottle. This was slid into the inner bottle after first cutting off the bottom of the beaker. Mastic was then used to make a good seal between the two components as tests showed that in certain conditions rain could enter the gauge at this point. Another place where rain could enter was where the two bottles joined. If the inner bottle was not firmly inserted into the outer, rain running down the side could enter. This was probably why some of my earlier records seem rather high.
Actual calibration of the gauge was not attempted due to the difficulty that arises from the curvature and tapering of the bottles. Instead a reading is obtained by measuring in cubic centimetres the amount of water collected. This is then divided by a factor to arrive at the rainfall in millimetres.
The calculation to obtain the factor involved finding the cross sectional area of the opening of the gauge. Again because of the flexibility of the material it was difficult to determine the diameter accurately. Instead a narrow strip was cut from around one of the bottles. This was laid out flat and measured to give the internal circumference. The radius was calculated by dividing the circumference by 2 pi. From this the area was found by multiplying pi by the radius squared.
This gave me an area of 56.73 sq.cm. One mm. of rain falling on this area will total 5.67 cu.cm. So to convert the amount of rain from the gauge as measured in cu.cm. it must be divided by 5.67 giving the rainfall in mm.
I've intentionally laboured over describing the calculations, so that anybody considering using the data can decide for himself or herself if it can be relied upon sufficiently for their purposes.
Once I began using a Met Office gauge I was able to compare the two gauges. I ran the DIY and the Met Office gauges together for a year to compare them. On analysing the results over this period it was possible to fine tune the factor which I had been using to convert the volume of rain collected in cu.cm to milli-metres of rainfall. The correct factor was found to be 5.93