You just have to take it as a definition.

For the purpose of model evaluation, observed features of extreme rainfall and temperature are compared with those in the control experiment. Daily rainfall data of about 1000 stations well distributed over India have been considered for the period 1901-90, depending on their availability. Using available daily rainfall data, the observed spatial patterns of rainy days, intensity and extreme rainfall have been analysed. A rainy day is defined as a day with a rainfall of 2.5 mm and above, as per the operational practice of the IMD. Using this definition, the average number of rainy days per year has been calculated for each station. Intensities were also calculated for these stations by dividing total rainfall on rainy days by the number of rainy days.

Regards

The amount of rain measured in weather stations over a long period of time gives that value. It is also given in inches. Less than that will not be called rainy. This is compared along the climatic regions. I believe snowfall is also so measured. It is not an arbitrary figure but as recorded in meteorological stations.

Obianuju P. Umeji

I don’t know the exact answer. 2.5 mm or 0.1 inch is well above what can be measured. My assumption would be a rainy day has enough rain to deliver water to plants, as some is absorbed by litter and leaves and likely to evaporate. But I suppose in arid climates, rainfall of less than 2.5 mm could be significant. Although best to stick with standard approach, I am doubtful there are any rules that would prevent using also 1 mm for an expanded analysis or description of conditions. Some areas experience some augmentation of rainfall through condensation onto tree or plant leaves.