1
$\begingroup$

I am seeking some assistance in how to take into account difference in trials periods when calculating average rainfall. I am investigating the effect of rain on plant disease severity. My trials have been conducted in different years, and the length of the trial varied in different year. I have been asked to use average rain instead of total rain to take into account difference in trial periods. Our weather station recorded total rain per day (at daily intervals), I need to calculate average rain per day. To achieve this, I have been asked to divide total rain in the trial period by number of days rain occurred (rainy days). Is it the right way to calculate average rain per day? Or we are misleading the results by calculating total rain per day instead (maybe in an incorrect way)? Please see blue column to see what I'm talking about. Thanks very much.

enter image description here

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

If we have (as in your example) within a particular year (2006) 41 rainy days and 156.3 cm of total rain, the average is indeed 3.81cm per rainy day. That said, your intuition is correct, this is not the whole story. It makes sense to also have the average rainfall per trial day, i.e. 2.27 cm per trial day. These two statistics help us capture different aspects of the rainfall's impact. The first one gives us an idea of how much it rained "overall" and the second indicator of how "strong" (temporally localised) the rainfall was.

For example, in a 20-day period if it rains every second day by 3cm that that means we have 30/20 = 1.5 cm of average rain per day and 30/10 = 3 cm per rainy day. For the same plant and period if we had only 2 days of 15 cm rainfall each, we would still have 30/20=1.5 cm of average rain per day but 30/2=15 cm per rainy day, which obviously will have a much different impact on the plant's well-being.

So in effect, neither you nor your collaborators were "right" or "wrong", just aimed to capture different aspects of the rainfall's pattern.

Improve this answer
$\endgroup$
8
  • $\begingroup$ Thanks for your comment. Perhaps I didn't explain my data correctly. Taking 2007 as an example, the trial period was 65 days. Rain occurred in 39 of 65 days, while 26 days were rainless. The average rain is 6.7 cm per day. How does this entail all the days of the year? I have weather data for 65 days only. I would imagine 6.7 cm rain occured in rainy days only (39 rainy days in 2007) because I have calculated average rain per day as total days/rainy days? Sorry, I am not a mathematician. I know avg rain per day is incorrect but looking for a simple answer to satisfy my supervisor. Thanks again $\endgroup$
    – Ahsk
    Commented Dec 13, 2022 at 19:44
  • $\begingroup$ I am only concerned with the trial period, not whole year. I planted plants on specific date and let them infected with disease plant materials. At the end of trial in each year,I recorded disease severity on each plant and tried to related them with weather conditions from planting to disease assessment date. I used total rainfall but I was told that this is not correct because there is a difference in trial period. To account for difference in trial period, I was told to calculate daily avg rain in RAINY DAYS. I know this is not correct but not sure how to explain this the concerned person. $\endgroup$
    – Ahsk
    Commented Dec 13, 2022 at 19:52
  • $\begingroup$ Thank you for the clarification. OK, I understand now. In that case it makes more sense to have the following "average rainfall per day within period" (i.e. total divided by the period length) and then "average rainfall per rainy day within period" (i.e. what you have now.) The first one gives you the idea how much it rained "generally" and the second indicator how heavy the rainfall was. For example, in a 20-day period if it rains every second day by 3cm that that means we have 30/20 = 1.5 cm of average rain per day and 30/10 = 3 cm per rainy day, for the same plant and period if we (cont.) $\endgroup$
    – usεr11852
    Commented Dec 13, 2022 at 22:29
  • $\begingroup$ have only 2 days of 15 cm rainfall each, we still have 30/20=1.5 cn of average rain per day but 30/2=15 cm per rainy day, which obviously will have a much different impact on the plant's well-being. (I will edit my answer) $\endgroup$
    – usεr11852
    Commented Dec 13, 2022 at 22:31
  • $\begingroup$ Thank you for the edited answer. This is exactly what I was looking for. In your example, same trial period and same rain but different distributions have yield completely different avg rain per rainy day. Since the experiment was conducted in natural environment and the rainfall distribution could not be controlled, the avg rainfall per day will likely lead to misleading results. I am communicating this example with my collaborators. I am upvoting your answer. Thanks for assistance. $\endgroup$
    – Ahsk
    Commented Dec 13, 2022 at 22:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.