Hi - sounds like you have some interesting data!  Any ideas why the decline is occurring? Has there been a lot of change in the surrounding land-cover?

To your question:

We assume simple linear trends, i.e., a constant rate of absolute average change per unit time (a straight line on a graph of rain versus time as with your regression).  We want the units to match ... so we need a target ... say mm/day per year ... i.e. the average decline in daily rain that takes place on average in one year ...

So to convert yours (per-year) to per-day simply divide by days in a year

-59.9 mm/year x 1/365.24 days/year = - 0.164 mm/day per year

To convert the 50 year expectation to a single year just divide by 50

-1 mm day-1 50 yrs-1 x 1/50  = -0.02 mm/day per year

And the 100 year trend? Simply divide by 100.

-0.5 and -1 mm day-1 per 100 yr x 1/100 = -0.005 to -0.01 mm/day per year.

if you want these values as rainfall decline per year per year (how much annual rain declines each year on average) you would just multiply up by 365.2 again.

So your decline is severe compared to what others have reported! 

Hope that helps

Douglas

Many thanks Doug, your answer is extremely helpful and enlightening! Best regards,

Alexis

i agry with Douglas Sheil reply for Alexis Cerezo Blandón question.

 For directly calculating the linear trends and the slopes, You can use Kendal's tau and Sen's slope estimators