Dear Dr Malka

I recommend you read the theses and articles as an attached file.

Best regards

Fit your series of rainfall data with a GEV.

For most utility, check to find out the detailed records or strip charts. If one rainfall reading or value a day is all there is, you are limited in addressing shorter period extremes. If data have been entered into excel, combining, sorting, etc. can be used to address 2day, 3 day, or other time combinations, by year, decade, etc. You or someone would have to program excel to do this. The initial time series over 30 years could have column for year, month, day and value in separate columns. Various added columns could be constructed for 2 day, 3 day, ... totals, etc. Plotting time series data can also be helpful in visualizing, and expanding Time on x axis allows for spread of data from many years as desired. Copying these various duration columns in a separate worksheet as values, instead of formulas, allows for sorting to find maximums if so desired, or can be done with visual inspection of data in columns.

The initial data could also be sorted from high to low, and ordered from high to low by number from highest day as 1, and lowest day as number 10,950. The high day value has a frequency of 1/10950, next high day is 2/10950, etc, etc, until all the days are calculated. There are tricks in excel to do these calculations quickly, by selecting example equation cell, and pulling the edge down through the rest of the column. Much like flow duration curves, these daily rainfall values could be plotted by frequency probability based on the 30 year data. There are other ways of plotting if there is desire to portray frequency curve for 30 years, and then attempt to extend curve beyond to 50 years or more. Confidence limits can be calculated, etc.

Calculating storm runoff from only daily rainfall would be difficult and likely poor quality. If you have one or more stream gauging stations in vicinity of rainfall station that might help apply or match daily rainfall to time adjusted flow. Various tools are sometimes used to estimate runoff, but they work best when validated or adjusted to conditions. Some use the SCS(NRCS) curve number approach. The Rational Equation is applied to small catchments or urban areas.