How to calculate standardized precipitation index (SPI).

I have rainfall data in .mat format. Data is 2D matrix mxn where m are row representing m days and n is column representing n locations. I wanted to calculated SPI for each location as well as whole country. Please do not suggest any softweaer.

[dpb]

MATLAB isn't "softwear"???

What is a SPI, precisely?

Where did you run into a problem in your implementation to calculate it, specifically?

[Govind Kumar]

yes matlab is softwear but i mentioned this because sometimes people provide link to download softwear instead of solution in matlab. SPI mean standardized precipitation index. It is a drought index which is used for drought trend analysis.

[dpb]

Well, what's the actual definition one could use to calculate from? We're not all specialists in various fields...

What have you tried so far?

[dpb]

Turns out this is quite a lot of work...see https://journals.ametsoc.org/doi/full/10.1175/JAMC-D-14-0032.1 and the definitive (apparently) work https://climate.colostate.edu/pdfs/relationshipofdroughtfrequency.pdf

They're fitting multiple distributions to find best fit of precipitation data over the given time period, then using that to compute a percentile.

Search the FileExchange; there are several submissions there that should at least give you a head start...this won't be a five-minute exercise.

Answer 1

[Chad Greene]

Hi Govind,

You might want to check out the spei function in the Climate Data Toolbox for Matlab.

Chad