Hi Rahul,

in my opinion, the answer is “both, and it depends”. The 31 station data set should provide a valuable record of rainfall at points throughout the catchment. Calculating drought statistics for these sites should offer valuable insight into spatial and temporal drought dynamics. You do not say where the stations are located, many pets of the world are covered by one or more gridded climate products which present interpolated rainfall data. Analysis of this kind of data set could be complementary to an analysis of the 31 station instrumental data set. I do not know the purpose of your study, but it could be informative to compare interpolated drought statistics based on the 31 stations to regional statistics obtained from an analysis of a gridded climate product.

hope this helps,

Martyn

What you mean by drought analysis? What the temporal sampling and spatial spacing of the rain-gauge-station data? What is the length of the record at each station?

Is the data heteroscedastic? Why?

Depends on topography and rainfall distribution/gradient. 45000km2 is commonly considered too big for useful analysis. You could also check the global raster data bases, they may not be better but have a higher resolution.