Hi Mohamed,

The equation that you stated is something called the I-D-F relation (Intensity, Duration and frequency or return period). However, to develop such equation, you would need the time series data for precipitation, so you could relate for various duration's and frequency to find the curve.

I think you could try the process of "regionalization" or parameterization of your watershed based on similar watershed with rainfall data available. For example, lets say you want to study the watershed 1(W1), but you have data only for the similar watershed 2(W2). You could relate the outflow at the similar watershed outlet (W2) to some parameters: area (A), precipitation(P), slop (s) and other characteristics of the watershed. Since the two watersheds are similar, you could approximate the outflow for W1 based on W2. 

Q(W2)=Q(W1)=f(A,P,s,etc)

You could do this by various methods, by i think linear regression should be a very good start.

Most of my read in this topic is portuguese, so I reserached some papper in English, hope helps:

https://s3.amazonaws.com/academia.edu.documents/30728098/awr2007_yadavwagenergupta.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=1503619441&Signature=ABY1g97dNI4U2eLSGP5u7Dm3z%2BA%3D&response-content-disposition=inline%3B%20filename%3DRegionalization_of_constraints_on_expect.pdf

http://ntur.lib.ntu.edu.tw/bitstream/246246/85363/1/5.pdf

https://scholar.google.com.br/scholar?hl=pt-BR&q=Hydrologic+Regionalization+of+Watersheds&btnG=&lr=

Yes you are right it is called (Intensity, Duration and frequency or return period) curve.

In my case I have no other similar watershed to use this "regionalization" method.

All I have is monthly rainfall data and maximum 24 hour precipitation for a return period of 30 years and I want to calculate rainfall intensity to get an idea about the peak discharge.

So the IDF method is not working because of lack of data.

Hi Francisco

I was wondering if I can extract these montana coefficient when I have the IDF curves.

I found one research paper for an area in Kongo and they draw the IDF curves.

Article Construction of Intensity–Duration–Frequency (IDF) curves fo...

Hi Mohamed,

You can use the following options:

1). If you have a gauged catchment outlet somewhere downstream of the catchment, then you can use the Soil and Water Assessment Tool model, SWAT to calculate the ungauged catchment extreme events ( high and low flow extremes). You may use either of the optimisers: Sufi-2, GLUE, MCMC, Parameter solution(ParaSol) or PSO. This works under the concept of reginalization. 

2). Use grid climate data ( including rainfall) and static data ( DEM, Soil, and land use) as input to predict the catchment discharge.

3). The rainfall data may be found under 'https://pmm.nasa.gov/data-access/downloads/trmm'  from NASA Precipitation Measuring Missions, (TRMM).  

I wouldn't use the rational formula  to calculate catchment peak discharge!

Sincerely, 

Gebiaw

You can try flow accumulation in Spatial Analyst tool of ArcGIS. You need to have a digital elevation model and average rainfall data

Hi

I'm in broad agreement with Francisco comments.  But I personally believe that geomorphological based approach is more reliable than other methods for estimation of peak discharge especially in data sparse regions. I suggest you to use Nash, GIUH, Snyder or ... methods to determine the catchment's unit hydrograph. Then by field observation estimate the maximum Q. For estimation of this discharge you should measure the maximum observed water tails in main channel and then by using manning equation calculate the maximum discharge.

finally by comparing this discharge and the peak of Unit hydrograph you can estimate the effective rainfall and use it in other methods.

the best

Azizian 

Hi again

Another solution that I think can be used is based on spatial pattern of rainfall in your study areas. In HEC-HMS you can find a method that called SCS Storm. in this method all USA divided into 4 groups and for each group a specific spatial pattern of rainfall  has been presented. In states that fall into first group you should use Storm Type I.

I think you should find or create curves like SCS Storm method.

the best

Azizian

The other solution is by using the Synthetic unit of hydrograph method published by Snyder. This method only is based on the characteristic of catchment area. Hope you can try it..

I have a same type of question...How to develop rainfall IDF curves for ungauged or data scarce sites/regions?

I ma asking for methodology/mechanism for future projection under changing climate. Can we use griddle data for future projection under changing climate scenarios for ungauged sites?