Monthly rainfall data is obtained by adding up daily data on monthly basis. The process is not reversible. That is,  daily data cannot be obtained from monthly data. Information lost in the process cannot be retrieved. 

I too agree with the answer of Dr. Ette Etuk.

I also agree with the answer 

I think noboday can do it.

It’s definitely impossible - I have had this problem before in that the Swazi Govt wouldn’t give me their daily data because of the vagaries of missed manual collections (which could have been statistically smoothed if not too frequent)

the only way this can be done to have any degree of accuracy is to use the daily data from a similar climate to spread the monthly data, even if that is a little dangerous 

However, when monthly rainfall can fall in one day or thirty using any non-proven algorithm to spread monthly rainfall will undermine any conclusions you come to as a result  

Statistically it is possible by means of a method for temporal disaggregation of time series. You can then validate your results by comparing with the daily rainfall of an adjacent watershed during the same period or comparing with the daily data for the same watershed during another period. However, to be able to credibly interpret your results you must know your watershed very well. For information on temporal disaggregation of time series check out http://www.oecd.org/std/21781478.pdf

Alternatively, your can google "temporal disaggregation of time series" for additional literature on the subject.

Dear,

It is impractical to have daily rainfall obtained from monthly data-set. However Global weather Data site is also providing daily rainfall statistics that is being used in ARC SWAT. You may also explore the same.....

confirming the answer of Mr John Diiwu.

Sincerly

Dear Newton,

It is practically difficult to proportionate the monthly data back to daily rainfall values. This is particularly true for rainfall as there are unpredictable records. You can correlate with other weather daily values (such as remote sensing derived data and other climate datasets) but still not recommended. 

There should be a solution to get the original daily data which produced the monthly data used in the conceptual SIMHYD model (7 parameter RRL model). 

- You will find the daily data from the BOM (Bureau of Metrology).

Regards, 

Gebiaw

I partly agree to John Diiwu. Statistically it is possible. Normally daily rainfall are recorded through ORG and hourly in SRRG, and converted to monthly. In the reversible process may not be very much accurate. In case of necessity these can be validated through records of similar catchments of the region.

However, daily rainfall can be saggregated more accurately in smaller time steps. 

You can get global 5-day precipitation data from the USGS CHIRPS (Climate Hazards group InfraRed Precipitation with Stations) dataset. They use remotely sensed cloud temperature to estimate precipitation (CHIRP) and then calibrate to local topography/climate/etc with on-the-ground Station data. The same USGS group will soon have 10-day ET data.

You can download 5-day (pentad), 10-day (decad), monthly, or annual precip data using the GeoCLIM tool and you will see a number of other features on their webpage.

I've used estimated ET + CHIRPS pentad precip data to calculate soil moisture deficit (SMD) and, using an estimated maximum SMD, groundwater recharge. When the CH group decad ET data are available, I will refine my estimates.

I agree with previous posts that the only thing you can do with monthly data is estimate average daily precip, give up, or use any surrounding daily data to interpolate (hard to imagine a situation where you have monthly data in the area of interest and daily data nearby).

Yes it is possible to get daily rainfall data from monthly data by a method created by Prof.Gommes . In AquaCrop model the method is used. You can get more information by reading his book : Pocket computers in Agrometeorology (1983)....Good Luck

You can use temporal disaggregation method there are a number of stestical methods. But at the same time you will loose quality of your data.

Thank you all for your suggestions.

Another useful reference on this topic:

Salas J. D., Delleur J. W., Yevjevich V., Lane W. L. 1980. Applied Modelling of Hydrologic Time Series. Water Resource Publications, Littleton, Colorado, USA. 484pp.

I know its not a new thread.... Its as others have said - its impossible to get back exactly the original data - but an approximation to is doable to a certain extent. Think about imposed distributions (with variation) of wet-days and amounts. You might also consider a markov-chain: If it rains on one day, there is probably a higher chance it will rain the next. Of course this all assumes that the statistical data you are basing it on is sufficient - and not changing (eg due to climate change!)

Hi there,

You may consider using weather generator to generate daily data from monthly data. An example of which is " MODAWEC_Monthly To Daily Weather Convertor ".

Also see: Using Monthly Weather Statistics to Generate Daily Data in a SWAT Model Application to West Africa. Ecological Modelling 201(3-4):301-311

download modawec video and follow the method. thanks