1. Rainfall Disaggregation models - Refer to work of Koutsoyiannis (Greek) or earlier work of Valencia and Schaake (1973), Lane (1979)
2. Artificial neural Network modelling (as already been suggested by other members)
Koutsoyiannis, D. Onof, C., and Wheater, H.S. (2001). Stochastic disaggregation of spatial-temporal rainfall with limited data. XXVI General Assembly of the European Geophysical Society Nice, France. Koutsoyiannis, D., and Manetas A. (1996). Simple disaggregation by accurate adjusting procedures, Water Resources Research, 32(7), 2105-2117. Koutsoyiannis, D., and Onof, C. (2001). Rainfall disaggregation using adjusting procedures on a Poisson cluster model. Journal of Hydrology, 246: 109-122.
Kumar, D.N., and Lall U. (2000). Multisite disaggregation of monthly to daily streamflow. Water Resources Research, 36(7): 1823–1833.
I would first calculate the correlation between both stations for monthly data, which you have and you can calculate for station A. If correlation is trustworthy, then I would use a double mass curve or use a model to infer daily data for station B. If you don't get a good correlation, I would probably search for other stations or you will have to use station A alone.
Always keep in mind that the interpolated data, by any method, it will not be very useful for predicting extreme values. So, it will depend on what you want to do with the data.
Hi Mr Sultan. Consider doing a regression analysis between the monthly data for stations A and B between the years 1970 to 1996. This will give you an indication of the level of correlation between the 2 stations. Then use one of the available models incorporating some environmental factors as proposed by Satish. Confirm that your generated data has the same correlation that the original data has (i.e. monthly data correlation).
Using the long term data from Station A, and the calculated values for Station B, determine the relationship between Station B and any other Stations you have data for. Apply this relationship to the days for which there are no data at Station B to the years for which you have no data. Make certain that you document the manner in which you generated the calculated values.
I agree with Jagat. More stations will generate a better value. Take care to examine the location and physiography of the stations you use. Microclimatology demonstrates that small changes in distance can create large changes in precipitation. Using the daily precipitation from Station A and any other stations for which you have data, determine the percentage of the monthly precipitation for each day and apply that to the monthly precipitation at Station B.
Take the monthly data and each month divide by 30 for an average on that station. Start taking daily readings from now on. If you want to guess on the coming rainfall I am sure there are many models that require money and educated guessing. You control the rain in your area. basic science using microbes id the answer. The biggest deserts can become ag fields and forests in about 5 years. Just need to start. Not difficult and no funding. Nature does the work. You provide the waste. Enjoy.
The reliability of the estimates will depend on how close station A and station B are and do they they fall on similar region. If you use more station to generate data will be better. If stations are in mountainous area, please use computer model that used isohytel method to estimate rainfalls. Jagat