Sri Lanka is an islandsituated southeast of the southern tip of the Indian sub-continent. Floods, Draughts, Landslides, and Storms are the common occurring disasters happen in Sri Lanka. Social, Environmental, Economical and Physical impacts arise due to these disasters. In order to minimize these impacts it is imperative to have earlier knowledge on possible disaster occurrences. These disasters are fully or partially connected with rainfall variation. In this background it is imperative to have a reliable forecast of rainfall. 
 The study was carried out for 17 districts out of 25 in the island and it attempts to assist Agriculture, Urban and Regional Infrastructure Management, Environment and Water Management of those areas to its best capability which can take early actions on flooding events and long term drought periods due to rainfall changes. In the study Numerical methods were used to find out the mathematical models of the Northeast and Southwest rainfall for upcoming years.The objective of this model is to forecast the rainfall for near future by using previous years’ rainfall data.Statistical methods are the most common methods using for rainfall forecasting although the existing statistical model cannot be used for feeding different data sets. Mathematical methods can be used to forecast annual rainfall as well as monthly rainfall even though statistical methods can be used to forecast either annual rainfall or monthly rainfall. 
 Polonnaruwa and Batticaloa districts shows the highest monthly rainfall variation and Hambantota and Mannar shows the lowest in 2014 and same in forecasted monthly rainfall in 2015. Kegalle district shows the highest mean annual rainfall in 2015, it has an increasing rainfall trend while Hambantota being the lowest as per the model and they also have slightly increasing rainfall trend. By using these findings people and authorized agenesis can take pre disaster responses beforehand.