We started with usage of 
AGNPS model. This model is based on a geographic cell basis used to represent upland and channel conditions. Cells are uniformly square areas subdividing the watersheds, allowing analysis at any point within the watershed. Model components use equations and methodologies that are well established and extensively used by agencies such as the USDA.
AGNPS model is a one of the basis of efficient decision support system and reliable work of this model will reduces the time-consuming data input steps and scenario testing.
AGNPS model provide realistic estimates of nutrient loads in run-off water for a wide range of storm events. Lot of scenarios will be created on different grid sizes, on historic hydrological and other type of parameters. After this period we will have own perspective of coefficient of efficiency (CE).
In literature CE value is in a range of >0.82 for peak flows, >0.7 for run-off volumes were obtained.
The CE values are >0.990 for nutrient loads for nitrogen and >0.96 for phosphorus (Southern Ontario project).
As is already mentioned this model can simulates surface runoff, sediment, and nutrient transport primarily from agricultural watersheds. The weather parameters are also included. When the model is active the Storm Type is also included as 24-hour rainfall distribution. Storm Type can be an IA as the least intense and type II the most intense short duration rainfall. Type III represents areas where tropical storms bring large 24-hour rainfall amounts.
