Abstract
Weather forecast is of prime attention of the researchers working in the
smart agriculture domain. In India, approximately 55% of the total crops are dependent
on weather (monsoon season). An accurate weather forecast model requires abundant
data to get the most accurate predictions. However, the weather forecast is a key area of
research and is always challenging from historical data. Hence, the current system used
for weather forecasting is an amalgamation of forecasting models, opinions, and
information trends, and specific patterns. This work presents the application of the
linear regression model and polynomial regression model for weather forecasting; like
a scheme to forecast rainfall, and precipitation using historical weather data. The
sample weather dataset covers 75 districts of Uttar Pradesh state which is received from
the Indian Meteorological Department (IMD). Furthermore, analysing the impact of
forecasts with different parameters is realized over six major crops Triticum (biological
name of wheat), Gram, Barley, Mustard, Sugarcane, and Maize of Uttar Pradesh State.
The main objective of the state-of-the-art is efficient crop management and passing the
appropriate message to farmers to make suitable decisions as per the weather
conditions.