Content

image

A Simple Weather Forecasting Model Using Mathematical Regression

Paras and Sanjay Mathur

Abstract

A simple model for weather forecasting has been described. The model is simple due to the fact that it uses simple mathematical equation using Multiple Linear Regression (MLR) equations that can be easily understood by a medium educated farmer. Weather data at a particular station is recorded which is a time-series data. The weather parameters like maximum temperature, minimum temperature and relative humidity have been predicted using the calculated features depending upon the correlation values in the weather data series over different periods from the weather parameter time-series itself. Relative humidity is also predicted using time series of maximum and minimum temperature and rainfall. The category of rainfall has been estimated using features of maximum and minimum temperature and relative humidity. The development phase of the model is to obtain MLR equations using input set and output parameter. The coefficients of these regression equations have been used to estimate the future weather conditions. The personal computer and simple data processing software like MS Excel can be used to make and validate the model by the user itself. The results obtained show that MLR model can estimate the weather conditions satisfactorily

Keyword: Weather forecasting; Time

Full Text

Archive Content | Society of Extension Eucation, Agra
image