Weather Forecasting Using Weather Forecasting

1534 Words4 Pages

Abstract—- Weather is an ever changing phenomenon. It has various components such as temperature, rainfall, cloud cover, humidity etc. Weather affects every sphere of human life, whether it is business or Entertainment. Weather Forecasting is one of the most technologically challenging tasks. Weather forecasting using Data mining is an innovative way of predicting the weather with minimal cost and maximum accuracy.

Introduction The need for weather forecasting has been amply justified over the years. Weather forecasting has existed since times unknown. The method for weather forecasting and the technology used has since massive improvement over the years. In ancient times the weather cock was used to determine the direction of the wind, the …show more content…

The common thread between all these methods is that the prediction is made on basis of the past occurrences of the characteristic of weather. It is thus noted that weather records for multiple hundreds of years are available for our perusal.
Since such humungous amounts of historical data is available to us using data mining to predict the weather seems to be a smart option. Predictive analysis is a sub type of data mining which is centered on pattern mining. Data mining to predict weather is a smart, effective and accurate way to get optimum results for predicting the weather.
Today, solar power has become part of our daily lives. Appliances like solar notebooks, solar air-conditioners, solar cars, etc. demonstrate the use of the sustainable power of the sun. As the adverse effects of burning of fossil fuels and the depletion rate of non-renewable energy sources increase, the future of solar energy looks bright. The problem with renewable sources of energy is that they are not easily predictable in advance and vary based on both weather as well as site specific conditions. …show more content…

To combat noisy data the dataset was first preprocessed and the missing values were replaced with a global constant. When we encounter a situation where the data to be considered is a missing value. The preliminary check performed identifies the global constant and the data from the previous day is used instead.
2. The second approach we adopted was to set a maximum and min criteria for the parameter to be predicted. We analyzed data for the past 5 years and we set values for the max and min occurrences for every month, if the system encounters a value which does not lie between the range of the maximum and minimum value it is scrapped as garbage value. In its place the previous day’s data is used.
3. Predictions are always tentative and many factors such as global warming affect the climate. If a prediction is made by the system which falters we adopt the feedback method. The predictions that we make are cross checked with the prediction from an authentic source and if the predictions are not found admissibly accurate the actual accurate predictions are placed in the system instead of the faltering

More about Weather Forecasting Using Weather Forecasting

Open Document