LEVERAGING MACHINE LEARNING FOR RAINFALL PREDICTION IN NORTH-CENTRAL NIGERIA: COMPARATIVE ALGORITHM STUDY

الملخص

Rainfall has been a major worry in recent times because of the weather patterns that are constantly changing. The last ten years, in particular, have seen significant changes in rainfall patterns due to global warming. The principal determinants of this rainfall pattern include precipitation, dew points, wind speed, pressure, temperature, and humidity. For the purposes of agricultural growth and precise rainfall forecasting, it is essential to comprehend the relationship between these variables and rainfall behaviour. This is particularly true for the north central region of Nigeria, which is known as the country’s “food basket.” This work aims to investigate how machine learning techniques might infer precipitation patterns in north-central Nigeria. This study looked at several algorithms and evaluated how well they performed in respect to each variable that was connected to the goal variable. We also examine the performance of several machine learning methods in predicting rainfall. The outcome demonstrates the potential of the Random Forest Regression Algorithm as a flexible method for comprehending and controlling rainfall patterns. Thus, it is advised that the Nigerian Meteorological Agency (NIMET) use the outcome in conjunction with the traditional NWP (Numerical Weather Prediction) method to further improve rainfall prediction. The result will help farmers optimise their planting and harvesting schedules, assist water resource managers in planning for different water usages, allow disaster management authorities to issue timely warnings, support the conservation of natural resources, and ultimately promote economic development through infrastructure planning.