Modise Wiston and Mphale KM
Weather forecasting is one the important science applications in our day-to-day planning activities. This is one prominent application that has played a significant role to humankind from long way back. Wherever humans have settled around the world, weather forecasting has always been part of their life for man has always been actively involved in their surroundings. Early humankind relied on their philosophical experience and other reoccurring weather phenomena to predict the weather and infer what was coming their way. This was the knowledge gathered over many years of observations and passed from one generation to another. However, it became evident that natural knowledge was inadequate to precisely gauge the atmospheric changes; a growing number of scientists realised the need for more advanced and better ways of predicting weather. Numerical weather prediction (NWP) is one of the modern-day forecasting techniques developed through simplified systems of physical laws of the atmosphere. This entails atmospheric observations, measurements of initial conditions and computer-based modelling to provide a practical means of predicting weather changes. While this is a positive development, weather forecasting is still faced with significant uncertainties that sometimes lead to erroneous and/or bad predictions. In this study, we trace the science of weather forecasting from early human civilizations through Bjerknes’ and Richardson’s numerical computation to modern day operational NWP. We highlight and discuss the concepts of precision, assessment and evaluation together with factors influencing weather forecasting. Also considered is how uncertainties arise and how the ‘chaos theory’ has shaped weather prediction.