Seasonal divisons and annual temperature predictions

This work is an explicit method for the analytical predictions of the seasonal divisions and the annual temperature behavior as a function of time at the Douala airport neighborhood in the equatorial region. The daily mean temperature measured during the whole year from 2000 to 2009, are analyzed in a thirteen months calendar. The experimental points are the monthly mean temperature. The nonlinear least square method is used to look for the theoretical function which fit to the experimental curve. The temperature changing with time from a year to another, a polynomial function of time is chosen to establish the annual behavior of the temperature which is then curved as a function of time for each of the ten years. The features of these functions are used to evaluate statistically, the average behavior of the temperature over the ten years which represents the best prediction of the temperature for the following year that is the year 2010. In fact, the average of the annual mean temperature and the annual mean standard deviation over the ten years were used to predict the seasonal divisions and their durations as well. Finally, these quantities were used to predict the temperature behavior of the year 2010. The periods of higher and lower temperatures are well identified and the durations of the four seasons observed in a year are then deduced. This new method is simple and very important to predict the weather and is applicable to other regions of the world. This work also teaches us how to fix the time and observe the appearance or the evolution of natural events in a standard calendar.