Relational characteristics of some natural phenomenon like solar radiation, rainfall and temperaturehave been realized by statistical analysis. The degrees of relationship existing between them havebeen computed to find out the strength of the dependency. For statistical analysis, solar radiation isconsidered as dependent variable on rainfall and temperature variation. Regression analysis hasbeen done for estimating the dependent variable solar radiation from the independent variablesrainfall and temperature. Due to the fact that radiation varies partially because of rainfall andpartially because of temperature variation, the partial regression coefficients have also beencomputed. A regression plane has represented these, in a 3-D rectangular co-ordinate system. Leastsquare regression planes fitting a set of data, points in a 3-D scattered diagram have been obtained.From these the coefficient of linear multiple correlation is calculated. It has been found that thecorrelation coefficient between rainfall and solar radiation is about 0.000737 and that betweenradiation and temperature is about 0.00281 (annually). Coefficient of multiple correlation is foundto be 0.00289. Standard error of estimate has also been computed. From this analysis, it isconcluded that a non-linear relationship may exist between solar radiation, rainfall and temperaturein this Ganges delta region. All the calculations were done with the data recorded in the visibleradiation from 9 a.m. To 6 p.m. And the average daily solar insolation (ADSI) has been estimatedfrom the information of the ADSI hours for Bangladesh which about 6KWh/day.
展开▼
