Mann-Kendall test to determine the significance of the changes
After collecting and classifying the meteorological data, they examined temporal trends. As it unevenly ascending or descending meteorological data sets have been showing the trend in the data. Any natural or unnatural changes, leading to changes in the process are given. For the presence or absence of data to be specified in the statistical analysis used. One of the tests used to determine the significance of trends in data Mann-Kendall is. With this test and programming in Matlab significance of the changes in parameters were investigated. In a particular series of data to determine the randomness of the equation (1) is used (Kendall, 1973):
(1)
The T statistic of Kendal and P total scores greater than row after it is My¬Gyrnd. P from equation (2) is:
(2)
The statistics for N> 10 is similar to a normal distribution with zero mean and variance. The significance test of the relationship (3) is calculated:
(3)
Where N is the total number of statistical years, tg level of significance test that is equal to 96.1 and 95% more likely it is significant, and is the Mann-Kendall statistics. Here tg equal to the critical value of the standard normal distribution (Z) with a test based on the 95% probability level is equal to 96/1. If (T) t> T> - (T) t no significant trend is not observed in the series and random series, and if T (T) t is a positive trend in the series will prevail.