Hello
I am trying to evaluate the impact of the climate variables on wilfire dynamics using Man-Kendall test, forecast future values and interpret them.
I saw in litterature several research but I don't know how to perform it properly using my dataset in R software and interpret the ouputs.
Dear Jean
The following packages in R can test the non-stationarity of data:
"Kendal" and "trend".
Also please check this post:
https://www.researchgate.net/post/How_can_I_use_R_software_to_perform_the_Mann-Kendall_Test_for_trend_analysis_and_abrupt_change
Let me know if you have any question.
Best,
Dear sir
I think you should provide the number of days in which their mean air temperature is above the pre-specified threshold (say 30 degree Celsius). So we have a series of n observations (i.e. n years). On the other hand we can gather the number of wild fires in a year. Thus the second series had n data equal to the number of years. So, you can run a correlation between these two series and test the coefficient of regression for its significant. If it was significant this implies that there is a direct linear relationship between the number of days with air temperature great than 30 degrees Celsius. This approach can be used for testing the hypothesis that there is a linear relationship between number of days with rain in 3 summer months of a year and the number of fire occurrences in a year. In this regard as the number of summer rains increases we anticipate that the number of fire occurrences decreases.
I hope this can solve your problem to some degree.
best wishes;
Dr Yagob
