In characterisation of extreme events connected with landsliding (almost always connected with intensive rainfalls) beside magnitude of the sliding ussualy losses of lives and goods is taken into consideration.For example, if some large landslide occur but if it doesn't influenced infrastructure or human lives, this is only high level natural hazard, but if landslide with "same" characteristics influenced and damaged and infratsructure and human lives, it produce also a lot of costs and losts that is a landslide with high risk and extreme event. In general, such events are not so frequefent but intensity, magnitude and effects on envoronmemt, infrastructures ald lives is high.

Thank you very much Prof. Kenneth and Prof. Milorad. Yes Im referring meteorological data like wet day frequency.  Here Im attaching the snap shot plot of wet day frequency for 100 years. I would like to understand the mechanism behind this extreme events. There are few works related to extreme events in dynamical systems (check the link).

I believe that the wet day frequency/ precipitation and all other meteorological phenomena should obey a deterministic or stochastic process. If we could understand this process, it would help us to understand the dynamics of extreme event. Also this would help us  understand the climate as well. 

It would be very helpful for me if you give your suggestions regarding this. 

http://arxiv.org/pdf/1311.5776.pdf

The data is taken from meteorological department of Tamilnadu (it is a state in India) from the year 1901 to 2000 AD. We can find the extreme events in very few places  around where the plot is done. IF we go away say 100km from this place there is no extreme events.

Also we can find a spatio-temporal relation in the wet day frequency/precipitation in the regions I have mentioned. 

Hi Suresh. You ak: 1) if the data have extreme events in past 100 years, how could we characterize this data? and 2) Also is there any connection between precipitation and wet day frequency?

About 1) it is common that such data contains a maximum positive value and a minimum positive one; they are historical-regional data of the two extreme values of days with or without rain. You define the concrete context of your data.

About 2) It is possible to define a dry day as one with zero rain and to measure its relative frequence from your data.There may be some connection between the relative frequence of rain days and those of dry days relative frequence.

You have two options: a) analize only rainy days with their two extreme values, and b) analyze all days (rainy and dry days) with some maximum extreme value and a minimum extreme value equal to zero. There are methods for both options.

Best wishes.

Doctor Towe, thanks for all your illustrating examples and diverse data cases.

What is the most frequent method that you apply to analyze in those cases to describe their univariate distributions and cumulative distribution probabilities?

Thanks, Emilio

Thanks Kenneth for your elegant answer to my question. I learned about some aspects of this data handling.

Go through Gowsamy Science Paper on extreme precipitation analysis

Dear Prof. Towe,

Please go through the attached figures. The figures.pdf shows the wet day frequency of the four places (spot 1 to spot 4) in Tamilnadu district for 1200 months. You can see marked spots in the attachment. 

While going through spot 1 to spot 4, one can see a quantitative changes in the dynamics of the wet day frequency. 

In spot 1 the average value of the wet day frequency is approximately  5 and in spot 4 the average value increased to 10. The value of the wet day frequency is increases through extreme event through spot 2 and spot 3.