How to define tails of a graph using Generalised Pareto Distribution

I am undertaking a study on extreme sea levels. I have extracted high tide residuals which I have binned into sea levels and their respective frequencies. This resembles a normal distribution for which I have outputted a simple bar chart in excel (-1.2 metres to +1.2 metres in 5 cm intervals). The problem is that the tails are poorly defined and I need to extrapolate that data up to 1000 years.

I am completely clueless as to how I go about this as I have never used matlab previously. Any help would be greatly appreciated.

Thanks

[phanta]

Hi Sean,

whats the goal of your distribution fit? You just want to plot the data as generalized pareto distribution (and say that it follows this type of distribution) or you want to use the information for further analyses?