Estimation of extreme precipitation quantiles in non stationary series

Preliminary analyses regarding how the presence of trend in precipitation series affects the sampling properties of estimated quantiles are illustrated. To this end, sampling properties of precipitation quantiles, namely bias and Mean Squared Error (MSE) are investigated with respect to the size of the estimation sample, assuming a trend in the parameters of the underlying distribution. In particular, analytical results are derived for the cases of exponential distribution, while more complex cases (e.g. Gumbel distribution) are investigated numerically by simulation. Also the effect of preliminary trend removal is investigated and compared to the case when trend is neglected.