Estimating N-Year Environmental Conditions: Correcting for Statistical Uncertainty

Abstract In engineering design, it is common to estimate N-year values of environmental variables. When based on small datasets, such estimates often show unconservative bias due to statistical uncertainty in distribution parameters. We study here the bias associated with various distributions (Gumbel, Weibull, GEV, GPD). Simple analytical bias corrections, suitable for spreadsheet calculations, are established for these various distribution models.