Flood Estimates: How Good Are They?

The few tests available suggest that procedures and assumptions in common use for estimating flood peak frequency for ungaged streams are subject to large errors and are biased toward overestimates. Economic analysis of costs and benefits is ignored by many agencies in favor of designing to a arbitrarily selected probability level, commonly 1%. In view of the large expenditures for storm drainage and flood damage mitigation, one may question the wisdom of this approach. This paper discusses the causes of the current situation and suggests some steps which might help to correct it.