When rating-curves of the form Q = gamma(h + alpha)(beta) are fitted by lea
st-squares, goodness of fit as measured by the coefficient of determination
r(2) is Often close to 1, suggesting that estimated discharges have high p
recision. This can be illusory if (a) no account is taken of the uncertaint
y in estimate of the parameter alpha, and/or (b) the stage h at which disch
arge is to be estimated is such that log(e)(h + alpha) lies far from the me
an value of this variable calculated using the data points (h,Q) that defin
e the rating-curve. Furthermore, since the annual maximum discharges in any
M year period of record are all estimated from the fitted rating-curve, th
ey will be correlated, even if the annual maximum stages in the M years are
statistically independent. The usual maximum likelihood procedures for fit
ting extreme-value distributions do not take account of this correlation. E
xpressions are given for the conditional (on the values of the M annual max
imum stages) and unconditional variances of the mean annual flood (Q) over
bar which take account of rating-curve uncertainties. (C) 1999 Published by
Elsevier Science B.V. All rights reserved.