No abstract

Hydroclimatological time series often exhibit trends. While trend magnitude can be determined with little ambiguity, the corresponding statistical significance, sometimes cited to bolster scientific and political argument, is less certain because significance depends critically on the null hypothesis which in turn reflects subjective notions about what one expects to see. We consider statistical trend tests of hydroclimatological data in the presence of long‐term persistence (LTP). Monte Carlo experiments employing FARIMA models indicate that trend tests which fail to consider LTP greatly overstate the statistical significance of observed trends when LTP is present. A new test is presented that avoids this problem. From a practical standpoint, however, it may be preferable to acknowledge that the concept of statistical significance is meaningless when discussing poorly understood systems.

An investigation is made of flood quantile estimators which can employ “historical” and paleoflood information in flood frequency analyses. Two categories of historical information are considered: “censored” data, where the magnitudes of historical flood peaks are known; and “binomial” data, where only threshold exceedance information is available. A Monte Carlo study employing the two‐parameter lognormal distribution shows that maximum likelihood estimators (MLEs) can extract the equivalent of an additional 10–30 years of gage record from a 50‐year period of historical observation. The MLE routines are shown to be substantially better than an adjusted‐moment estimator similar to the one recommended in Bulletin 17B of the United States Water Resources Council Hydrology Committee (1982). The MLE methods performed well even when floods were drawn from other than the assumed lognormal distribution.

We consider the appropriateness of "rating curves" and other log linear models to estimate the fluvial transport of nutrients. Split-sample studies using data from tributaries to the Chesapeake Bay reveal that a minimum variance unbiased estimator (MVUE), based on a simple log linear model, provides satisfactory load estimates, even in some cases where the model exhibited significant lack of fit. For total nitrogen (TN) the average difference between the MVUE estimates and the observed loads ranges from -8% to +2% at the four sites. The corresponding range for total phosphorus (TP) is -6% to +5%. None of these differences is statistically significant. The observed variability of the MVUE load estimates for TN and TP, which ranges from 7% to 25% depending on the case, is accurately predicted by statistical theory.

This paper extends the work of Gilliom and Helsel (1986) on procedures for estimating descriptive statistics of water quality data that contain “less than” observations. Previously, procedures were evaluated when only one detection limit was present. Here we investigate the performance of estimators for data that have multiple detection limits. Probability plotting and maximum likelihood methods perform substantially better than simple substitution procedures now commonly in use. Therefore simple substitution procedures (e.g., substitution of the detection limit) should be avoided. Probability plotting methods are more robust than maximum likelihood methods to misspecification of the parent distribution and their use should be encouraged in the typical situation where the parent distribution is unknown. When utilized correctly, less than values frequently contain nearly as much information for estimating population moments and quantiles as would the same observations had the detection limit been below them.