Fundamentals of Statistical Hydrology by Mauro Naghettini

Author:Mauro Naghettini
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

8.2.3 Plotting of Historic Events

Measuring rainfall depths is usually less demanding than gauging river flows since it does not require the previous definition of rating curves and in-river measurement of flow velocities. As a result, rainfall gauging networks are usually denser and have longer periods of records than flow gauging networks, on a worldwide scale. Bayliss and Reed (2001) report that, up until 1999, the mean length of annual flood peak records, available for regional flood studies in the UK, was just 23.4 years, which is probably much less than the rainfall homologous statistic. However, as large floods usually leave marks on river reaches, on river cross sections, and on human-made constructions, informal records on past events can be retrieved from several sources and then incorporated into flood frequency analysis.

Possible sources of informal (or non-systematic) records of past floods include peak levels marked on plaques, bridges, and buildings; church and local authorities archives; drawings and blueprints of past engineering works; and newspapers. These are termed historical floods and should be distinguished from the so-called paleofloods. Paleoflood reconstruction , by tracing and interpreting geological and botanical evidences that remained preserved long after the passage of floods, is another possible source of informal records of extreme floods that have occurred in the past. A new science, named Paleoflood Hydrology, has emerged from the development of new techniques and methods, over recent decades, to reconstruct ancient floods and incorporate them into frequency analysis. The reader is referred to House et al. (2002) and to Benito and Thorndycraft (2004) for a full account on the developments of Paleoflood Hydrology. In Chap. 11 an example of Bayesian flood frequency analysis, with the incorporation of paleofloods, illustrates such an interesting topic. This subsection focuses on how to assign empirical probabilities (or plotting positions) to historical floods.

Historic flood marks indicate the water levels that have risen above a fixed threshold during some historic period. For some, occurrences of floods greater than such a reference threshold are known but the magnitudes of flood peaks are usually unknown or poorly defined. Flood series that contain such a type of data are said to be censored and the threshold is the censoring value (NERC 1975, Bayliss and Reed 2001). Stedinger and Cohn (1986) refer to such a flood series as a binomially censored series since each flood should belong to one of the two possible states: either above or below the threshold. For some other historic floods, both their occurrences and magnitudes above the threshold are known or can be determined. In both cases, censored historic series should be analyzed alongside systematic series. Assigning empirical probabilities or plotting positions to such a combined series requires an attentive interpretation of flood records.

Following Hirsch (1987), assume that, in a period of N years, there are g known flood occurrences, of which, the k largest are singled out. Of these N years, s years correspond to the systematic period of records, such that s < N. Of the k largest floods in N years, e

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.