Research and Applications Needs in Flood Hydrology Science: A Summary of the October 15, 2008 Workshop of the Planning Committee on Hydrologic Science (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Topic IV How Can Advances in Techniques for the Accurate Analysis of Ancient Flood Events Aid Estimation of Future Flood Magnitudes and Frequency, and Understanding of the Generative Processes for Extreme Flood Phenomena? PRESENTATION Lisa Ely of Central Washington University then spoke on applications of paleoflood hydrology, types of paleoflood records, and recent advances in paleoflood hydrology. She began by defining “paleofloods” as past floods that were neither recorded by direct hydrological measurement nor in the written historical record. She noted that this definition could include modern floods in ungaged basins that were later inter- preted, using paleoflood techniques, after the fact. Her take-home message was that evidence of past floods is present in many if not most watersheds. Paleoflood records may not be the ultimate answer but why not use it? Ely then gave a summary of major applications of paleoflood hydrology. These include flood magni- tude and frequency to extend the record over long time scales (i.e., 100s to 1,000s of years); flood station- arity; evaluating design flood values for engineering design, flood risk assessment and river management; response of extreme floods to climatic changes; geomorphic impacts of extreme floods (i.e., what role do they play in the changing geometry of the river), which has applications for wetlands management; and communicating flood hazards to public. She followed this with an outline of types of paleoflood evidence that we look for in the geologic re- cord, and are, in fact, derived from physical effects on natural indicators. She began with three indicators of estimated peak stage: • Slackwater flood deposits, which are fine sand and silt deposited in areas of low velocity along stable, confined channels; • Erosional scour lines or silt lines; and • Flood scars on trees. She then summarized three other, less direct techniques, namely: • Alluvial floodplain stratigraphy; • Changes in the channel geometry over time; and 13 

14 Research and Applications Needs in Flood Hydrology Science • Changes in the stream’s competence (i.e., the maximum particle size that a stream can transport), as inferred from the presence of boulders in the stream sediments. She then gave one or two examples of each. Ely proceeded to describe some of the recent advances in paleoflood hydrology. The first of these is improved resolution of standard types of dating methods and development of new ones. Second, there have been many advances in hydraulic modeling of paleodischarge, from the models themselves to topog- raphic databases and rendering. A third is probability-based flood-frequency methods that incorporate isolated peak flow events. A fourth is the development of applications in a greater variety of geomorphic settings and geographic locations. A final advance has been the development of global paleoflood data- sets for recognizing climatic patterns. But there is no “standardized treatment” for every watershed, Ely cautioned. One has to be careful, for example, with channels that are not stable over time. More needs to be done to understand the relation between paleostage indicators and peak water-surface elevations. Finally, she noted that non-meteorological floods also occur—referring to those caused by landslides, catastrophic failure of moraines or landslide debris, etc.—that need to be considered. Ely gave four examples of the use of paleoflood data, and then concluded by suggesting some poten- tial future directions for research and applications. These included implications of future climate changes on floods; regional comparisons of global paleoflood records; targeting climatically sensitive or hazard- prone areas; questions related to stationarity, including non-meteorological floods, flood frequency and risk assessment using updated techniques; new methodologies in geochronology and geomorphology; and techniques for communicating flood hazards to the public. PLENARY DISCUSSION Victor Baker of the University of Arizona led the subsequent discussion, which began with the ques- tion “How can one incorporate urbanization in paleoflood studies?” Ely noted that in some watersheds one could compare pre-urbanization floods with post-urbanization floods, incorporating paleoflood data, and Baker stated that for extreme floods, the urban effect is usually overridden anyway. Difficulties in finding paleoflood data in more populated regions such as the eastern U.S. were noted, and Ely stated that there has been work on floodplain stratigraphy that examines grain size changes that may not even be de- tectable visibly. Baker noted that caves can sometimes preserve paleoflood deposits in the Appalachians. Several participants mentioned that countries such as China, India, and Egypt have long written re- cords of floods—up to 5,000 years in some cases. While the quality of the description varies, the dates tend to be very accurate. Ely said this can be used to put the modern record into context, and that we should be careful not to reject paleoinformation solely because it is not consistent with the present system. Finally, there was discussion of the potential value of a major survey of paleofloods around the coun- try. Given the scarcity of paleoflood data in a single basin, should scientists be doing more “trading off of space for time” to get a big picture of prehistoric behavior? BREAKOUT SESSION REPORT According to session rapporteur James O’Connor, the discussion ranged widely but could be distilled to two main issues—the primary benefits of paleoflood studies, and limits to their applicability. 

Topic IV 15 The primary benefits of paleoflood analyses are often seen as an extension of the observed record of extreme events with information of actual prehistoric occurrences of large floods. Such observations, in addition to leading to improved magnitude-frequency relationships for specific locations (especially im- portant for critical structures), can provide physical evidence supporting understanding of watershed flood-generation mechanisms and perhaps improve the poor spatial resolution of probable maximum flood estimates. Paleoflood studies also offer opportunities to investigate relations between climate condi- tions and flood generation—a factor affecting stationarity assumptions in flood frequency analysis. O’Connor said that two main issues requiring research, and which may be factors limiting the appli- cability of paleoflood information, were raised at the breakout session. These were (1) investigation of the discrepancy between maximum floods observed from paleoflood analyses and those predicted from probable maximum precipitation studies (a topic also mentioned in the breakout session following Topic I). It was noted that for most rivers analyzed so far the maximum floods determined by paleohydrologic techniques (from physical evidence of actual floods) are much smaller than the Probable Maximum Floods (PMF) determined by probable maximum precipitation studies. Does this discrepancy owe to shortcomings in the methods of paleohydrology and/or PMP analyses? Or does it reflect some sort of bias in where paleoflood studies have been conducted? And (2) developing a better understanding of the uncertainties involved with paleoflood techniques. The group also discussed the viability of some sort of a national paleoflood research program aimed at extending flood records for many rivers. Many of the group, however, concluded that some systematic targeted studies aimed at some of the issues noted above would be the most efficient approach to making progress at the two questions noted above.