Topic II
What Can and Should Be the Role of New Observing Methods, Both In Situ (Including New Sensor Technologies) and Remote Sensing? How Might Approaches to the Estimation of Hydrologic Extremes Differ Based on the Richness of the Historic Observations?

PRESENTATION

Doug Alsdorf of The Ohio State University spoke on the topic of new observing methods for flood hydrology science, emphasizing his overall theme that floods are two-dimensional (2D), and one particularly useful remote sensing technique to look at floodwaters beneath inundated vegetation—interferometric synthetic aperture radar (InSAR). Flood waters move laterally across floodplains, wetlands, or urban environments and this movement is not bounded like that of typical channel flow. This two dimensional flow is obvious, but not well measured. Water flow within channels is measured by essentially one-dimensional methods. The water surface elevation is routinely gauged and combined with periodically collected velocity profile data to form a rating curve indicative of discharge. This approach, however, does not capture the complexity of floodwater flows because these are unbounded and have velocities that vary spatially as well as temporally.

According to Alsdorf, there are almost no two-dimensional height mappings of these flood water heights (e.g., essentially nothing like that of a topographic digital elevation map where elevations are mapped in a spatially continuous manner). Instead, the measurements often come from high water marks on the sides of buildings, bridges, or vegetation. High water marks fail to capture the temporal dynamics of rising and falling waters which are important for calibrating and validating two-dimensional hydrodynamic models. Occasionally, small devices are deployed before the arrival of a flood (e.g., level loggers, etc.) and are used to measure the temporal variations in flood water heights. These can be surveyed to provide a slope measurement between two or more devices. Nevertheless, these devices still represent a one-dimensional measurement, not a blanketing 2D measurement.

Hydrodynamic flood models that depend on water height measurements are only as accurate as the calibrating and validating data, Alsdorf stated. Because of the lack of 2D measurements, the models can often produce results that may be reasonable but incorrect. For example, models of Amazon floodplain flow do not match the measured height changes that occur during the passage of the annual flood wave. These models thus predict incorrect water routing and incorrect water heights. Models of smaller floodplains in the U.K. do match 2D mappings of inundated area but the degree to which they match flood water elevations is unknown.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 6
Topic II What Can and Should Be the Role of New Observing Methods, Both In Situ (Including New Sensor Technologies) and Remote Sensing? How Might Approaches to the Estimation of Hydrologic Extremes Differ Based on the Richness of the Historic Observations? PRESENTATION Doug Alsdorf of The Ohio State University spoke on the topic of new observing methods for flood hy- drology science, emphasizing his overall theme that floods are two-dimensional (2D), and one particularly use- ful remote sensing technique to look at floodwaters beneath inundated vegetation—interferometric synthetic aperture radar (InSAR). Flood waters move laterally across floodplains, wetlands, or urban environments and this movement is not bounded like that of typical channel flow. This two dimensional flow is obvious, but not well measured. Water flow within channels is measured by essentially one-dimensional methods. The water surface elevation is routinely gauged and combined with periodically collected velocity profile data to form a rating curve indicative of discharge. This approach, however, does not capture the complexity of floodwater flows because these are unbounded and have velocities that vary spatially as well as temporally. According to Alsdorf, there are almost no two-dimensional height mappings of these flood water heights (e.g., essentially nothing like that of a topographic digital elevation map where elevations are mapped in a spa- tially continuous manner). Instead, the measurements often come from high water marks on the sides of build- ings, bridges, or vegetation. High water marks fail to capture the temporal dynamics of rising and falling wa- ters which are important for calibrating and validating two-dimensional hydrodynamic models. Occasionally, small devices are deployed before the arrival of a flood (e.g., level loggers, etc.) and are used to measure the temporal variations in flood water heights. These can be surveyed to provide a slope measurement between two or more devices. Nevertheless, these devices still represent a one-dimensional measurement, not a blan- keting 2D measurement. Hydrodynamic flood models that depend on water height measurements are only as accurate as the calibrating and validating data, Alsdorf stated. Because of the lack of 2D measurements, the models can often produce results that may be reasonable but incorrect. For example, models of Amazon floodplain flow do not match the measured height changes that occur during the passage of the annual flood wave. These models thus predict incorrect water routing and incorrect water heights. Models of smaller flood- plains in the U.K. do match 2D mappings of inundated area but the degree to which they match flood wa- ter elevations is unknown. 6

OCR for page 6
Topic II 7 He summarized his presentation with five points. First, because we need more precise measurements of natural events on the Earth's surface, we should “get into space,” in other words he suggests increasing remote sensing techniques as one of the solutions. Second, water flow across floodplains is more complex than implied by 1D, point-based measurements. Third, flow paths and water sources are not fixed in space and time, but rather vary with flood water elevations. Fourth, hydrodynamic models do show promise for improving our understanding of floodplain hydraulics. And finally, to fulfill this promise, we need much more high-resolution topography and 2D mapping of water levels, their changes through time and space, and inundated area. PLENARY DISCUSSION The discussion session was led by Charles Vorosmarty. He asked participants to think in terms of packages of observations—i.e., the Decadal Survey approach—including InSAR and the proposed Sur- face Water Ocean Topography (SWOT) satellite mission, and then to think in terms of “What are the sci- ence questions that we would be able to answer?” Alsdorf emphasized that SWOT will not be “poin- table” at a given flood, but rather will catch many floods worldwide but at random. It is oriented more toward storage and discharge of large rivers on about a weekly basis. As for other kinds of questions we could answer better if we had improved 2D waterbody data, he mentioned two examples: • What are the methane and carbon dioxide fluxes from the hydrosphere to the atmosphere? These are partly a function of water depth. • How fast are Arctic lakes disappearing? Several other participants emphasized the importance of in-situ, ground-deployed sensors. These have several advantages over satellite-based sensors, in that they are cheaper, can be quickly deployed (48 hours before a major storm event, for example), have orders-of-magnitude better spatial resolution, and can be programmed to resolve phenomena at very short time intervals. Another participant emphasized the need to coordinate other measurements with the many Lidar airborne topography missions that are being flown. Merging data at disparate scales was viewed by some as a very difficult challenge in many cases, but the National Weather Service (NWS) has done this on an ad hoc basis, for example, using radar information on rainfall to constrain a stochastic model on storms. One participant said that the NWS has noted that climate related trends in hydrologic parameters model output are often smaller than the errors in the estimates themselves. The NWS has to prioritize its research based on its potential impact on applications, and in some cases (e.g., a project with only a 25- year life cycle) one might ask, “Is climate change in a 25-year timeframe so small that it doesn’t matter?” Another participant gave an example of the opposite, namely recent increases in summer precipitation in many parts of the highly populated coastal Eastern U.S. And in the case of the truly extreme events, as one participant noted, one has to either expand one’s observation in space with satellites—or in time with paleohydrology—to achieve both scientific understanding and predictive capability. Finally, a participant noted that we might feel more comfortable with our decision-making under un- certainty if we had a better scientific framework on which to hang not only climate-related hydrologic trends but those that may be due to urbanization, regulation, levees, and floodplain alteration.

OCR for page 6
8 Research and Applications Needs in Flood Hydrology Science BREAKOUT SESSION REPORT Rapporteur Geoff Bonnin (NOAA, National Weather Service) summarized the breakout session as having focused primarily on the need for more data—and better use of the data—in modeling and for de- cision support. He said that many participants in the session stated a strong need for many more observa- tions both spatially and temporally and of more physical elements and of complementary elements that allow confirmation by comparison between physical elements. The need exists in the context of the ne- cessity to establish these observations as part of a long-term and consistent observing system. The theme of new sensing technologies was secondary to the need for more observations. Additionally, observing resources are limited and so there is a need to determine how to specify observing systems in order to maximize the benefit of those resources. Bonnin noted that an underlying theme was a recognition that many types of data and many different periods of record exist in many locations and forms. Many of those present felt there is a need to develop analyses and techniques for analysis and assimilation that extract more information out of current and fu- ture observing records and collections of records. The group, he stated, discussed how traditionally we have observed physical elements such as rain- fall, streamflow, and water surface elevation. However, these are characteristics of phenomena that create impacts on society. Society is now asking questions about the impacts themselves such as: Will/did my house get flooded? How much of the house was submerged? They are also asking questions in areal, in addition to point, terms such as: How much of New Orleans was flooded? And how much of Georgia is without sufficient water to sustain industry? If scientists wish to help answer these questions, it would seem the best answer is to improve our recording of the impacts themselves, rather than just the character- istics of the phenomena. Bonnin said that many in the group believed that access to both the raw data and analyses of the data need to be improved. It is not enough, they said, to have collected the data, but it is essential to make in- vestments in data analysis to develop useful information products such as precipitation frequency analy- ses and flood frequency analyses. Real-time access to information for decision support and rapid access for analysis would also be important, and a participant noted that the Federal Emergency Management Agency has funds for post-event data collection in presidentially declared disasters. Finally, Bonnin repeated some closing comments that Robert Hirsch had made at the end of the ses- sion. Hirsch had summarized what he believed it would take to provide society with estimates of risk at any point as follows: • We need to understand the potential effect of climate change, but it is important to start by analyzing the information we already have. For example, USGS state-wide analyses of flood frequency are greater than 30 years old in some cases and NOAA precipitation frequency and duration analyses are similarly out of date. Given the interest in climate change as it affects flood risks, the first thing to attend to is to make sure that our analyses consider the most recent data. Then, look at the data and see if we can identify or describe changes in precipitation or flooding that may be related to changes in climate. • Consider the watershed. What changes have occurred (urbanization, conversion from forest to agri- culture)? What changes will occur? And what differences do the watershed changes make? • Consider the channel and floodway. What changes have occurred (bridges, culverts, buildings, changes in channel dimension or roughness, floodplain land use changes)? What about the role of land-surface subsi- dence (changing slopes)? • Consider flood control structures. We should know what and where they are, what their current char- acteristics are and how they actually operate in addition to how they were and are supposed to operate. What

OCR for page 6
Topic II 9 are their modes of failure and how do those failure modes affect the watershed and its people and environ- ment? What do we actually know about how flood risk changes after these structures are put in place—risk from smaller floods versus risk from great floods.