OHD/HL/HSMB - Hydraulics: Mapping

Flood Forecast Mapping

8/1/2008

The NWS recently began providing static inundation maps at selected river forecast points through the Advanced Hydrologic Prediction Services (AHPS) web pages. The map libraries served to the public through AHPS provide information about the approximate extent of flooding associated with river stage forecasts at pre-defined stage increments. Emergency managers and other users have provided positive feedback on these maps. Therefore, efforts are underway to partner with additional communities and expand NWS forecast inundation mapping services.

However, in order to expand this program efficiently and effectively, we need to (1) identify the most suitable sites for producing static maps and sites where dynamic mapping is required, (2) enhance our operational, unsteady hydraulic modeling capabilities for the dynamic mapping sites, (3) identify efficient technological mapping solutions that are compatible with NWS operational systems, (4) develop improved techniques to verify maps, and (5) better techniques to compute and convey uncertainty.

Static vs. Dynamic

The assumptions that go into developing static inundation maps may be too conservative of an approximation at some forecast locations; therefore, dynamic mapping approaches must be considered. Static maps will only be accurate if the flood wave evolution at a point of interest does not vary much among events with similar peak flows. Dynamic mapping is required when flood maps vary substantially among events with similar peak flows. Many situations can cause this effect. Some examples include:

• Mildly sloping channels that have looped rating curves.
• Markedly different boundary conditions from event to event, e.g.
• Backwater effects on local tributaries that depend on flow states in the main river
• Time-varying tidal conditions.
• Time-varying reservoir levels or gate releases.
• Dam or levee break situations

Science and Technology Challenges

Although a number of technological solutions for inundation mapping are available, a recent review of four NWS dynamic mapping pilot projects by the Real-Time (Dynamic) Inundation Mapping Evaluation Team highlights difficulties in selecting the most appropriate technology for dynamic forecast mapping. Among key issues identified were the (1) incompatibilities between the architecture directions of industry leading GIS software and AWIPS (i.e. ESRI’s emphasis on Windows for desktop software and dropping of support for desktop applications on Unix and the AWIPS move to Linux), and (2) the difficulties of the available pilot hardware and software in meeting the high computational requirements for dynamic mapping. Unfortunately, none of the GIS software technologies used in any of the four pilots is compatible with AWIPSII.

The computational and software platform requirements for operational dynamic mapping must be strongly considered as we move forward. Also, there are key science and implementation questions that will impact the overall computational requirements such as:

1. Can we produce reliable probabilistic forecast maps?
2. How often do forecast maps need to be updated?
3. What are the requirements for the accuracy and resolution of the topographic data?
4. When and where will 2-D models provide significantly improved results over 1-D modeling techniques?
5. How can we verify the accuracy of competing techniques?

Plans

The hydraulics group is now beginning to scope new projects to improve our understanding of the issues described above and help make decisions about how to cost effectively expand and improve inundation mapping services.