Advisory Committee on Water Information
| Subcommittees |
Subcommittee on Hydrology -- Minutes -
|
|||||||||||||||||||||||
|
National Weather Service Office Room 14316, 1325 East West Highway, Silver Spring, MD AGENDA
Immediately following the meeting, Advanced Hydrologic Prediction Service (AHPS) Update Presentation on Community Hydrology Prediction System SUMMARY OF MEETINGPARTICIPATING Don Woodward, American Forests MEETING HIGHLIGHTSDon Frevert called the meeting to order at 9:35 30 a.m.
Changes to Terms of Reference by Don Woodward as proposed in
the October 2, 2003 meeting minutes were approved by SOH.
The e-mail of January 20th had as an attachment a paper that was jointly prepared by Martin and Will Thomas (see Appendix A.2). It outlined areas of their agreement and disagreement on debris flows and how they should be classified in the USGS database. Their primary difference was Martin’s reliance on FEMA’s regulatory definition of a flood (44CFR 59.1) that includes debris flows as floods and Will’s reliance on a report by the National Research Council in 1982 that does so for purposes of the National Flood Insurance Program (NFIP). At the meeting, Martin provided an excerpt from a 1985 memorandum from the Federal Insurance Administration (FIA) stating that the regulatory definition was the most appropriate definition of the two. Will acknowledged the opinion of the 1985 position paper (see Appendix A.4). Martin expressed the opinion that FEMA has established a definition of a flood that is universal because it reflects the position of Congress and a formal rule making process by FEMA. Although that definition includes debris flows as floods, it is not being used in USGS’ databases as such. He reiterated that the issue is a conflict between real life, and a science fair approach at the USGS where debris flows are not classified as floods. He drew the analogy of the USGS’ failure to classify debris floods as floods in its database to the accounting errors at World Com that cost the public approximately five billion dollars. It was his opinion that if the science fair version of debris flows rather than the regulatory version of debris flows continues to be used by the USGS in its database, future damages to life and property could occur at levels similar to the amount of the World Com damages. The reason is because without flow of record data, the proper hazard protection will not be adopted. After comments by some SOH members defending the USGS procedure, Martin asked the group whether they would be willing to have their families live in a house in a debris flow prone area not properly mapped because the USGS had not recorded a debris flow as a flood. Martin was concerned about the true risk of flooding and feared that data collection and analysis procedures that in some instances threw out the flood flow estimates associated with the largest floods of records because they were debris flows would result in a significant misrepresentation of the risk of flooding. He indicated that the recent Waterman Canyon flood with its associated loss of life illustrated the real life problems experienced in San Bernardino County, California. Further, if the USGS had such problems determining what the correct flood flow value was in those instances where the USGS has removed peak flow data from the online databank for the public, how could they decide that the value they had withdrawn from the database was wrong? Note: Federal Insurance Administration (FIA) Review of the Mudslide, I. E., Mudflow Coverage of the National Flood Insurance Program After the discussion by various members of the SOH, Martin Becker voiced his position that the classification of debris flows in accord with the regulatory definition and the concurrent recording by the USGS in its database was a life or death issue. For that reason, he moved that he, along with Steve Blanchard of USGS and Geoff Bonnin of NWS prepare a fact sheet reflecting the comments from the meeting on the subject of the proper classification debris flows that would be presented to the Office of Information and Regulatory Affairs of OMB with an invitation for it to attend the next SOH meeting to discuss the issue. This motion died for lack of a second. 1. Federal Insurance Administration (FIA) Review of the Mudslide, I. E., Mudflow Coverage of the National Flood Insurance Program In the discussion individuals made their comments below. Will Thomas: The FEMA definition of "Flood" includes mudslides (mud flows) because of the desire to provide flood insurance coverage to those who suffer damages from these hazards. The 1982 NRC report and USGS Open-File Report 85-276 indicate that mudslides (mud flows) are a type of landslide, not a flood. However, both reports indicate that mud floods (debris flows) are a type of flood. FEMA's definition of "Flood" should be used for the National Flood Insurance Program (NFIP) but should not be the controlling definition of flood within the Federal government because this definition includes mudslides, a type of landslide, and is not consistent with other government or organizational publications. There is no need for a universal definition of "Flood".USGS should provide a footnote in their peak flow file (web site) that identifies peak flows affected by significant debris and have a link to another file that describes all data (indirect measurement, pictures, cross sectional data, profiles, etc.) related to the debris-affected peak flow. Users of the data can utilize the debris-affected flood as they consider appropriate. Consistent with statements by Robert Mason, USGS, in the Subcommittee on Hydrology August 2003 minutes, USGS should continue research related to estimation and use of debris-affected peak flows. Steve Blanchard: I appreciate Martin raising this question on the definition of a flood. Hazards are an important issue. Martin raising the issue related to the definition of floods and debris flows will result in constructive dialogue that will hopefully resolve some confusion and misconceptions about the issue. The role of the USGS is to correctly characterize hazard events. That is, to use the best technical and scientific methods we have available to accurately describe these events and what happened. In relation to Martin's request to the SOH that a universal
definition for a "flood" be adopted by Federal agencies,
and that the definition should be the definition specified in
the US Code (USC) of Federal Regulations for the National Flood
Insurance Program, the USGS disagrees with Martin's The USGS believes it is wrong to use an equation such as Manning's equation, which is for Newtonian fluids (water), to estimate a discharge for a debris flow, which is a non-Newtonian flow event. What the USGS has done for debris flows, is to include the date and stage (flow height) of the event in our peak flow file, but we have not included a discharge. We are exploring options for making the debris flow information available through another database. A description of and the data for debris flows events continues to be available from the USGS office which worked the event. In regards to Martin's motion to have a small group work on a white paper on this issue and to include OMB, the USGS declined the motion because of the vagueness of the motion, especially related to the white paper - it was unclear what the purpose of the white paper was, what issues would be addressed by the white paper, etc. The USGS declined the motion but did agree to work with Martin and others to try and accurately document the content of the SOH meeting discussion on this subject. Don Woodward: Raised the question of whether debris flows could be flagged or annotated in a flood flow data base. John Costa: I support Will's position that the FEMA definition of floods
is far more inclusive and broad than that of any other agency
or organization, including ASCE, AGI (American Geological Institute),
TRB (Transportation Research Board), and USGS, among others.
None of these recognize mud or USGS policy is never to destroy or remove data on flow events. In fact, recent efforts have increased the amount of information available on debris flows at gaging stations. You just can't get a discharge value from the web page, but there is at least one other indication that a large event occurred - the stage record. If anyone wants additional information, they have to ask the local office, who will be glad to provide information, even inappropriate indirect measurements that may have been done. USGS can't control who uses our data, or for what purpose. The evidence is present in our databases for large flow events that may not be floods at gaging stations. It is the responsibility of the user to be sufficiently familiar with the hydrology of an area to know how to recognize those indications. It is not difficult to do. Our primary statutory responsibility is to provide the most scientifically correct information possible to the public. Doug James: Observed that there is a need to treat both flood and landslide/debris flow data – even though mixing the two may not be technically correct. Don Frevert: Inquired as to whether anyone monitors landslides and debris flows on a national or local basis. (The prompt response was that USGS has a national landslide mapping program) David Wingerd: • Not all peak flow should be used in hydrologic runoff studies. An example of a peak flow on the Snake River in Idaho. This flow was the result of the failure of Teton Dam. • How various agencies’ definition of a flood is influenced by how they use the data. For example NRC defines a mudslide as occurring only in a river. Commonly mudslides can occur anyplace, but NRC is only concerned with mudslides in a river. The Corps’ Geotech folks took a look of the agency definitions and found that none were universal. They have different definitions for Mudflow, Mudslides, and Debris flows. Kevin Long: Noted that FEMA’s definition of a flood was driven by insurance considerations. Gene Stallings: At the most recent meeting, the debate on debris flows or debris floods was lively with some strong differences of opinion of what is a flood. The National Hydrologic Warning Council members have much interest in the subject. Just prior to our meeting, I received an e-mail from one of our members in California, raising some of the same issues as Martin Becker. (Since I attended the Meeting and prepared my Executive Summary for the NHWC members, I have received additional comments—this is an aside). My comments were that ice jams and dam breaks are also floods. This is my first comment. A flood is a flood. The damage can be the same. Second, I heard several comments during the one hour discussion period, there were new to me. I made the recommendation that this valuable information should not be lost by requesting to document their comments. Whether we need a Work Group on the subject is not known. But if we do form a group, Terms of Reference should be defined and understood by all. Based on the interest of the NHWC in this subject, we may be able to take an active role. Jery Stedinger: The public is exposed to a real risk from floods with a range of characteristics and there are two issues. One is the collection of accurate data. The USGS is appropriately concerned if records of floods that included significant debris are accurately described. It appears that USGS may not want to include estimated flows for such events in their records because they cannot accurately compute a flow from the stage records. The other issue is how to compute risk of flooding for residents and occupants of the floodplain. For the second task it is imperative for the analysis to recognize the risk of floods whatever their characteristics. My concern is that the risk of flooding may not be accurately evaluated because the needed data is not collected or reported, perhaps because such an effort was not viewed as being part of the mission of the USGS or some other agency. I believe the hydrology subgroup should not believe the public safety is served by just including stage information in USGS records with a footnote describing data problems, along with the assumption that the public and consultants will then do the appropriate risk computation. This subgroup, representing the expertise from the relevant federal agencies, should take responsibility to ensure that appropriate methods and analysis procedures are available and employed for computing the actual risks of flooding. Christopher M. Knopp: I questioned the validity of establishing flood plain maps given that the debris flows discussed resulted from the consequences of high intensity rain on burned over slopes. Without the fire, there was negligible risk. While the frequency of fire can be determined, fire frequency is not a statistic figure that lends itself to a meaningful return interval determination. I would add to my earlier discussion by saying that debris flows of this type should have their own delineation, separate from flood plain mapping. Lumping this sort of information with flood mapping would dilute the importance of both. FERC: Potential Failure Modes (PFMs): Each possible failure mode (FM) that could threaten a dam’s structural integrity and ultimately cause loss of a reservoir should be developed and listed with adverse and positive factors. Those factors could affect the potential for each FM to happen as the triggering mechanism of a dam failure. Weighing associated adverse and positive factors, each PFM can be classified in the following categories:
For example, we consider the following general PFMs of a concrete dam’s flood-induced failure: • Overtopping flows causing dam failure due to inadequate
spillway capacity (e.g., resulting from insufficient available
capacity and operational and/or procedures difficulties such as
inoperable or malfunctioned gate system because of losing power,
mechanical trouble or floating debris strict; improper reservoir
operation, etc.) Merits of PFMA Process: Conventional dam inspections apply standards-based criteria to identify a dam’s “traditional” PFMs (i.e., deficiencies). However, PFMA is a site specific approach to integrate those criteria with risk-based judgment in order to identify and classify “traditional” and “non-traditional” PFMs. Thus PFMA provides enhanced understanding and insight of the risks associated with a dam. The PFMA-oriented dam inspection has the following advantages:
As described above, PFMA is systematically conducted for a very focused and directed dam inspection process. As a result of practice, PFMA has been acknowledged as an efficient and effective tool functioning as a decision-support system for dam safety management. PFMA Guidelines Website: The PFMA guidance is posted on FERC’s website, titled “Chapter 14 Dam Safety Performance Monitoring Program (DSPMP) - Potential Failure Modes Analysis.” The website is listed below http://www.ferc.gov/industries/hydropower/safety/dspmp.asp
Doug James had no new developments to report. USDA Forest Service: Chris Knopp reported that the Forest Service will hold an internal conference on Watershed Management Issues in October, 2005. NWS: Tom Donaldson reported that the NWS Advanced Hydrologic Prediction Service (AHPS) was fully funded, $6 million, for FY 2004. However, one major program for which NWS has signed agreements was not funded so that the funding for AHPS had to make up the difference. That difference was $1.3 million. The impact that this had on the AHPS program this year was to reduce the number of forecast locations planned for conversion to AHPS. The number now stands at 419 additional locations. The good news is that AHPS was fully funded, the bad news is that the program took a hit from internal. CORPS OF ENGINEERS – HEC CENTER: Jeff Harris reported that Version 1.0 of HEC-ResSim has been released. HEC-ResSim will replaces HEC-5, our DOS based reservoir routing software. Efforts are ongoing to add water quality and sediment transport capabilities to HEC-RAS. The HEC-HMS software is undergoing a major rewrite using the JAVA language. Several enhancements are being added including detailed continuous soil moisture accounting, rain-on-snow runoff capabilities, detailed areal rainfall reduction techniques and an improved graphical interface. We are also undertaking an effort to upgrade our statistical frequency analysis software in one Windows based package. Currently, we have 4 pieces of software which are DOS based. The most visible of these is the HEC-FFA software which performs frequency analysis based on Bulletin 17B criteria. ResSim 2.0, RAS 3.1.1 and HMS 2.2.2 are available from the HEC website.
Steve Blanchard reported as follows:
The second is a Fact Sheet "Effects of Urban Development
on Floods" by Christopher Konard, USGS Fact Sheet FS-076-03.
This Fact Sheet is available on-line electronically at URL: FEMA: Kevin Long reported that FEMA had received funding for their map modification efforts.
Will Thomas reported that the Stormwater Management Committee
of the ASFPM provided comments to the USGS on December 16, 2003
on their proposal entitled "Improving Information Needed
for Flood Mapping in the United States". The Stormwater Management
Committee of ASFPM is supportative of the proposal and recommended
that the proposal be Will Thomas also reported that the ASFPM annual floodplain conference will be held in Biloxi, MS during the week of May 16-21, 2004. There will be about 15 workshops given either before or after the conference. Will and selected colleagues from Michael Baker will present a 4-hour workshop on Hydrology and Hydraulics for Floodplain Managers. NATIONAL HYDROLOGIC WARNING COUNCIL Eugene Stallings reported that he has been active on the Hydrology Subcommittee (or Committee depending on the time) off and on since the early 1970’s. Historically, he can not remember such a large turnout at the meetings on a consistent basis. Apparently, the Subcommittee is working on topics of considerable interest to the hydrology community. The future looks very bright for the Subcommittee. Keep up the good work. AMERICAN FORESTS: FHWA: CORPS OF ENGINEERS: Dave Wingerd reported the inconsistent river flow data issue as follows: The Corps, USGS, NWS and others received the stage data electronically via telemetry often from the same gage. A rating curve at the receiving office translates stage data into flow data. When the river changes the rating curves need to be revised. The USGS monitors the gaging stations and, when necessary modifies the curves. If the new rating curve is not rapidly deployed to NWS or the Corps, the computed flow values on the three agencies' web page could be different. If there is a significant change, the change in the rating curve could effect how the Corps regulates it projects. Also this is becoming a serious PR problem when automated data is displayed on the Internet. Theoretically three different agencies could report three different flows for the same point. It has been a difficult, unsolved issue that local field offices in different agencies were trying to solve in different ways. Actions toward the solution are underway. The USGS assigned a
focal point and USGS, the Corp and NWS each put together a team
of end users from various field offices that understand and use
the rating curves. Also being addressed is the issue of extending the rating curves at its upper and lower ends surfaced. BUREAU OF RECLAMATION: Don Frevert reported that the Bureau of Reclamation held its River Systems Management Workshop November 4-6, 2003 in Fort Collins Colorado. The workshop was attended by about 100 managers and technical specialists from Reclamation, other federal and state agencies, universities and private industry. The next workshop is scheduled for March, 2006. Although Reclamation's overall budget and its funding for research have remained relatively flat over the past two years, Reclamation's funding for the interagency Watershed and River Systems Management Program has been cut severely and progress on the Reclamation portion of the program has been seriously impeded. Several SOH member agencies are involved in the ASCE/EWRI Watershed Management Conference which will be held in Williamsburg, VA July 19-22, 2005. This conference is held every five years. A call for papers was distributed at the meeting and is also being distributed in electronic form. Proposals for sessions as well as abstracts for individual papers are being solicited. BLM: Eric Janes reported that Ms. Heidi Hadley has reported as the BLM Salinity Control Coordinator to the Upper Colorado Region Office of the U.S. Bureau of Reclamation at Salt Lake City. Heidi will provide the lead coordination for the BLM’s role in management of saline sediment and salt uptake/ transport on public lands for the seven Basin States. Heidi will report to Ed Shepard, Assistant Director for Renewable Resources and Planning, BLM Headquarters. Eric also mentioned that the BLM is actively participating with other federal departments (e.g. USDA-Forest Service) in a fast track Web-Based Field Implementation Guide for the President’s Healthy Forests Initiative and the Healthy Forests Restoration Act of 2003. The principal area of interest to the Subcommittee in this endeavor will be the watershed-based risk assessment for municipal water supply watersheds in areas with dangerous levels of wildland fuel accumulation that need hazardous fuels treatment. Much more information will be available on this effort for the April, 2004 meeting. There will be a change in the primary representation to the Subcommittee very soon. Eric Janes will be retiring on Feb. 29, 2004 after three decades of service with the BLM, the last 6 of which Eric served on the SOH as a Headquarters representative. For the next several months, Jim Fogg, BLM Alternate Member of the BLM’s National Science and Technology Center will provide the continuity for our representation to the SOH. By late spring or summer, the BLM hopefully will communicate with ACWI on who the Primary Member will be.
The meeting was adjourned at 1:00 p.m. A brief presentation from NWS followed the meeting.
George Smith presented a summary of NWS’ efforts
in developing a Community Hydrologic Prediction System (CHPS)
using the PowerPoint slides of Appendix B as a hand-out material
.
Appendix A References for the Meeting Agenda
Item 8: “Discussion of Potential for a Universal Definition
of a Flood” In an effort to present more than one viewpoint,
I invited Will Thomas to participate in the preparation of a paper
for the SOH. Of note is the fact that Will disagrees with me regarding
the reach of FEMA’s definition of a flood throughout the
Federal Government. Presented below are several definitions of
“flood” that have been gathered by Will and me. Please
note that 44CFR 59.1 is the only definition of “flood”
that is the result of a formal rule-making process. Another point
of disagreement between Will and me is that it is Will’s
opinion that the peak flow data relating to debris-effected events
should be shown in a secondary database and available to the public,
but not is the main database. It is my opinion that the peak flow
data should be shown in both databases. Appendix A.2 A discussion of floods, debris flows, mudslides and landslides prepared for the January 29, 2004 meeting of the Subcommittee on Hydrology (Prepared by Will Thomas and Martin Becker, January 20, 2004) (A) FEMA National Flood Insurance Program (NFIP) Regulations (a) A general and temporary condition of partial or complete inundation of normally dry land areas from:
FEMA further defines Mudslide (i.e., mudflow) as: “Mudslide (i.e., mudflow) describes a condition where there is a river, flow or inundation of liquid mud down a hillside usually as a result of a dual condition of loss of brush cover, and the subsequent accumulation of water on the ground preceded by a period of unusually heavy or sustained rain. A mudslide (i.e., mudflow) may occur as a distinct phenomenon while a landslide is on progress, and will be recognized as such by the Administrator only if the mudflow, and not the landslide, is the proximate cause of damage that occurs.” National Research Council (NRC) 1982 report on “Selecting
a Methodology for Delineating Mudslide Hazard Areas for the National
Flood Insurance Program FEMA asked the NRC Committee on Methodologies for Predicting Mudflow Areas for a “review of the Congressional intent undergirding the mudflow coverage and for a comparison of the flood insurance definitions of mudflow with the many-varied scientific classifications that which mix mudflow as a peril with the separate peril of landslide.” A major intent of the NRC report was to clarify how mudslides were different from other landslides for identification purposes and insurance claims within the NFIP. The NRC Committee distinguished between mud floods (debris flows) and mud flows (mudslides), the latter of which is considered a type of landslide. NRC defined “mud flood” as a flood in which the water carries heavy loads of sediment (as much as 50 percent by volume), including coarse debris. NRC defined “mud flow” as a specific subset of landslides whose dominant transporting mechanism is that of a flow having sufficient viscosity to support large boulders within a matrix of smaller-sized particles. NRC referenced the congressional record but did not consider it in their classification system. The following classification system was provided by NRC:
Figure 1 from the 1982 NRC report. As shown in Figure 1, floods were identified as occurring in floodplains, drainage channels and alluvial fans. Landslides were identified as occurring on hillslopes with the exception that mud flows (mudslides) occasionally occur in drainage channels and on alluvial fans. As shown in Figure 1, damages from clear water floods, mud floods (debris flows), and mud flows (mudslides) are covered by flood insurance within the NFIP. The NRC report provided some recommendations of which the first was “A clear decision by FEMA as to which phenomena are to be included under the mudslide provisions of the NFIP and which are to be excluded, with reference to a standard classification scheme for earth movements.” (B) U.S. Geological Survey Fact Sheet 176-97, “Debris-Flow Hazards in the United States”, dated 1997 USGS Fact Sheet 176-97 defines debris flows as fast-moving landslides. USGS further states that “Debris flows, sometimes referred to as mudslides, mudflows, lahars, or debris avalanches, are common types of fast-moving landslides”. Pictures of debris flows in Fact Sheet 176-97 indicate that debris flows can occur solely on hillslopes. USGS Fact
Sheet 176-97 is available on the USGS (Geologic Division)
web site: (C) USGS Open-File Report 85-276 A-D, Feasibility of Nationwide Program for the Identification and Delineation of Hazards from Mud Flows and Other Landslides, dated 1985, edited by Russell Campbell Russell Campbell, the editor of USGS Open-File Report 85-276 A-D, was a member of the NRC Committee on Methodologies for Predicting Mudflow Areas that wrote the 1982 NRC report. USGS Open-File Report 85-276A indicates “The assertion in the regulatory definition that heavy rainfall and absence of brush cover are usual conditions for the occurrence of “mudslides (i.e., mudflows)” is inaccurate. Neither rainfall nor absence of brush cover are requisite conditions for the formation of mud flows, which begin on well-vegetated slopes at least as commonly as on slopes that have lost brush cover, and mud flows occur in association with snowmelt as well as rainfall; nor is either of those conditions required for the initiation of a mud slide.” USGS Open-File Report 85-276A provides a more detailed description
of the various types of landslides and indicates that “wet
non-cohesive flows”, not “wet cohesive flows”
are intended for coverage under the NFIP. This report is consistent
with the NRC 1982 report in defining Mud Floods (debris flows)
as a type of flood. As is the case the 1982 NRC report, mudflows
are shown as being covered by the NFIP. In USGS Open-File Report 97-438 (D) Other Issues In addition to the issues discussed in this paper, there are other issues relating to the debris flow topic. The issues that we have discussed and should discuss include:
Appendix A.3 Martin
Becker’s email of November 13, 2003 When you read the article, below, you will note
a discussion regarding Deer Creek. One of the reasons that the
area below the the Deer Creek debris basin has become populated
is because the USGS expunged the flow of record from its data
bank because it was a debris flow. Therefore, the event of record
was not be fully considered in the analysis of potential flooding. Martin Becker Appendix A.4 FIA
MEMORANDUM
Appendix B - “A Community Hydrologic Prediction
System” Presentation Handout PowerPoint.Show.8 Appendix B – NWS Precipitation Frequency Estimates and Federal Responsibilities (Double click here to view PowerPoint presentation) |
|||||||||||||||||||||||
