You are on page 1of 4

Thus, using the same data, two different conclusions are obtained.

It is the writer's opinion that when dealing with the specific case, the
alluvial expansion angle is known and therefore this fact should be used in
determining the probability of flooding across the alluvial fan in the partic-
ular instance. The methodology that uses the alluvial expansion angle data
is the one presented by the writer, the conclusions of which support the
Downloaded from ascelibrary.org by Universidad de los Andes on 10/03/18. Copyright ASCE. For personal use only; all rights reserved.

uniform probability distribution. It is therefore suggested that the uniform


probability distribution continue to be used until enough data are compiled
to show otherwise.

Closure by Richard H. French 6

First, the writer is obliged to disclose his relationships with those who
provided discussions. Flippin completed her graduate studies under the
supervision of the writer. Fuller and O'Brien are colleagues and friends of
the writer and, as the references in the discussions suggest, collaborators
on some projects. Stevens previously contacted the writer seeking additional
data and information. The writer encouraged his colleagues to discuss this
publication and is pleased that busy professionals found time to write these
discussions. The lack of discussions from the regulators and their consultants
should be noted.
Flippin provides useful insight into the behavior of the model with dif-
ferent sets of fan and LP3 parameters. The points made in this discussion
are valid.
O'Brien and Fuller raise important technical and philosophical issues.
First, as suggested in their discussion, the FEMA model is and has been a
controversial approach to a difficult and critically important problem in arid
and semiarid climates. Given the episodic nature of flooding in this envi-
ronment, the limited historic data available, and the lack of a focused re-
gional research and data-collection program, this model, or any other model,
is likely to remain controversial in the future. At least as early as 1984,
appeals for regional research and data-collection programs were made (Dawdy
1984); and to date, no creditable regional program has been undertaken.
The writer does note and commend the recent efforts by the Flood Control
District of Maricopa County to undertake a data-collection program (Al-
luvial 1992). While this is a noteworthy effort, it will have limited regional
impact because of its local focus.
Second, there can be no debate over the statement that the FEMA model
has limitations. The primary issue is whether in a specific situation the
limitations of the model constitute a fatal flaw. This issue is not unique to
the FEMA model; rather, it is a generic issue that applies to all models. In
this context, the use of the terminology alluvial fan can result in confusion
and errors in judgment. As used in the regulatory and hydraulic engineering
environments, alluvialfan refers to a landform of a specified shape on whose
surface certain hydraulic processes are expected to take place. Therefore,
a landform may be an alluvial fan from a geomorphologic viewpoint but
not from the viewpoint of hydraulic engineering. While there are a number
of cases where the limitations of the FEMA model have been violated and
the results produced in error, there are also cases of alluvial fan flooding
6Res. Prof., Water Resour. Ctr., Desert Res. Inst., P.O2 Box 19040, Las Vegas,
NV 89132-0040.
1320

J. Hydraul. Eng., 1993, 119(11): 1320-1323


where the underlying assumptions of the F E M A model have been met and
the events have behaved as predicted. The writer has personally observed
(postevent) a channel that was flow formed by a flood event on an alluvial
fan of the Gillis Range east of Hawthorne, Nev. During this event, the
channel also avulsed. The regulatory and engineering communities must
recognize that in dealing with complex situations, a model cannot be fitted
Downloaded from ascelibrary.org by Universidad de los Andes on 10/03/18. Copyright ASCE. For personal use only; all rights reserved.

to the situation; rather, a model appropriate to the situation must be found


or developed. A good example of appropriate use of the available models
to delineate flood hazard in arid environment can be found in Schmeltzer
et al. (1993). Many of the problems with the results of the FEMA model
noted by O'Brien and Fuller derive from misuse of the model; poor engi-
neering judgment; and/or lack of a detailed field inspection by qualified
professionals rather from any flaw in the model itself. For example, the
FEMA model is sensitive to the value of the LP3 skew coefficient, which
is generally estimated from flow events derived from simulated rainfall-
runoff models. As indicated in O'Brien (1990) and Schmeltzer et al. (1993)
appropriate adjustments to the rainfall-runoff model must be made to achieve
reasonable and creditable runoff values and estimated LP3 skew coefficient
values. In many applications of the F E M A model, this fundamental concept
has neither been recognized nor addressed. Erroneous results deriving from
such efforts are not a problem with model but with the knowledge and
judgment of the user of the model.
Third, there has never been the suggestion that the F E M A model or its
derivatives should be applied within urban areas. O'Brien and Fuller fail to
recognize that many important and critical facilities such as low-level ra-
dioactive, mixed, and hazardous-waste management sites; industrial facili-
ties; mine support facilities; and transportation alignments are located on
alluvial fans where no other improvements exist. The statement of O'Brien
and Fuller to discourage the use of the F E M A model in urban or improved
situations is accepted with the caveat that there are situations, such as fringe
development areas where the use of the model is entirely appropriate; see
for example, Hydrologic (1990).
Fourth, there is the issue of what is intended by the word design. The
paper suggests that the model can be used to estimate the design flow of a
specified return period to a location. There is no suggestion in the paper
of a method to estimate the depth or velocity of flow at a specified point;
the discussion confuses the F E M A use of the model to delineate hazard
zones with the intention of the paper. It is further suggested that the pro-
posed model not be used for design. In a subsequent paper by O'Brien
et al. (1993), it is stated

Flood hazards on alluvial fans are presently delineated with a simplistic


probabilistic model . . . Since the F E M A method doesn't simulate
flood hydrographs, it is inappropriate for the design of flood mitigation
structures such as levees, flood containment w a l l s . . . (O'Brien et al.
1993).

In O'Brien et al. (1993), the authors fail to identify why only an estimate
of peak flow is not adequate for design. Further, the model discussed in
the paper can be easily modified to estimate a conservative hydrograph at
point of interest, as appropriate to the situation.
Fifth, the subject of engineering conservatism must also be addressed.
The discussors fail to define what they mean by conservative. The writer
1321

J. Hydraul. Eng., 1993, 119(11): 1320-1323


cannot comment on the referenced article (O'Brien and Fullerton 1992),
which is not easily available; however, in the case of O'Brien and Fullerton
(1990) a similar claim is made. The case discussed in this paper, Hiko Springs
Wash, is an example of the misapplication of the FEMA model. Hiko
Springs Wash is a deeply incised channel with a meandering ephemeral
stream in the bottom. It is not an environment that meets the hydraulic
Downloaded from ascelibrary.org by Universidad de los Andes on 10/03/18. Copyright ASCE. For personal use only; all rights reserved.

criteria for the FEMA model. An upcoming paper by Flippin and French
(1993) specifically addresses the issue of conservatism.
The discussion by Stevens also raises important issues and presents new
data. The writer takes issue with the implication of the phrase "how the
data were used to produce the conclusions." Given prior written commu-
nications after the publication of the paper, Stevens is well aware that the
writer appropriately used the data available to him. Further, the effect of
the expansion angle on the absolute value of the channel angular deviation
from the medial radial line is addressed, using the data available to the
writer, in Fig. 2. Unfortunately, Stevens does not plot his data in a similar
figure. The data presented by Stevens is a valuable addition to our knowl-
edge, and the profession would benefit by its publication in the peer-re-
viewed literature.
In response to the substantiative assertions of Stevens, the writer would
question whether alluvial fans in Alberta exhibit the same hydraulic behavior
as alluvial fans in the arid and semiarid southwestern United States. There
are very significant temperature, precipitation depth and duration, vege-
tative, and tectonic activity differences that may result in process and be-
havior differences. For example, statistical testing of Anstey's (1965) data
regarding the size and slope of alluvial fans in the United States and Pakistan
demonstrates that there are significant differences between the fans in these
two countries. The concept of a uniform probability distribution refutes the
physical observations of Bull (1964), which is quoted in the paper. The
writer requested the complete data set used by Stevens and has received
these data. The writer intends to evaluate these data, and will, hopefully
with the cooperation of Stevens and ASCE, publish a technical note in the
future.
In conclusion, the writer would note there is one point of agreement
among all involved with this paper and the discussions; that is, the current
state of the art of hydraulic design on arid and semiarid alluvial fans is not
adequate. In this important area of hydraulic engineering, there is a clear
basis for valid disagreements among professionals, and there is also the clear
potential for misuse of the existing models. Further, no single model will
adequately address all alluvial fan flood hazard issues. The need for data
collection and further analysis has been accurately identified. Finally, the
writer thanks those who prepared discussions for their time, effort, and
comments.

APPENDIX. REFERENCES
Alluvial fan data collection and monitoring study. (1992). Flood Control District of
Maricopa County, Phoenix, Ariz.
Dawdy, D. R. (1984). "Mud flood, mudslide and alluvial fan hazards." Proc. of a
Workshop Improving the Effectiveness of Floodplain Management in Western State
High-Risk Areas, Special Publication 9, The Association of State Floodplain Man-
agers, Madison, Wisc., 39-41.
Flippin, S. J., and French, R. H. (1993). "Verification of an alluvial fan drainage
design methodology for transportation alignments." J. Irrig. Drain. Engrg., ASCE.
1322

J. Hydraul. Eng., 1993, 119(11): 1320-1323


Hydrologic criteriaand drainage design manual. (1990). Clark County Regional Flood
Control District, Las Vegas, Nev.
O'Brien, J. S. (1990). Off-site alluvial fan flood hazard delineation for the Primm
Property Stateline, Nevada. Lenzotti & Fullerton Consulting Engineers, Inc.,
Breckenridge, Colo.
O'Brien, J. S., Julien, P. Y., and Fullerton, W. T. (1993). "Two-dimensional water
flood and mudflow simulation." Y. Hydr. Engrg., ASCE, 119(2), 244 261.
Downloaded from ascelibrary.org by Universidad de los Andes on 10/03/18. Copyright ASCE. For personal use only; all rights reserved.

O'Brien, J. S., and Fullerton, W. T. (1990). "Two-dimensional modeling of alluvial


fan flows." Proc. of the Int. Syrup. on Hydraulics~Hydrology of Arid Lands, R.
H. French, ed., ASCE, New York, N.Y., 262-267.
Schmeltzer, J. S., Miller, J. J., and Gustafson, D. L. (1993). "Flood assessment at
the Area 5 radioactive waste management site and hazardous waste storage unit,
DOE/Nevada Test Site, Nye County, Nevada." Raytheon Services Nevada, Las
Vegas, Nevada.

1323

J. Hydraul. Eng., 1993, 119(11): 1320-1323

You might also like