RECORDS--Policy on the Publication of Estimated Flows

In Reply Refer To:
WGS-Mail Stop 415                         March 8, 1990


Subject:  RECORDS--Policy on the Publication of Estimated Flows

Flow records on some streams are poor at times because of channel
and flow conditions or because of monitoring problems.  The
U.S. Geological Survey is sometimes asked to obtain flow records
on streams that are very difficult and/or very expensive to gage.
Flows can be estimated for some of these sites as accurately, at
times, from flow routing or regression procedures or, possibly in
some cases, from precipitation-runoff modeling, as from a flow
record collected at the site.  This policy statement provides
guidelines for the publication of estimated flows.

In situations where there is missing or unreliable record and
these periods of record must be estimated, methods such as flow
routing or regression analysis may be used to estimate the flow
during these periods.  Although periods of routed or modeled flows
used to fill missing periods can be published in an annual data
report and stored in WATSTORE, long periods computed for special
projects should not be.  The estimated record, however, must be
clearly identified and qualified, as noted in the "WRD Data
Reports Publication Guide," if it is considered to be of a differ-
ent accuracy than the monitored flow record.

Flow records estimated from precipitation-runoff models may be of
statistical value but may not be an adequate representation of
daily or peak flows at a site because of the uncertainty that the
precipitation record is representative for a given day or storm
event.  Therefore, flows estimated from precipitation-runoff
models should not be published in the annual data reports or
stored in WATSTORE.

Flows estimated from flow routing or regression models or from
precipitation-runoff models may be published in other reports
provided that it is clearly identified that the records were
obtained from a model and are not from monitored data at that
site.  The methods and assumptions for the modeled flows also
should be clearly described.  Flow records computed from a
hydraulic routing model such as BRANCH, where stage data are
obtained at the gage site and at a site either upstream or down-
stream from the primary gage, can be published in the annual data

Records published in the annual data reports should be based on
one or more types of measured correlative data of the stream, such
as stage or velocity, at the site.  Estimates of missing record
should be made using the most accurate method available and then
should be qualified as appropriate.  Flow records published in the
annual data reports and stored in WATSTORE should represent the
best possible flow record at the site indicated.  These records
are sometimes used in court and are also used for a variety of
other water-related decisions.  The records must be computed in
such a way that the U.S. Geological Survey can support published
discharges and computational procedures in court or in dealings
with other agencies and the public.

                           Charles W. Boning
                           Chief, Office of Surface Water

WRD Distribution:  A, B, S, FO, PO

This WRD Memorandum does not supersede any previous WRD