Guidelines for Frequency of Discharge Measurements during Ice-effected Periods

In Reply Refer To:                                January 26, 1994
Mail Stop 415


SUBJECT:  Guidelines for Frequency of Discharge Measurements 
          during Ice-effected Periods

The purpose of this memorandum is to provide guidelines for the 
frequency of discharge measurements during periods of ice-affected 
streamflow at gaging stations.  Office of Surface Water (OSW) 
technical reviewers have noted during District reviews that the 
number of discharge measurements during ice-affected streamflow 
conditions has gradually declined over the past several decades.  
There is concern that this situation has progressed to the point 
where the integrity of many winter records has fallen below 
desirable levels.

Determination of daily mean discharge values during winter ice 
periods is based on discharge measurements and in most cases on 
comparison of hydrographs for nearby stations.  The accuracy of 
discharge estimates during ice-affected periods is, therefore, 
directly related to the number and quality of discharge 
measurements.  About one half of all ice-affected discharge 
measurements are considered "poor," according to Janice Fulford's 
recent investigation of measurement accuracy reported in
U.S. Geological Survey (USGS) Open-File Report 92-493 
"Characteristics of U.S. Geological Survey Discharge Measurements 
for Water Year 1990."  There are limitations in improving the 
accuracy of ice-affected discharge measurements, but more frequent 
measurements can be made.

Several USGS publications provide guidance on discharge- 
measurement frequency. Water Supply Paper 337 "The Effects of Ice 
on Streamflow" by W.G. Hoyt, 1913, (p. 67) states:  "The number of 
current-meter measurements necessary to determine the winter flow 
of a stream depends so largely on local conditions that no 
definite rule can be given, but the axiom applicable to open-water 
measurements--the greater the number the greater the accuracy--
applies with equal force to measurements made when ice is present.  
If the control section of a station is always clear of ice or if 
the flow is affected only by anchor ice, no more measurements are 
necessary than are necessary to determine the open-water flow at 
the same station.  If surface ice is present at stations on 
streams uncontrolled by storage, the number of discharge 
measurements will depend almost entirely on the temperature; when 
conditions of temperature and runoff are uniform, one measurement 
a month may be sufficient to insure good records; when thaws are 
frequent the number increases rapidly, and during certain periods 
in the winter, it may be necessary to make measurements almost 

Water Supply Paper (WSP) 2175, Vol. 2, by S.E. Rantz (p. 360) 
states:  "Because of the variability of the backwater effect, 
discharge measurements should be made as frequently as is feasible 
when the stream is under ice cover, particularly during periods of 
freeze-up and thaw when flow is highly variable.  (Procedures for 
making measurements under ice cover are described in the section 
in chapter 5 titled, "Current-Meter Measurements from Ice Cover.")  
In midwinter the frequency of measurements will depend on climate, 
accessibility, size of stream, winter runoff characteristics, and 
required accuracy of the discharge record.  As a general rule, two 
measurements per month is the recommended frequency.  At stations 
below powerplants that carry a variable load, it may be necessary 
to make two measurements during each winter visit--one at the high 
stage of the regulated flow and the other at the low stage.  The 
backwater effects may be markedly different at the two stages.  In 
very cold climates where winter ice-cover persists and winter 
discharge shows a relatively smooth recession, fewer winter 
measurements are needed than in a climate that promotes the 
alternate freezing and thawing of river ice."

In the 1987-88 study of winter discharge record accuracy described 
in USGS Open-File Report 90-554, "Evaluation of Selected Methods 
for Determining Streamflow During Periods of Ice Effect" by
N.B. Melcher and J.F. Walker, discharge measurements were made at 
1-5 day intervals during periods of ice effect on three Iowa 
streams.  These measurements were used to establish a base record 
which was compared with the results of a variety of computational 
techniques based on discharge measurements made at a 6-week 
frequency.  Statistical evaluation of results from the commonly 
used "Hydrographic-and-climate comparison" technique indicated 
that 60 percent of estimated daily discharges had relative errors 
greater than 10 percent, and 9 percent had relative errors greater 
than 50 percent.

An acceptable measurement frequency during winter ice periods may 
vary significantly from stream to stream depending on the relative 
significance of winter flows to annual discharge, accessibility 
during winter months, the economics of frequent measurements, and 
the accuracy goals for given stations.  Although judgment that 
considers these constraints must be exercised in winter data 
collection, Districts should be knowledgeable of the potential 
reduced accuracy of discharge estimates with less frequent 
measurements.  In general, the OSW endorses the recommendations 
given in WSP 337 and 2175 and adds the following guidance:

1.  For streams whose surfaces typically stay frozen for all or 
    most of the winter period but respond to temperature changes
    with change in streamflow, a discharge measurement should be
    made as soon as possible after the ice sheet formation is 
    safe, and no less frequently than once every few weeks 
    thereafter until breakup occurs.

2.  For streams whose surfaces typically stay frozen for the 
    entire winter period and that are characterized by gradual 
    streamflow recessions, a discharge measurement should be 
    made as soon as possible after the ice sheet formation is 
    safe, and at least once every 6-weeks thereafter until 
    breakup.  Particular attention is needed to obtain a discharge 
    measurement at minimum flow, which often occurs just prior to 

3.  For streams whose surfaces stay frozen for a period, break 
    up and then refreeze, at least one discharge measurement 
    should be made during each frozen period.

4.  For streams whose records have good hydrographic comparison,
    discharge measurements at least every four weeks are 
    recommended and additional discharge measurements should be 
    obtained at the best (index) station, particularly during 
    winter temperature extremes to improve definition of the 
    discharge hydrograph.

Districts are encouraged to carefully analyze the quality of ice-
affected gaging station records, taking into consideration the 
number and quality of discharge measurements, groupings of 
stations used for hydrographic comparisons, applicability of 
weather records, and to downgrade the accuracy of published 
records that are derived from limited basic information sources.  
In execution of data programs, it also is recommended that the 
Districts' quality assurance plan address winter discharge 
records, specifying the accuracy goal for various stations or 
groups of stations in the District.  Such accuracy goals may be 
stated in terms of reduced measurement frequency for those 

                                 Charles W. Boning, Chief
                                 Office of Surface Water