Updating the Hydro-Climatic Data Network (HCDN)

To: "A  - Division Chief and Staff",
        "B  - Branch Chiefs and Offices",
        "FO - State, District, Subdistrict and other Field Offices",
        "PO - Project Offices"
cc: "CD - All Data Chiefs",
        " , Computer Policy Advisory Committee, Reston, VA ",
        " , WRD Archive File, Reston, VA ",
        "Annette L Ledford, Clerk Typist, Reston, VA "
From: "Thomas H Yorke Jr., Chief, OSW, Reston, VA "
Subject: OSW Technical Memorandum No. 97.05
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Tue, 08 Jul 1997 09:58:35 -0400
Sender: "Annette L Ledford, Clerk Typist, Reston, VA "

In Reply Refer To:                                         July 3, 1997
Mail Stop 415

Office of Surface Water Technical Memorandum No. 97.05

Subject: Updating the Hydro-Climatic Data Network (HCDN)

The Hydro-Climatic Data Network (Water-Resources Investigations Report 
93-4076) has been one of our most popular data products; it is 
consistently in the 20 favorites on the wwwrvares server and has been 
cited in papers in refereed journals as diverse as Freshwater Biology and 
Journal of Water Resources Planning and Management. This collection of 
73,231 water years of streamflow data measured at 1659 sites throughout 
the United States and its territories has been of great value to those 
interested in the long-term and regional variations of streamflow. The 
purpose of the HCDN "is to provide a single, division-wide compilation of 
discharge stations having periods of record for which the monthly and 
annual streamflow values are suitable for climatological analysis." 
(Office of Surface Water (OSW) Technical Memorandum No. 89.15) The 
current HCDN dataset, sometimes known as HCDN92, contains periods of 
record from water year 1874 through water year 1988.

While all of our data are valuable, the HCDN has special scientific 
interest as an organized, documented, and quality-assured collection of 
streamflow records representing national conditions. Furthermore the 
publication of this dataset on CD-ROM allows others, such as university 
professors, to mount the data on their own local servers for access by 
students in their hydrology classes. We feel the special effort - above 
and beyond putting all of our streamflow data on the H2O server - is 
warranted to keep this excellent product in the hydrologic community's 

This memorandum is a request for each District to assist in updating the 
records that qualify for inclusion in the HCDN. Briefly, the criteria 
(first defined in OSW Memorandum No. 88.05) are that each water year of 
discharge which meets the following four criteria should be included in 
the HCDN:

1.  Long records: n>=20; preferably, n>=50, but the 20-year threshold
    was chosen to allow the inclusion of the benchmark stations.

2.  Good quality data; this pertains to the quality of the stage record,
    the discharge measurements, and the stage-discharge relation.

3.  Broad geographic and diverse climatologic representation across the

4.  Minimal anthropogenic effects in basin, including minimal regulation
    of the stream discharge or reduction of base flow due to extreme
    ground-water pumping.

(The fifth criteria stated in OSW 88.05 of the basin size being within a 
certain range was dropped in OSW 89.15 as irrelevant, given all the 
exceptions to it.) A detailed discussion of the criteria was given in OSW 
89.15, the relevant portion of which is attached as an Appendix.

We are requesting that each District:

1.  Review the list of acceptable record at HCDN stations they operate
    (this will be sent electronically to the District Data Chief unless
    the District Chief designates someone else) and report the results
    of the review to Robin Sevin (rssevin@usgs.gov) by August 15, 1997.
    Your review will determine 

    a.  what part of the discharge record from water year 1989 through
        water year 1996 at stations already in the HCDN and in operation
        at the end of water year 1988 qualify for adding to the HCDN and

    b.  what record at stations not already in the HCDN should be added;
        i.e., are there stations that didn't have 20 years of record in
        1988 but otherwise qualified and that now have 20 or more years
        of record which qualifies for inclusion?

2.  Assure that all of the discharge data to be added to the new HCDN
    (to  be known as HCDN97) are available from the national web server
    by September 2, 1997. The update software is available from the
    Illinois District as announced in a recent email from Verne Schneider
    to the field.

We plan to retrieve the new data from the H2O server in early September 
and have new copies of the CD-ROM available by December 1997. Your 
cooperation is supporting the HCDN and helping to make the data available 
by CD-ROM are appreciated. Your efforts will further the use of our 
streamflow data as an invaluable scientific tool. 

                            (from OSW 89.15)

The accuracy of this list [the HCDN] is critical--there is great interest 
in using the USGS data in climate change studies throughout the 
hydrologic community. Please ensure that great care is taken to review 
this list bearing in mind the potential embarrassment of having included 
unsuitable periods of record or having missed key stations. THE 
There will be no higher level technical screening of the individual 
records which the District chooses.

Below is a clarification and expansion of the selection criteria of OSW 
Technical Memorandum No. 88.05. Please remember there may be exceptions 
to any of these criteria. We need to document such exemptions on a 
station-by-station basis by means of an explanatory comment in the list.

Criterion of record length: At least 20 years of suitable record. If a 
station has been operated for 50 years and only the first 20 years are 
free of defects discussed below, then only that 20 years should be 
selected. If, however, a shorter record is all that is available to fill 
a gap in geographic or climatic representation, it is acceptable. A 
specific example is the inclusion of short records from islands in the 
Pacific because that is all that is available for that geographic region.

Criterion of data quality: A rating of "good" or "excellent" on nearly 
all of the annual daily discharge records. A few days in some years rated 
"poor" or "fair" because of estimated record or similar considerations 
would not necessarily disqualify a station, nor would a few years rated 
"poor" or "fair" out of many rated "good" or "excellent."

Criterion of coverage: Broad geographic and diverse climatic 
representation across the country. We prefer to have too many stations 
rather than too few. It is always more desirable for the individual users 
to have the option of paring down the list according to whatever density 
or accounting criteria they choose than to be supplied with an 
insufficient number of stations. Thus we now consider the suggestion of 
one or two stations in each accounting unit or areal rectangle as, at 
best, a desirable minimum. ALL stations and periods of record which fit 
the other criteria should be included in the list.

Criterion of natural conditions: Minimal anthropogenic effects such as 
diversion, augmentation, land-use changes, or regulation of the stream 
discharge or reduction of base flow due to extreme ground-water pumping. 
This is the most difficult criterion. The objective is to include a 
discharge station if the anthropogenic effects are not felt to mask the 
effects of climate variation in the monthly and annual values.

    a.  Stations subject to regulations by dams are generally not
        acceptable. However, as an example, a low-head hydropower dam
        which has only a transient effect on high or low flows, but not
        on the monthly values, is acceptable. This effect should be

    b.  Stations subject to diversion or transfer for any use are
        generally not acceptable. However, if the diversion has existed
        essentially unchanged for the entire period of record, the
        station may be acceptable. For example, a station in, say, the
        Nevada desert may be subject to diversions for irrigation but 
        the practice has remained unchanged over the entire period of
        record. So, while the diversions exist, and do affect the actual
        values, the process by which they alter the record is, in a
        sense, unchanging and does not significantly confound the
        influence of climate on the record. These considerations should
        be noted.

    c.  A station in a basin which has undergone a substantial change in
        usage, say forest to agriculture or urban, is probably not
        suitable even if the change was gradual. If such a station is
        included, such change should be noted.

PLEASE NOTE: Even though it will be stressed in the report that the 
selected periods have been qualified so that at least the monthly 
averages meet our criteria, users of this data set may wish to do 
analyses on a shorter time step, possibly daily. Please indicate any 
conditions which would make the discharge records unsuitable for analysis 
at less than a monthly resolution.

Inclusion of formerly suitable stations: If a station is not presently 
suitable because of regulation, diversion, augmentation, or because it 
has been discontinued but had a period of record which on its own is 
suitable, the suitable period should be listed. Likewise, the fact that a 
station might soon be discontinued or become regulated is not cause to 
exclude it. We are making no statement about the suitability of records 
for water year 1997 or later. The heart of this effort is to identify 
existing periods of record suitable for climate analysis. 

Exclusion of constructed records: Records that are constructed (e.g., the 
Four Rivers Index of the Sacramento Basin) or reconstructed (e.g., the 
"natural flow" for the Colorado River at Lee's Ferry) from other sites or 
information on diversions, augmentation, pumping, regulation, etc. are 
generally not suitable. While such records may be indicative of current 
conditions, we wish to avoid the question of whether the computational 
algorithm might disguise the effects of climate. We do not mean to 
exclude stations for which the effect of diversions, etc., are easily and 
routinely corrected. A comment to this effect should be included for 
these records.


                                 Thomas H. Yorke
                                 Chief, Office of Surface Water

WRD Distribution: A, B, FO, PO