PUBLICATIONS - Water temperature records - thermographs 


                                          September 23, 1959

QUALITY OF WATER BRANCH MEMORANDUM NO. 60.13

To:       District Engineer, SW, Portland, Oregon
             (Through Chief, Surface Water Branch)

From:     Chief, Quality of Water Branch

Subject:  PUBLICATIONS - Water temperature records - 
                         thermographs

     Thank you for your interest and the comments on the 
subject records noted in your memorandum of September 3, 
1959.  The discrepancies which you pointed out have again 
alerted us to the need for careful review of our data before 
publication, and the probable need for repeatedly cautioning 
the field offices on the necessity for adequate checking.  So 
far as we know, of the items you mentioned in Water Supply 
Paper 1400 only the minimum for May 9, page 458 was a 
typographical error.  The remaining data were completed the 
same as submitted to us for publication.  We do not
know why records have been transmitted that are manifestly
inadequately checked, but suspect the very simplicity of the
tabulations and the relatively small number of thermographs 
in most individual districts is a reason for casual treatment 
in many cases.

     The items you note are almost all based on the fact that 
on a thermograph record the maximum temperature on a given 
day cannot be correctly reported as lower than the minimum
temperature on an adjoining day.  For the record of the 
Schuykill River at Passayunk Ave., Philadelphia, p. 74, WSP 
1400, this occurrence does not necessarily mean an error, 
because the record is neither a thermograph nor a.m. and p.m. 
readings as you assumed; the record was obtained from the 
Atlantic Refining Company who has an observer reading a 
thermometer, presumably every few hours.  Our manuscript 
heading was imperfect in omitting the temperature RECORDS 
AVAILABLE as well as not describing how the record was 
obtained; certainly values on November 22 and May 25 also 
should have been revised.

     You did not mention our general memorandum of 
instructions, subject:  TEMPERATURE - Surface Waters, to: All 
District Chiefs and Staff Officials for the Quality of Water 
and Surface Water Branches dated October 1, 1956, in which we 
call attention specifically to maximum and minimum values 
from thermograph records.  We hoped that those instructions 
would result in satisfactory data for publication.  Time and 
personnel have not been available in this office for detailed 
checking of each table and those which were thought to be 
straightforward and complete have usually been only scanned 
and spot checked in review.

     In spite of the volume of thermograph records (about 155
scheduled for 1957) we do not believe the accuracy of these
tables would necessarily be improved at the present time by
putting on tape or using automatic processing.  The handling 
and manipulations involved in punching tape and using machine 
checks would undoubtedly cost more than simple thorough 
checking of tables prepared by normal methods, and still not 
completely eliminate errors.  However, we are currently 
developing a coded punch tape system for processing water 
quality and temperature data, but this will not entirely 
eliminate the need for personal checking of data.

     We want to present accurate meaningful data in our 
annual reports and will require more strict adherence to the
instructions on future tabulations.

                              (signed)


                              S. K. Love

WRD Distribution: A, B, S4, FO-L4, S2, FO-L2

WRD Distribution: A, B, S4, FO-L4, S2, FO-L2

                                        September 3, 1959

To:       Chief, Surface Water Branch, Washington, D.C.

From:     District Engineer, SW, Portland, Oregon

Subject:  Water temperature records -- Obvious errors in 
          WSP 1400


     We are going to go back on our word (see last paragraph 
of our memo of August 13, 1956, copy enclosed) and make one 
more complaint about the type of errors involving a maximum 
daily temperature that is lower than the minimum of an 
adjoining day.

     Yesterday we received WSP 1400 and decided to check for 
the max-min error to see if records had improved any.  We 
checked 12 daily tables from 11 different States and found 
that type of error in 7 of the records involving 6 States.  
We checked one record each for Kentucky, Tennessee, Illinois, 
Michigan, and Maryland and found no errors although on page 
458 it seems obvious that minimum on May 9 should be 68o not 
58o (probably a
type.).

     The errors in the 7 records mentioned above occur as
follows:

Page   State        No. of errors         Days involved

 39     N.Y.            5           Oct. 6-7, 16-17, Nov. 21-
                                    11, 22-23. Dec. 12-13
 41     N.Y.             1          Sept. 21-22
 74     Pa.          88 errors out of 247 days, as many as 18
                        days in one month are involved and
                            3 errors are 6o or 7o.  Possibly
                            the headings for these columns
                           should be a.m. or p.m. rather than
                           max. & min. but nothing to so
                           indicate.  For that matter the
                           description on page 72 does not
                           even refer to water temperatures
                           under Records available.
 142    Va.              3          Dec. 28-29, Jan. 16-17,
                                    Aug. 7-8.
 351    Ohio             2          Jan. 7-8, Apr. 22-23.
 442    Ind.             3          Dec. 16-17, Mar. 30-31,
                                    Apr. 1-2.
 469    N.C.             1          June 7-8, a 3o error, but
                            perhaps this is a type. and max.
                            for Jan. 8 (Note:writer probably
                            means June 8), should be 68o.

     Possibly water temperature data could be put on tape to 
let the datatron pick out the errors.  At any rate a serious 
effort should be made to avoid publishing figures that make 
us look ridiculous in the eyes of any critical user.

                              /s/ Kenneth N. Phillips
                                   District Engineer