***************************************************************************************
FEQ--Summary
of user notices and pending bug fixes since Version 8.92.
Version 8.92 has
been renumbered to Version 9.15. See release.txt in the
distribution files or
http://il.water.usgs.gov/proj/feq/feqdoc/release.html
for further details and
additional notices. This list is superseded by the
release.txt distributed with
subsequent versions of FEQ. User notices and bug
reports for subsequent releases
of FEQ and FEQUTL may also apply to this
version.
The version numbers are for reference only--the fixes have not been applied
to Version 9.15.
(FEQUTL notices follow)
***************************************************************************************
Version 8.93
July 8, 1997:Version 9.52
--Metric units and HECDSS for diffuse runoff values will not function properly because no unit conversion is done on the HECDSS unit-area runoff intensity values. Currently FEQ assumes that the units for the diffuse runoff values in the HECDSS are inches per time interval and the interval is of fixed length for all files involved. The DTSF files, on the other hand, are assumed to have the runoff intensity in feet per second for the customary US set and meters per second for the metric set of units.
November 9, 1999:Version 9.65
--Factor on value in HECDSS files in the INPUT FILES block has a default of zero when it should have a default of 1.0.
March 7, 2000:Version 9.80
--Significant errors found in the STDCX governing-equation option. The error is in the computation of some of the derivatives of the residual function as well as an error in one part of one function-table lookup routine. The governing-equation options other than STDX has little use and therefore some errors have been present for some time. Recommend that users use only STDX until further notice.
March 15, 2001:Version 9.84
--An error has been discovered in the computation of the Hager side-weir correction factor that has been present in the side-weir routine (Code 14) of FEQ for at least 10 years. The factor is in the correct range (<= 1.0) but has been coded in FEQ to return larger values than are referenced in the FEQ documentation in section 8.1.2.1.3.2 (p. 83) to Hager (1987).Testing indicates that the effect of the error is most dominant at low heads in the range of 0.2 to 0.4 feet where simulated flows may exceed the correct Hager computation by about 25 percent. At heads near one foot the simulated flows exceed the Hager computation by about 5 percent and at 2 feet or above, the error is approximately 2 percent. Therefore, the effect of the error on existing models depends on the operational head. Because the computed flow approximations over side weirs at low heads have large uncertainties, results from physical model testing are often implemented as a rating in the model instead of using the side-weir routines.
The above error will be fixed in a future FEQ update. To obtain a copy of the fixed code now, send e-mail to h2osoft@usgs.gov or alishii@usgs.gov.
Problem exists in interpolation of cross sections when a bottom slot is present. Recommend that users do not interpolate cross sections when a bottom slot is present.
October 1, 2001:Version 9.91
The point flows block is no longer supported. Point flows should be connected to the model at a junction created for each point flow with a dummy branch added to provide the boundary node to reference the point flow.
May 17, 2002:Version 10.25
Found a subtle bug in the computations of one derivative in the side-weir code. This appears only if the weight factor on head in the source channel differs from 0.5.The multiplying factor for the forced-boundary instruction, code 6, is applied to both the base value and the time-series value. The base value serves as a constant lower limit on the flow or elevation at that location.
September 27, 2005:**********************************************************************************************
An error has been discovered in reporting the time step when using a diffuse time-series file (DTSF). The reported time offset depends on the initial hour of each event. If the starting hour is zero then the offset is zero. Otherwise the offset depends on the starting hour. If the starting hour is 1, then the offset is 1 - 1/24.0 hours, which is slightly less than one hour. Thus the maximum offset in the reported time is one hour or less. It appears that this discrepancy has been present for some time. To prevent this bug, use a starting time of 0.0 when using a DTSF.
The version numbers are for reference only--the fixes have not been applied
to Version
4.85.
**********************************************************************************************
Version 5.06
January 18, 2000:Version 5.31
--Roughness factor greater than 0.60 in CULVERT command will cause failure of a table lookup.Error found in the default values in the EMBANKQ command. CSHIFT is defaulted to 1.0 when it should have been defaulted to 0.0. This results in an increase of crest elevation for the embankment of 1.0 feet. Do not use the default value of CSHIFT.
December 21, 2000:Version 5.32
--The low-head weir coefficient in EMBANKQ and in flow over the road in CULVERT is improperly allowed to vary with the submerged total head as an argument in computing submerged weir flow. The argument should be held fixed at the free-flow total head value. Limited testing shows that the effect is generally smaller than 0.05 percent of the flow. Some submerged flows less than 10 cfs, appear to to affected by less than 0.5 percent. The only values affected are those in which a significant approach velocity head is present. The flows are unaffected if the velocity head is negligible.
January 22, 2001:Version 5.35
--An erroneous error message is printed if UFGATE has less than two gate openings given.An error exists in computing LEFT and RIGHT options for cross sections where either or both points of intersection do not match an existing point on the cross section. A check for consistency of elements in the cross-section function table before it was output should give notice if any of these errors cause significant differences in the computed elements, however until an update is issued, all LEFT and RIGHT options should intersect with an existing point on the cross section.
February 9, 2002:
--Problem exists in interpolation of cross sections when a bottom slot is present. Recommend that users do not interpolate cross sections when a bottom slot is present.