Model Change Bulletin
              
                                    MCB#2                         05/07/96

                             ISCST3 (dated 96113)
                             ISCLT3 (dated 96113)


    This Model Change Bulletin documents changes made to the Industrial
Source Complex Dispersion Model (ISC3) (see Code Changes below).  In addition,
this MCB identifies an anomalous behavior in the dry deposition algorithm (see
Dry Deposition Anomaly below).  [NOTE:  New ISC3 users are urged to read ISC3
MCB#1, Amendment #1 for important information.]


CODE CHANGES

    Changes from the ISCST3 (dated 95250) and ISCLT3 (dated 95250) include a
correction to subroutine DEPCOR in DEPFLUX.FOR for both models.  A valid
ELSEIF line was incorrectly commented out.  The change occurs on line 201
in the DEPFLUX.FOR code file.  This correction only affects the dry depletion
option of the models.  This change did NOT cause a change in the test case
results (TEST-ST.ZIP and TEST-LT.ZIP).  Additional limited testing has shown
no changes in model output, using input conditions that would likely show
differences.

    The ISCST3 (dated 96113) model also includes a few minor changes to the
SETUP.FOR and OUTPUT.FOR files for compatibility with the ISCEV3 (EVENT) model,
and a minor change to subroutine HRLOOP in ISCST3.FOR to correct a potential
problem with the STARTEND keyword for non-sequential meteorological data sets.


DRY DEPOSITION ANOMALY
                      
    In addition to this correction, users should be aware of potentially
anomalous behavior of the dry depletion algorithm under some circumstances.
Slight discontinuities in the concentration distribution with downwind distance
have been noted for some stability categories (rural C and D, urban D, E, and
F) for particles in the 5 to 10 micron range when dry depletion is implemented.
The discontinuities have been traced to a discontinuity in the sigma-z profile
with distance after the point where the plume centerline "touched down."  The
plume touchdown distance will vary depending on the source characteristics,
but will generally be from a few hundred meters to a few thousand meters
downwind for stack sources, and less for low-level or ground-based sources.
Up to the touchdown point, the sigma-z is based on the standard sigma-z curves
for ISC.  Beyond the touchdown point, the sigma-z is adjusted to account for a
change in the vertical profile of the plume due to continued settling of the
particles above the plume centerline.  It is currently believed that this
discontinuity occurs at the touchdown point and is due to problems with
convergence of the SZFORM2 subroutine, and possibly with the overall
convergence of the depletion integral by subroutine QATR2.

    Based on limited tests that have been performed, the magnitude of the
anomaly appears to be slight (i.e., may potentially affect results by about 5
to 10 percent or less), and the largest affect will be beyond the point of
maximum concentration or deposition flux.  This appears to be associated with
a discontinuity in sigma-z at the touchdown point and with convergence of
sigma-z in SZFORM2, which seem to gradually disappear at distances well beyond
the touchdown point (i.e., about 5 times the touchdown distance).  It is not
anticipated that the anomaly will continue to get worse with distance.  Also,
it should be emphasized that this affects the dry depletion option only, and
has no effect on the deposition algorithm itself.  The anomaly is still being
investigated, and an update to the models will be issued once it is resolved.
For questions about the appropriateness of use of the dry depletion algorithm
for a specific application, contact the Regional Modeling Contact.


                           *****************
                           *   IMPORTANT   *
                           *****************

    This program will henceforth be identified by its name and a Julian date,
i.e.: ISCST3 (dated 96113) and ISCLT3 (dated 96113).  Future changes to this
model will be identified by an updated Julian date found at the beginning of
the model source code and followed by a reference to a SCRAM BBS Model Change
Bulletin (MCB#n) which will explain the nature and contents of the changes.
All Model Change Bulletins for this program will always be available for back
reference on this BBS.  In the future, notification of changes (MCBs) to this
program will be made in the SCRAM ALERTS section of this BBS immediately
following the LOGON procedure.