March 5, 1991


MEMORANDUM

SUBJECT:  New Jersey DEP Comments on Valley Model

FROM:     Dean A. Wilson, Meteorologist
          Techniques Evaluation Section, SRAB (MD-14)

TO:       Raymond Werner, Chief
          Impact Assessment Section, Region II, 2AWM-AP


     In response to your request, Russ Lee and I have reviewed
"Valley Model Evaluation" by Chris N. Salmi, of the New Jersey
Department of Environmental Protection, with regard to alleged
errors in the Valley model.

     As noted, the Valley model has 112 built in receptors.  Mr.
Salmi incorrectly states that "the receptors are located along
the sixteen wind direction radials with seven rings."  The Valley
Model User's Guide, on page 3-8, states that the receptors "are
located approximately on 16 radials (seven receptors each) at
increments of 22.5þ angular separation....  The decimal point of
each printed concentration indicates the receptor location.  The
receptors along a given radial are not spaced equally, except due
east and west...." (the underscore is mine).  Additional emphasis
is given on page 4-13 of that User's Guide as follows:  "It is
emphasized that the only lines of receptors which are equally
spaced lie due east and due west of center; the only lines of
receptors which do not deviate somewhat from the radials are
those in the four cardinal directions."

     It should be noted that this characteristic of the receptor
array is by design, not by mistake.  The design is to permit
placing map overlays over the printouts, so that the
concentrations will appear at the locations indicated by the
decimal points on the printouts.  Mr. Salmi conducted an analysis
with receptors on 16 equally spaced radials and 7 equidistant
rings, which produced different results.  This is because his
receptors were placed at different locations than in Valley, and
not due to an error in Valley.

     Given this understanding, it seems that the only difference
between the NJDEP version of Valley and the original USEPA Valley
is that the receptors are deployed at the intersections of
radials and concentric rings in the NJDEP version.  There is 
precedent for accepting this change, since the "Valley-
equivalent" option of COMPLEX-I also deploys receptors at the
intersections of radials and concentric rings.  If, in fact, the
only difference between the NJDEP version of Valley and the USEPA
Valley is this difference in the deployment of receptors, then
the NJDEP version can be accepted for regulatory purposes.
However, to confirm this, the State should set up a test case and
perform an equivalency demonstration (with the Valley model)
pursuant to Section 3.2.2 of the "Guideline on Air Quality Models
(Revised)."

     If you have any questions, please contact Russ at FTS
629-5638.

cc:  A. Colecchia  
     D. Wilson