Question regarding wrf_rh function

[lmm_ch]

Hi,

I found the NCL source code for relative humidity calculation as follows:

c---------------------------------------------------------

C NCLFORTSTART
SUBROUTINE DCOMPUTERH(QV,P,T,RH,NX)

IMPLICIT NONE
INTEGER NX
DOUBLE PRECISION QV(NX),P(NX),T(NX),RH(NX)
C NCLEND
DOUBLE PRECISION SVP1,SVP2,SVP3,SVPT0
PARAMETER (SVP1=0.6112D0,SVP2=17.67D0,SVP3=29.65D0,SVPT0=273.15D0)
INTEGER I
DOUBLE PRECISION QVS,ES,PRESSURE,TEMPERATURE
DOUBLE PRECISION EP_2,R_D,R_V
PARAMETER (R_D=287.D0,R_V=461.6D0,EP_2=R_D/R_V)
DOUBLE PRECISION EP_3
PARAMETER (EP_3=0.622D0)

DO I = 1,NX
PRESSURE = P(I)
TEMPERATURE = T(I)
c es = 1000.*svp1*
ES = 10.D0*SVP1*EXP(SVP2* (TEMPERATURE-SVPT0)/
+ (TEMPERATURE-SVP3))
c qvs = ep_2*es/(pressure-es)
QVS = EP_3*ES/ (0.01D0*PRESSURE- (1.D0-EP_3)*ES)
c rh = 100*amax1(1., qv(i)/qvs)
c rh(i) = 100.*qv(i)/qvs
RH(I) = 100.D0*DMAX1(DMIN1(QV(I)/QVS,1.0D0),0.0D0)
END DO

RETURN
END

c----------------------------------------------

But I am confused that where does that "(1.D0-EP_3)" comes from for QVS calculation? Should it be simply QVS = EP_3*ES/ (0.01D0*PRESSURE- ES)?

Thanks.

Chris