SUBROUTINE ADJUST(TEMP,Q,BETA,KMAX)
      implicit none
C     ****
C     ****     PERFORMS CONVECTIVE ADJUSTMENT ON TEMPERATURE COLUMN
C     ****       TEMP(KMAX) NEUTRAL TEMPERATURE COLUMN FOR
C     ****         CONVECTIVE ADJUSTMENT
C     ****       BETA(KMAX)
C     ****         BETA(K) = 1./PI(K)
C     ****           WHERE:
C     ****             PI(1) = (P(1))**(G/CP(1))
C     ****             PI(K) = PI(K-1)*(P(K)/P(K-1))**(G/CP(K-1))
C     ****       Q(KMAX)
C     ****         Q(1) = PI(1)*(P(1)-P(2))
C     ****         Q(K) = PI(K)*(P(K-1)-P(K+1))
C     ****
C     ****       N(KMAX), THETA(KMAX), S(KMAX) AND T(KMAX) ARE
C     ****         WORK ARRAYS
C     ****
C     ****       OUTPUT IS CORRECTED TEMP ARRAY
C     ****
!
! Args:
      integer,intent(in)  :: kmax
      real,intent(in)     :: q(kmax), beta(kmax)
      real,intent(inout)  :: temp(kmax)
!
! Local:
      integer :: k,nn,l
      real :: ss,tt,ttheta
      real :: s(kmax), t(kmax), theta(kmax)
      integer :: n(kmax)
!
    1 K = 1
      N(1) = 1
      L = 2
      THETA(1) = BETA(1)*TEMP(1)
    2 NN = 1
      TTHETA = BETA(L)*TEMP(L)
    3 IF(THETA(K).LE.TTHETA)GO TO 6
      IF(NN.GT.1)GO TO 4
      SS = Q(L)
      TT = SS*TTHETA
    4 IF(N(K).GT.1)GO TO 5
      S(K) = Q(L-NN)
      T(K) = S(K)*THETA(K)
    5 NN = NN+N(K)
      S(K) = SS+S(K)
      SS = S(K)
      T(K) = TT+T(K)
      TT = T(K)
      TTHETA = TT/SS
      IF(K.EQ.1)GO TO 7
      K = K-1
      GO TO 3
    6 K = K+1
    7 IF(L.EQ.KMAX)GO TO 8
      L = L+1
      N(K) = NN
      THETA(K) = TTHETA
      GO TO 2
    8 IF(NN.EQ.1)GO TO 11
    9 TEMP(L) = TTHETA/BETA(L)
      IF(NN.EQ.1)GO TO 11
   10 L = L-1
      NN = NN-1
      GO TO 9
   11 IF(K.EQ.1)RETURN
      K = K-1
      L = L-1
      NN = N(K)
      TTHETA = THETA(K)
      GO TO 8
      END