Surpirising output when using SOULSBY VAN RIJN in combination with TYPE OF WAVES .EQ.1
Summary
In bedload_soulsby.f, the wave orbital velocity is converted from the peak orbital velocity to the mean orbital velocity (using a division by sqrt(2).
! ************************************************* !
! VI - SPEED INDUCED BY THE CURRENT AND WAVES !
! ************************************************* !
URMS = UW%R(I)/SQRT(2.D0)
VTOT = SQRT(UNORM%R(I)**2+(0.018D0/CD)*URMS**2)
This corresponds to a multiplication with SQRT(2) done in gaia-prepare_step.f
! ! WAVE ORBITAL VELOCITY: UW ! IF(HOULE) THEN ! Forcing done inside gaia, set through the steering file IF(UW%TYPR.NE.'Q') THEN CALL CALCUW(UW%R,HN%R,HW%R,TW%R,GRAV,NPOIN,TYPE_HOULE) ELSE ! Monochromatic wave forcing: UW has just been calculated ! in CALCUW IF(TYPE_HOULE.EQ.1)THEN CONTINUE ! Coupling with TOMAWAC (irregular waves) ! UW calculated from sprectum is an Urms! ! transformation for a Jonswap spectrum (Soulsby 1993) ELSEIF(TYPE_HOULE.EQ.2)THEN DO I=1,NPOIN UW%R(I)=SQRT(2.D0)UW%R(I) ENDDO ELSE WRITE(LU,)'VALUE OF TYPE OF WAVES IS NOT OK' CALL PLANTE(1) STOP ENDIF ENDIF ENDIF
This works correctly in case TYPE_HOULE.Eq.2 (the default) and HOULE = .TRUE. However, in all other cases, there is division too much by sqrt 2, leading to an incorrect answer. So my proposition would be to change bedload_soulsby.f to:
! ************************************************* !
! VI - SPEED INDUCED BY THE CURRENT AND WAVES !
! ************************************************* !
IF (HOULE.AND.HOULE_TYPE.EQ.2) THEN
URMS = UW%R(I)/SQRT(2.D0)
ELSE
URMS = UW%R(I)
ENDIF
VTOT = SQRT(UNORM%R(I)**2+(0.018D0/CD)*URMS**2)
Please let me know what you think.