!_EAL:MODEL_LAYER:INITIALIZATION #ifndef VERT_UNIT ! This MODULE holds the routines which are used to perform various initializations ! for the individual domains, specifically for the Eulerian, mass-based coordinate. !----------------------------------------------------------------------- MODULE module_initialize_real USE module_bc USE module_configure USE module_domain USE module_io_domain USE module_model_constants USE module_state_description USE module_timing USE module_soil_pre USE module_date_time USE module_llxy #ifdef DM_PARALLEL USE module_dm #endif REAL , SAVE :: p_top_save INTEGER :: internal_time_loop CONTAINS !------------------------------------------------------------------- SUBROUTINE init_domain ( grid ) IMPLICIT NONE ! Input space and data. No gridded meteorological data has been stored, though. ! TYPE (domain), POINTER :: grid TYPE (domain) :: grid ! Local data. INTEGER :: idum1, idum2 CALL set_scalar_indices_from_config ( head_grid%id , idum1, idum2 ) CALL init_domain_rk( grid & ! #include "actual_new_args.inc" ! ) END SUBROUTINE init_domain !------------------------------------------------------------------- SUBROUTINE init_domain_rk ( grid & ! #include "dummy_new_args.inc" ! ) USE module_optional_input IMPLICIT NONE ! Input space and data. No gridded meteorological data has been stored, though. ! TYPE (domain), POINTER :: grid TYPE (domain) :: grid #include "dummy_new_decl.inc" TYPE (grid_config_rec_type) :: config_flags ! Local domain indices and counters. INTEGER :: num_veg_cat , num_soil_top_cat , num_soil_bot_cat INTEGER :: loop , num_seaice_changes INTEGER :: ids, ide, jds, jde, kds, kde, & ims, ime, jms, jme, kms, kme, & its, ite, jts, jte, kts, kte, & ips, ipe, jps, jpe, kps, kpe, & i, j, k INTEGER :: imsx, imex, jmsx, jmex, kmsx, kmex, & ipsx, ipex, jpsx, jpex, kpsx, kpex, & imsy, imey, jmsy, jmey, kmsy, kmey, & ipsy, ipey, jpsy, jpey, kpsy, kpey INTEGER :: ns ! Local data INTEGER :: error INTEGER :: im, num_3d_m, num_3d_s REAL :: p_surf, p_level REAL :: cof1, cof2 REAL :: qvf , qvf1 , qvf2 , pd_surf REAL :: p00 , t00 , a REAL :: hold_znw LOGICAL :: were_bad LOGICAL :: stretch_grid, dry_sounding, debug INTEGER IICOUNT REAL :: p_top_requested , temp INTEGER :: num_metgrid_levels REAL , DIMENSION(max_eta) :: eta_levels REAL :: max_dz ! INTEGER , PARAMETER :: nl_max = 1000 ! REAL , DIMENSION(nl_max) :: grid%dn integer::oops1,oops2 REAL :: zap_close_levels INTEGER :: force_sfc_in_vinterp INTEGER :: interp_type , lagrange_order , extrap_type , t_extrap_type LOGICAL :: lowest_lev_from_sfc , use_levels_below_ground , use_surface LOGICAL :: we_have_tavgsfc INTEGER :: lev500 , loop_count REAL :: zl , zu , pl , pu , z500 , dz500 , tvsfc , dpmu LOGICAL , PARAMETER :: want_full_levels = .TRUE. LOGICAL , PARAMETER :: want_half_levels = .FALSE. !-- Carsel and Parrish [1988] REAL , DIMENSION(100) :: lqmi ! Dimension information stored in grid data structure. CALL get_ijk_from_grid ( grid , & ids, ide, jds, jde, kds, kde, & ims, ime, jms, jme, kms, kme, & ips, ipe, jps, jpe, kps, kpe, & imsx, imex, jmsx, jmex, kmsx, kmex, & ipsx, ipex, jpsx, jpex, kpsx, kpex, & imsy, imey, jmsy, jmey, kmsy, kmey, & ipsy, ipey, jpsy, jpey, kpsy, kpey ) its = ips ; ite = ipe ; jts = jps ; jte = jpe ; kts = kps ; kte = kpe CALL model_to_grid_config_rec ( grid%id , model_config_rec , config_flags ) ! Check to see if the boundary conditions are set properly in the namelist file. ! This checks for sufficiency and redundancy. CALL boundary_condition_check( config_flags, bdyzone, error, grid%id ) ! Some sort of "this is the first time" initialization. Who knows. grid%step_number = 0 grid%itimestep=0 ! Pull in the info in the namelist to compare it to the input data. grid%real_data_init_type = model_config_rec%real_data_init_type ! To define the base state, we call a USER MODIFIED routine to set the three ! necessary constants: p00 (sea level pressure, Pa), t00 (sea level temperature, K), ! and A (temperature difference, from 1000 mb to 300 mb, K). CALL const_module_initialize ( p00 , t00 , a ) ! Fix the snow (water equivalent depth, kg/m^2) and the snowh (physical snow ! depth, m) fields. IF ( ( flag_snow .EQ. 0 ) .AND. ( flag_snowh .EQ. 0 ) ) THEN DO j=jts,MIN(jde-1,jte) DO i=its,MIN(ide-1,ite) grid%snow(i,j) = 0. grid%snowh(i,j) = 0. END DO END DO ELSE IF ( ( flag_snow .EQ. 0 ) .AND. ( flag_snowh .EQ. 1 ) ) THEN DO j=jts,MIN(jde-1,jte) DO i=its,MIN(ide-1,ite) ! ( m -> kg/m^2 ) & ( reduce to liquid, 5:1 ratio ) grid%snow(i,j) = grid%snowh(i,j) * 1000. / 5. END DO END DO ELSE IF ( ( flag_snow .EQ. 1 ) .AND. ( flag_snowh .EQ. 0 ) ) THEN DO j=jts,MIN(jde-1,jte) DO i=its,MIN(ide-1,ite) ! ( kg/m^2 -> m) & ( liquid to snow depth, 5:1 ratio ) grid%snowh(i,j) = grid%snow(i,j) / 1000. * 5. END DO END DO END IF ! For backward compatibility, we might need to assign the map factors from ! what they were, to what they are. IF ( ( config_flags%polar ) .AND. ( flag_mf_xy .EQ. 1 ) ) THEN DO j=max(jds+1,jts),min(jde-1,jte) DO i=its,min(ide-1,ite) grid%msfvx_inv(i,j) = 1./grid%msfvx(i,j) END DO END DO IF(jts == jds) THEN DO i=its,ite grid%msfvx(i,jts) = 0. grid%msfvx_inv(i,jts) = 0. END DO END IF IF(jte == jde) THEN DO i=its,ite grid%msfvx(i,jte) = 0. grid%msfvx_inv(i,jte) = 0. END DO END IF ELSE IF ( ( config_flags%map_proj .EQ. PROJ_CASSINI ) .AND. ( flag_mf_xy .EQ. 1 ) ) THEN DO j=jts,jte DO i=its,min(ide-1,ite) grid%msfvx_inv(i,j) = 1./grid%msfvx(i,j) END DO END DO ELSE IF ( ( .NOT. config_flags%map_proj .EQ. PROJ_CASSINI ) .AND. ( flag_mf_xy .NE. 1 ) ) THEN DO j=jts,jte DO i=its,ite grid%msfvx(i,j) = grid%msfv(i,j) grid%msfvy(i,j) = grid%msfv(i,j) grid%msfux(i,j) = grid%msfu(i,j) grid%msfuy(i,j) = grid%msfu(i,j) grid%msftx(i,j) = grid%msft(i,j) grid%msfty(i,j) = grid%msft(i,j) ENDDO ENDDO DO j=jts,min(jde,jte) DO i=its,min(ide-1,ite) grid%msfvx_inv(i,j) = 1./grid%msfvx(i,j) END DO END DO ELSE IF ( ( .NOT. config_flags%map_proj .EQ. PROJ_CASSINI ) .AND. ( flag_mf_xy .EQ. 1 ) ) THEN IF ( grid%msfvx(its,jts) .EQ. 0 ) THEN CALL wrf_error_fatal ( 'Maybe you do not have the new map factors, try re-running geogrid' ) END IF DO j=jts,min(jde,jte) DO i=its,min(ide-1,ite) grid%msfvx_inv(i,j) = 1./grid%msfvx(i,j) END DO END DO ELSE IF ( ( config_flags%map_proj .EQ. PROJ_CASSINI ) .AND. ( flag_mf_xy .NE. 1 ) ) THEN CALL wrf_error_fatal ( 'Neither SI data nor older metgrid data can initialize a global domain' ) ENDIF ! Is there any vertical interpolation to do? The "old" data comes in on the correct ! vertical locations already. IF ( flag_metgrid .EQ. 1 ) THEN ! <----- START OF VERTICAL INTERPOLATION PART ----> ! Variables that are named differently between SI and WPS. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%tsk(i,j) = grid%tsk_gc(i,j) grid%tmn(i,j) = grid%tmn_gc(i,j) grid%xlat(i,j) = grid%xlat_gc(i,j) grid%xlong(i,j) = grid%xlong_gc(i,j) grid%ht(i,j) = grid%ht_gc(i,j) END DO END DO ! A user could request that the most coarse grid has the ! topography along the outer boundary smoothed. This smoothing ! is similar to the coarse/nest interface. The outer rows and ! cols come from the existing large scale topo, and then the ! next several rows/cols are a linear ramp of the large scale ! model and the hi-res topo from WPS. We only do this for the ! coarse grid since we are going to make the interface consistent ! in the model betwixt the CG and FG domains. IF ( ( config_flags%smooth_cg_topo ) .AND. & ( grid%id .EQ. 1 ) .AND. & ( flag_soilhgt .EQ. 1) ) THEN CALL blend_terrain ( grid%toposoil , grid%ht , & ids , ide , jds , jde , 1 , 1 , & ims , ime , jms , jme , 1 , 1 , & ips , ipe , jps , jpe , 1 , 1 ) END IF ! Filter the input topography if this is a polar projection. IF ( config_flags%map_proj .EQ. PROJ_CASSINI ) THEN #if ( defined( DM_PARALLEL ) && ( ! defined( STUBMPI ) ) ) ! We stick the topo and map fac in an unused 3d array. The map scale ! factor and computational latitude are passed along for the ride ! (part of the transpose process - we only do 3d arrays) to determine ! "how many" values are used to compute the mean. We want a number ! that is consistent with the original grid resolution. DO j = jts, MIN(jte,jde-1) DO k = kts, kte DO i = its, MIN(ite,ide-1) grid%t_init(i,k,j) = 1. END DO END DO DO i = its, MIN(ite,ide-1) grid%t_init(i,1,j) = grid%ht(i,j) grid%t_init(i,2,j) = grid%msftx(i,j) grid%t_init(i,3,j) = grid%clat(i,j) END DO END DO # include "XPOSE_POLAR_FILTER_TOPO_z2x.inc" ! Retrieve the 2d arrays for topo, map factors, and the ! computational latitude. DO j = jpsx, MIN(jpex,jde-1) DO i = ipsx, MIN(ipex,ide-1) grid%ht_xxx(i,j) = grid%t_xxx(i,1,j) grid%mf_xxx(i,j) = grid%t_xxx(i,2,j) grid%clat_xxx(i,j) = grid%t_xxx(i,3,j) END DO END DO ! Get a mean topo field that is consistent with the grid ! distance on each computational latitude loop. CALL filter_topo ( grid%ht_xxx , grid%clat_xxx , grid%mf_xxx , & grid%fft_filter_lat , & ids, ide, jds, jde, 1 , 1 , & imsx, imex, jmsx, jmex, 1, 1, & ipsx, ipex, jpsx, jpex, 1, 1 ) ! Stick the filtered topo back into the dummy 3d array to ! transpose it back to "all z on a patch". DO j = jpsx, MIN(jpex,jde-1) DO i = ipsx, MIN(ipex,ide-1) grid%t_xxx(i,1,j) = grid%ht_xxx(i,j) END DO END DO # include "XPOSE_POLAR_FILTER_TOPO_x2z.inc" ! Get the un-transposed topo data. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%ht(i,j) = grid%t_init(i,1,j) END DO END DO #else CALL filter_topo ( grid%ht , grid%clat , grid%msftx , & grid%fft_filter_lat , & ids, ide, jds, jde, 1,1, & ims, ime, jms, jme, 1,1, & its, ite, jts, jte, 1,1 ) #endif END IF ! If we have any input low-res surface pressure, we store it. IF ( flag_psfc .EQ. 1 ) THEN DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%psfc_gc(i,j) = grid%psfc(i,j) grid%p_gc(i,1,j) = grid%psfc(i,j) END DO END DO END IF ! If we have the low-resolution surface elevation, stick that in the ! "input" locations of the 3d height. We still have the "hi-res" topo ! stuck in the grid%ht array. The grid%landmask if test is required as some sources ! have ZERO elevation over water (thank you very much). IF ( flag_soilhgt .EQ. 1) THEN DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) ! IF ( grid%landmask(i,j) .GT. 0.5 ) THEN grid%ght_gc(i,1,j) = grid%toposoil(i,j) grid%ht_gc(i,j)= grid%toposoil(i,j) ! END IF END DO END DO END IF ! Assign surface fields with original input values. If this is hybrid data, ! the values are not exactly representative. However - this is only for ! plotting purposes and such at the 0h of the forecast, so we are not all that ! worried. DO j = jts, min(jde-1,jte) DO i = its, min(ide,ite) grid%u10(i,j)=grid%u_gc(i,1,j) END DO END DO DO j = jts, min(jde,jte) DO i = its, min(ide-1,ite) grid%v10(i,j)=grid%v_gc(i,1,j) END DO END DO DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) grid%t2(i,j)=grid%t_gc(i,1,j) END DO END DO IF ( flag_qv .EQ. 1 ) THEN DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) grid%q2(i,j)=grid%qv_gc(i,1,j) END DO END DO END IF ! The number of vertical levels in the input data. There is no staggering for ! different variables. num_metgrid_levels = grid%num_metgrid_levels ! The requested ptop for real data cases. p_top_requested = grid%p_top_requested ! Compute the top pressure, grid%p_top. For isobaric data, this is just the ! top level. For the generalized vertical coordinate data, we find the ! max pressure on the top level. We have to be careful of two things: ! 1) the value has to be communicated, 2) the value can not increase ! at subsequent times from the initial value. IF ( internal_time_loop .EQ. 1 ) THEN CALL find_p_top ( grid%p_gc , grid%p_top , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) #if ( defined( DM_PARALLEL ) && ( ! defined( STUBMPI ) ) ) grid%p_top = wrf_dm_max_real ( grid%p_top ) #endif ! Compare the requested grid%p_top with the value available from the input data. IF ( p_top_requested .LT. grid%p_top ) THEN print *,'p_top_requested = ',p_top_requested print *,'allowable grid%p_top in data = ',grid%p_top CALL wrf_error_fatal ( 'p_top_requested < grid%p_top possible from data' ) END IF ! The grid%p_top valus is the max of what is available from the data and the ! requested value. We have already compared <, so grid%p_top is directly set to ! the value in the namelist. grid%p_top = p_top_requested ! For subsequent times, we have to remember what the grid%p_top for the first ! time was. Why? If we have a generalized vert coordinate, the grid%p_top value ! could fluctuate. p_top_save = grid%p_top ELSE CALL find_p_top ( grid%p_gc , grid%p_top , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) #if ( defined( DM_PARALLEL ) && ( ! defined( STUBMPI ) ) ) grid%p_top = wrf_dm_max_real ( grid%p_top ) #endif IF ( grid%p_top .GT. p_top_save ) THEN print *,'grid%p_top from last time period = ',p_top_save print *,'grid%p_top from this time period = ',grid%p_top CALL wrf_error_fatal ( 'grid%p_top > previous value' ) END IF grid%p_top = p_top_save ENDIF ! Get the monthly values interpolated to the current date for the traditional monthly ! fields of green-ness fraction and background albedo. CALL monthly_interp_to_date ( grid%greenfrac , current_date , grid%vegfra , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) CALL monthly_interp_to_date ( grid%albedo12m , current_date , grid%albbck , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Get the min/max of each i,j for the monthly green-ness fraction. CALL monthly_min_max ( grid%greenfrac , grid%shdmin , grid%shdmax , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! The model expects the green-ness values in percent, not fraction. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%vegfra(i,j) = grid%vegfra(i,j) * 100. grid%shdmax(i,j) = grid%shdmax(i,j) * 100. grid%shdmin(i,j) = grid%shdmin(i,j) * 100. END DO END DO ! The model expects the albedo fields as a fraction, not a percent. Set the ! water values to 8%. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%albbck(i,j) = grid%albbck(i,j) / 100. grid%snoalb(i,j) = grid%snoalb(i,j) / 100. IF ( grid%landmask(i,j) .LT. 0.5 ) THEN grid%albbck(i,j) = 0.08 grid%snoalb(i,j) = 0.08 END IF END DO END DO ! Compute the mixing ratio from the input relative humidity. IF ( flag_qv .NE. 1 ) THEN CALL rh_to_mxrat (grid%rh_gc, grid%t_gc, grid%p_gc, grid%qv_gc , .TRUE. , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) END IF ! Two ways to get the surface pressure. 1) If we have the low-res input surface ! pressure and the low-res topography, then we can do a simple hydrostatic ! relation. 2) Otherwise we compute the surface pressure from the sea-level ! pressure. ! Note that on output, grid%psfc is now hi-res. The low-res surface pressure and ! elevation are grid%psfc_gc and grid%ht_gc (same as grid%ght_gc(k=1)). IF ( flag_tavgsfc .EQ. 1 ) THEN we_have_tavgsfc = .TRUE. ELSE we_have_tavgsfc = .FALSE. END IF IF ( ( flag_psfc .EQ. 1 ) .AND. & ( flag_soilhgt .EQ. 1 ) .AND. & ( flag_slp .EQ. 1 ) .AND. & ( .NOT. config_flags%sfcp_to_sfcp ) ) THEN CALL sfcprs3(grid%ght_gc, grid%p_gc, grid%ht, & grid%pslv_gc, grid%psfc, & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) ELSE IF ( ( flag_psfc .EQ. 1 ) .AND. & ( flag_soilhgt .EQ. 1 ) .AND. & ( config_flags%sfcp_to_sfcp ) ) THEN CALL sfcprs2(grid%t_gc, grid%qv_gc, grid%ght_gc, grid%psfc_gc, grid%ht, & grid%tavgsfc, grid%p_gc, grid%psfc, we_have_tavgsfc, & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) ELSE IF ( flag_slp .EQ. 1 ) THEN CALL sfcprs (grid%t_gc, grid%qv_gc, grid%ght_gc, grid%pslv_gc, grid%ht, & grid%tavgsfc, grid%p_gc, grid%psfc, we_have_tavgsfc, & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) ELSE CALL wrf_error_fatal ( 'not enough info for a p sfc computation' ) END IF ! If we have no input surface pressure, we'd better stick something in there. IF ( flag_psfc .NE. 1 ) THEN DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%psfc_gc(i,j) = grid%psfc(i,j) grid%p_gc(i,1,j) = grid%psfc(i,j) END DO END DO END IF ! Integrate the mixing ratio to get the vapor pressure. CALL integ_moist ( grid%qv_gc , grid%p_gc , grid%pd_gc , grid%t_gc , grid%ght_gc , grid%intq_gc , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) ! Compute the difference between the dry, total surface pressure (input) and the ! dry top pressure (constant). CALL p_dts ( grid%mu0 , grid%intq_gc , grid%psfc , grid%p_top , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) ! Compute the dry, hydrostatic surface pressure. CALL p_dhs ( grid%pdhs , grid%ht , p00 , t00 , a , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute the eta levels if not defined already. IF ( grid%znw(1) .NE. 1.0 ) THEN eta_levels(1:kde) = model_config_rec%eta_levels(1:kde) max_dz = model_config_rec%max_dz CALL compute_eta ( grid%znw , & eta_levels , max_eta , max_dz , & grid%p_top , g , p00 , cvpm , a , r_d , cp , t00 , p1000mb , t0 , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF ! The input field is temperature, we want potential temp. CALL t_to_theta ( grid%t_gc , grid%p_gc , p00 , & ids , ide , jds , jde , 1 , num_metgrid_levels , & ims , ime , jms , jme , 1 , num_metgrid_levels , & its , ite , jts , jte , 1 , num_metgrid_levels ) IF ( flag_slp .EQ. 1 ) THEN ! On the eta surfaces, compute the dry pressure = mu eta, stored in ! grid%pb, since it is a pressure, and we don't need another kms:kme 3d ! array floating around. The grid%pb array is re-computed as the base pressure ! later after the vertical interpolations are complete. CALL p_dry ( grid%mu0 , grid%znw , grid%p_top , grid%pb , want_full_levels , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! All of the vertical interpolations are done in dry-pressure space. The ! input data has had the moisture removed (grid%pd_gc). The target levels (grid%pb) ! had the vapor pressure removed from the surface pressure, then they were ! scaled by the eta levels. interp_type = 2 lagrange_order = grid%lagrange_order lowest_lev_from_sfc = .FALSE. use_levels_below_ground = .TRUE. use_surface = .TRUE. zap_close_levels = grid%zap_close_levels force_sfc_in_vinterp = 0 t_extrap_type = grid%t_extrap_type extrap_type = 1 ! For the height field, the lowest level pressure is the slp (approximately "dry"). The ! lowest level of the input height field (to be associated with slp) then is an array ! of zeros. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%psfc_gc(i,j) = grid%pd_gc(i,1,j) grid%pd_gc(i,1,j) = grid%pslv_gc(i,j) - ( grid%p_gc(i,1,j) - grid%pd_gc(i,1,j) ) grid%ht_gc(i,j) = grid%ght_gc(i,1,j) grid%ght_gc(i,1,j) = 0. END DO END DO CALL vert_interp ( grid%ght_gc , grid%pd_gc , grid%ph0 , grid%pb , & num_metgrid_levels , 'Z' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Put things back to normal. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%pd_gc(i,1,j) = grid%psfc_gc(i,j) grid%ght_gc(i,1,j) = grid%ht_gc(i,j) END DO END DO END IF ! Now the rest of the variables on half-levels to inteprolate. CALL p_dry ( grid%mu0 , grid%znw , grid%p_top , grid%pb , want_half_levels , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) interp_type = grid%interp_type lagrange_order = grid%lagrange_order lowest_lev_from_sfc = grid%lowest_lev_from_sfc use_levels_below_ground = grid%use_levels_below_ground use_surface = grid%use_surface zap_close_levels = grid%zap_close_levels force_sfc_in_vinterp = grid%force_sfc_in_vinterp t_extrap_type = grid%t_extrap_type extrap_type = grid%extrap_type CALL vert_interp ( grid%qv_gc , grid%pd_gc , moist(:,:,:,P_QV) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) CALL vert_interp ( grid%t_gc , grid%pd_gc , grid%t_2 , grid%pb , & num_metgrid_levels , 'T' , & interp_type , lagrange_order , t_extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) #ifdef RUC_CLOUD ! Add -DRUC_CLOUD to ARCHFLAGS in configure.wrf file to activate the following code num_3d_m = num_moist num_3d_s = num_scalar IF ( flag_qr .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_m IF ( im .EQ. P_QR ) THEN CALL vert_interp ( grid%qr_gc , grid%pd_gc , moist(:,:,:,P_QR) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF IF ( flag_qc .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_m IF ( im .EQ. P_QC ) THEN CALL vert_interp ( grid%qc_gc , grid%pd_gc , moist(:,:,:,P_QC) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF IF ( flag_qi .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_m IF ( im .EQ. P_QI ) THEN CALL vert_interp ( grid%qi_gc , grid%pd_gc , moist(:,:,:,P_QI) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF IF ( flag_qs .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_m IF ( im .EQ. P_QS ) THEN CALL vert_interp ( grid%qs_gc , grid%pd_gc , moist(:,:,:,P_QS) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF IF ( flag_qg .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_m IF ( im .EQ. P_QG ) THEN CALL vert_interp ( grid%qg_gc , grid%pd_gc , moist(:,:,:,P_QG) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF IF ( flag_qni .EQ. 1 ) THEN DO im = PARAM_FIRST_SCALAR, num_3d_s IF ( im .EQ. P_QNI ) THEN CALL vert_interp ( grid%qni_gc , grid%pd_gc , scalar(:,:,:,P_QNI) , grid%pb , & num_metgrid_levels , 'Q' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF END DO END IF #endif #ifdef DM_PARALLEL ips = its ; ipe = ite ; jps = jts ; jpe = jte ; kps = kts ; kpe = kte ! For the U and V vertical interpolation, we need the pressure defined ! at both the locations for the horizontal momentum, which we get by ! averaging two pressure values (i and i-1 for U, j and j-1 for V). The ! pressure field on input (grid%pd_gc) and the pressure of the new coordinate ! (grid%pb) are both communicated with an 8 stencil. # include "HALO_EM_VINTERP_UV_1.inc" #endif CALL vert_interp ( grid%u_gc , grid%pd_gc , grid%u_2 , grid%pb , & num_metgrid_levels , 'U' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) CALL vert_interp ( grid%v_gc , grid%pd_gc , grid%v_2 , grid%pb , & num_metgrid_levels , 'V' , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END IF ! <----- END OF VERTICAL INTERPOLATION PART ----> ! Save the grid%tsk field for later use in the sea ice surface temperature ! for the Noah LSM scheme. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%tsk_save(i,j) = grid%tsk(i,j) END DO END DO ! Protect against bad grid%tsk values over water by supplying grid%sst (if it is ! available, and if the grid%sst is reasonable). DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( ( grid%landmask(i,j) .LT. 0.5 ) .AND. ( flag_sst .EQ. 1 ) .AND. & ( grid%sst(i,j) .GT. 170. ) .AND. ( grid%sst(i,j) .LT. 400. ) ) THEN grid%tsk(i,j) = grid%sst(i,j) ENDIF END DO END DO ! Take the data from the input file and store it in the variables that ! use the WRF naming and ordering conventions. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) IF ( grid%snow(i,j) .GE. 10. ) then grid%snowc(i,j) = 1. ELSE grid%snowc(i,j) = 0.0 END IF END DO END DO ! Set flag integers for presence of snowh and soilw fields grid%ifndsnowh = flag_snowh IF (num_sw_levels_input .GE. 1) THEN grid%ifndsoilw = 1 ELSE grid%ifndsoilw = 0 END IF ! We require input data for the various LSM schemes. enough_data : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE (LSMSCHEME) IF ( num_st_levels_input .LT. 2 ) THEN CALL wrf_error_fatal ( 'Not enough soil temperature data for Noah LSM scheme.') END IF CASE (RUCLSMSCHEME) IF ( num_st_levels_input .LT. 2 ) THEN CALL wrf_error_fatal ( 'Not enough soil temperature data for RUC LSM scheme.') END IF CASE (PXLSMSCHEME) IF ( num_st_levels_input .LT. 2 ) THEN CALL wrf_error_fatal ( 'Not enough soil temperature data for P-X LSM scheme.') END IF END SELECT enough_data ! For sf_surface_physics = 1, we want to use close to a 30 cm value ! for the bottom level of the soil temps. fix_bottom_level_for_temp : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE (SLABSCHEME) IF ( flag_tavgsfc .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%tavgsfc(i,j) END DO END DO ELSE IF ( flag_st010040 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st010040(i,j) END DO END DO ELSE IF ( flag_st000010 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st000010(i,j) END DO END DO ELSE IF ( flag_soilt020 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%soilt020(i,j) END DO END DO ELSE IF ( flag_st007028 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st007028(i,j) END DO END DO ELSE CALL wrf_debug ( 0 , 'No 10-40 cm, 0-10 cm, 7-28, or 20 cm soil temperature data for grid%tmn') CALL wrf_debug ( 0 , 'Using 1 degree static annual mean temps' ) END IF CASE (LSMSCHEME) CASE (RUCLSMSCHEME) CASE (PXLSMSCHEME) IF ( flag_tavgsfc .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%tavgsfc(i,j) END DO END DO ELSE IF ( flag_st010040 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st010040(i,j) END DO END DO ELSE IF ( flag_st040100 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st040100(i,j) END DO END DO ELSE IF ( flag_st100200 .EQ. 1 ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) = grid%st100200(i,j) END DO END DO ELSE CALL wrf_debug ( 0 , 'No 10-40 cm or 40-100 cm soil temperature data for grid%em_tmn') CALL wrf_debug ( 0 , 'Using 1 degree static annual mean temps' ) END IF DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%tmn(i,j) =(10 * grid%st000010(i,j) + 30 * grid%st010040(i,j) + & 60 * grid%st040100(i,j) + 100* grid%st100200(i,j) )/200 grid%tmn(i,j) = grid%st010040(i,j) !grid%tmn(i,j) = grid%st040100(i,j) !grid%tmn(i,j) = grid%st100200(i,j) END DO END DO END SELECT fix_bottom_level_for_temp ! Adjustments for the seaice field PRIOR to the grid%tslb computations. This is ! is for the 5-layer scheme. num_veg_cat = SIZE ( grid%landusef , DIM=2 ) num_soil_top_cat = SIZE ( grid%soilctop , DIM=2 ) num_soil_bot_cat = SIZE ( grid%soilcbot , DIM=2 ) CALL nl_get_seaice_threshold ( grid%id , grid%seaice_threshold ) CALL nl_get_isice ( grid%id , grid%isice ) CALL nl_get_iswater ( grid%id , grid%iswater ) CALL adjust_for_seaice_pre ( grid%xice , grid%landmask , grid%tsk , grid%ivgtyp , grid%vegcat , grid%lu_index , & grid%xland , grid%landusef , grid%isltyp , grid%soilcat , grid%soilctop , & grid%soilcbot , grid%tmn , & grid%seaice_threshold , & num_veg_cat , num_soil_top_cat , num_soil_bot_cat , & grid%iswater , grid%isice , & model_config_rec%sf_surface_physics(grid%id) , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! surface_input_source=1 => use data from static file (fractional category as input) ! surface_input_source=2 => use data from grib file (dominant category as input) IF ( config_flags%surface_input_source .EQ. 1 ) THEN grid%vegcat (its,jts) = 0 grid%soilcat(its,jts) = 0 END IF ! Generate the vegetation and soil category information from the fractional input ! data, or use the existing dominant category fields if they exist. IF ( ( grid%soilcat(its,jts) .LT. 0.5 ) .AND. ( grid%vegcat(its,jts) .LT. 0.5 ) ) THEN num_veg_cat = SIZE ( grid%landusef , DIM=2 ) num_soil_top_cat = SIZE ( grid%soilctop , DIM=2 ) num_soil_bot_cat = SIZE ( grid%soilcbot , DIM=2 ) CALL process_percent_cat_new ( grid%landmask , & grid%landusef , grid%soilctop , grid%soilcbot , & grid%isltyp , grid%ivgtyp , & num_veg_cat , num_soil_top_cat , num_soil_bot_cat , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte , & model_config_rec%iswater(grid%id) ) ! Make all the veg/soil parms the same so as not to confuse the developer. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) grid%vegcat(i,j) = grid%ivgtyp(i,j) grid%soilcat(i,j) = grid%isltyp(i,j) END DO END DO ELSE ! Do we have dominant soil and veg data from the input already? IF ( grid%soilcat(its,jts) .GT. 0.5 ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%isltyp(i,j) = NINT( grid%soilcat(i,j) ) END DO END DO END IF IF ( grid%vegcat(its,jts) .GT. 0.5 ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%ivgtyp(i,j) = NINT( grid%vegcat(i,j) ) END DO END DO END IF END IF ! Land use assignment. DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%lu_index(i,j) = grid%ivgtyp(i,j) IF ( grid%lu_index(i,j) .NE. model_config_rec%iswater(grid%id) ) THEN grid%landmask(i,j) = 1 grid%xland(i,j) = 1 ELSE grid%landmask(i,j) = 0 grid%xland(i,j) = 2 END IF END DO END DO ! Adjust the various soil temperature values depending on the difference in ! in elevation between the current model's elevation and the incoming data's ! orography. adjust_soil : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE ( SLABSCHEME , LSMSCHEME , RUCLSMSCHEME, PXLSMSCHEME ) CALL adjust_soil_temp_new ( grid%tmn , model_config_rec%sf_surface_physics(grid%id) , & grid%tsk , grid%ht , grid%toposoil , grid%landmask , flag_soilhgt , flag_tavgsfc , & grid%st000010 , grid%st010040 , grid%st040100 , grid%st100200 , grid%st010200 , & flag_st000010 , flag_st010040 , flag_st040100 , flag_st100200 , flag_st010200 , & grid%st000007 , grid%st007028 , grid%st028100 , grid%st100255 , & flag_st000007 , flag_st007028 , flag_st028100 , flag_st100255 , & grid%soilt000 , grid%soilt005 , grid%soilt020 , grid%soilt040 , grid%soilt160 , & grid%soilt300 , & flag_soilt000 , flag_soilt005 , flag_soilt020 , flag_soilt040 , & flag_soilt160 , flag_soilt300 , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) END SELECT adjust_soil ! Fix grid%tmn and grid%tsk. fix_tsk_tmn : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE ( SLABSCHEME , LSMSCHEME , RUCLSMSCHEME, PXLSMSCHEME ) DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( ( grid%landmask(i,j) .LT. 0.5 ) .AND. ( flag_sst .EQ. 1 ) .AND. & ( grid%sst(i,j) .GT. 170. ) .AND. ( grid%sst(i,j) .LT. 400. ) ) THEN grid%tmn(i,j) = grid%sst(i,j) grid%tsk(i,j) = grid%sst(i,j) ELSE IF ( grid%landmask(i,j) .LT. 0.5 ) THEN grid%tmn(i,j) = grid%tsk(i,j) END IF END DO END DO END SELECT fix_tsk_tmn ! Is the grid%tsk reasonable? IF ( internal_time_loop .NE. 1 ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( grid%tsk(i,j) .LT. 170 .or. grid%tsk(i,j) .GT. 400. ) THEN grid%tsk(i,j) = grid%t_2(i,1,j) END IF END DO END DO ELSE DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( grid%tsk(i,j) .LT. 170 .or. grid%tsk(i,j) .GT. 400. ) THEN print *,'error in the grid%tsk' print *,'i,j=',i,j print *,'grid%landmask=',grid%landmask(i,j) print *,'grid%tsk, grid%sst, grid%tmn=',grid%tsk(i,j),grid%sst(i,j),grid%tmn(i,j) if(grid%tmn(i,j).gt.170. .and. grid%tmn(i,j).lt.400.)then grid%tsk(i,j)=grid%tmn(i,j) else if(grid%sst(i,j).gt.170. .and. grid%sst(i,j).lt.400.)then grid%tsk(i,j)=grid%sst(i,j) else CALL wrf_error_fatal ( 'grid%tsk unreasonable' ) end if END IF END DO END DO END IF ! Is the grid%tmn reasonable? DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( ( ( grid%tmn(i,j) .LT. 170. ) .OR. ( grid%tmn(i,j) .GT. 400. ) ) & .AND. ( grid%landmask(i,j) .GT. 0.5 ) ) THEN IF ( model_config_rec%sf_surface_physics(grid%id) .NE. LSMSCHEME ) THEN print *,'error in the grid%tmn' print *,'i,j=',i,j print *,'grid%landmask=',grid%landmask(i,j) print *,'grid%tsk, grid%sst, grid%tmn=',grid%tsk(i,j),grid%sst(i,j),grid%tmn(i,j) END IF if(grid%tsk(i,j).gt.170. .and. grid%tsk(i,j).lt.400.)then grid%tmn(i,j)=grid%tsk(i,j) else if(grid%sst(i,j).gt.170. .and. grid%sst(i,j).lt.400.)then grid%tmn(i,j)=grid%sst(i,j) else CALL wrf_error_fatal ( 'grid%tmn unreasonable' ) endif END IF END DO END DO interpolate_soil_tmw : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE ( SLABSCHEME , LSMSCHEME , RUCLSMSCHEME, PXLSMSCHEME ) CALL process_soil_real ( grid%tsk , grid%tmn , & grid%landmask , grid%sst , & st_input , sm_input , sw_input , st_levels_input , sm_levels_input , sw_levels_input , & grid%zs , grid%dzs , grid%tslb , grid%smois , grid%sh2o , & flag_sst , flag_soilt000, flag_soilm000, & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte , & model_config_rec%sf_surface_physics(grid%id) , & model_config_rec%num_soil_layers , & model_config_rec%real_data_init_type , & num_st_levels_input , num_sm_levels_input , num_sw_levels_input , & num_st_levels_alloc , num_sm_levels_alloc , num_sw_levels_alloc ) END SELECT interpolate_soil_tmw ! Minimum soil values, residual, from RUC LSM scheme. For input from Noah or EC, and using ! RUC LSM scheme, this must be subtracted from the input total soil moisture. For ! input RUC data and using the Noah LSM scheme, this value must be added to the soil ! moisture input. lqmi(1:num_soil_top_cat) = & (/0.045, 0.057, 0.065, 0.067, 0.034, 0.078, 0.10, & 0.089, 0.095, 0.10, 0.070, 0.068, 0.078, 0.0, & 0.004, 0.065 /) ! 0.004, 0.065, 0.020, 0.004, 0.008 /) ! has extra levels for playa, lava, and white sand ! At the initial time we care about values of soil moisture and temperature, other times are ! ignored by the model, so we ignore them, too. IF ( domain_ClockIsStartTime(grid) ) THEN account_for_zero_soil_moisture : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE ( LSMSCHEME ) iicount = 0 IF ( ( FLAG_SM000010 .EQ. 1 ) .OR. ( FLAG_SM000007 .EQ. 1 ) ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( (grid%landmask(i,j).gt.0.5) .and. ( grid%tslb(i,1,j) .gt. 170 ) .and. & ( grid%tslb(i,1,j) .lt. 400 ) .and. ( grid%smois(i,1,j) .lt. 0.005 ) ) then print *,'Noah -> Noah: bad soil moisture at i,j = ',i,j,grid%smois(i,:,j) iicount = iicount + 1 grid%smois(i,:,j) = 0.005 END IF END DO END DO IF ( iicount .GT. 0 ) THEN print *,'Noah -> Noah: total number of small soil moisture locations = ',iicount END IF ELSE IF ( FLAG_SOILM000 .EQ. 1 ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%smois(i,:,j) = grid%smois(i,:,j) + lqmi(grid%isltyp(i,j)) END DO END DO DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( (grid%landmask(i,j).gt.0.5) .and. ( grid%tslb(i,1,j) .gt. 170 ) .and. & ( grid%tslb(i,1,j) .lt. 400 ) .and. ( grid%smois(i,1,j) .lt. 0.005 ) ) then print *,'RUC -> Noah: bad soil moisture at i,j = ',i,j,grid%smois(i,:,j) iicount = iicount + 1 grid%smois(i,:,j) = 0.005 END IF END DO END DO IF ( iicount .GT. 0 ) THEN print *,'RUC -> Noah: total number of small soil moisture locations = ',iicount END IF END IF CASE ( RUCLSMSCHEME ) iicount = 0 IF ( ( FLAG_SM000010 .EQ. 1 ) .OR. ( FLAG_SM000007 .EQ. 1 ) ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%smois(i,:,j) = MAX ( grid%smois(i,:,j) - lqmi(grid%isltyp(i,j)) , 0. ) END DO END DO ELSE IF ( FLAG_SOILM000 .EQ. 1 ) THEN ! no op END IF CASE ( PXLSMSCHEME ) iicount = 0 IF ( FLAG_SM000010 .EQ. 1 ) THEN DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) grid%smois(i,:,j) = MAX ( grid%smois(i,:,j) - lqmi(grid%isltyp(i,j)) , 0. ) END DO END DO ELSE IF ( FLAG_SOILM000 .EQ. 1 ) THEN ! no op END IF END SELECT account_for_zero_soil_moisture END IF ! Is the grid%tslb reasonable? IF ( internal_time_loop .NE. 1 ) THEN DO j = jts, MIN(jde-1,jte) DO ns = 1 , model_config_rec%num_soil_layers DO i = its, MIN(ide-1,ite) IF ( grid%tslb(i,ns,j) .LT. 170 .or. grid%tslb(i,ns,j) .GT. 400. ) THEN grid%tslb(i,ns,j) = grid%t_2(i,1,j) grid%smois(i,ns,j) = 0.3 END IF END DO END DO END DO ELSE DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( ( ( grid%tslb(i,1,j) .LT. 170. ) .OR. ( grid%tslb(i,1,j) .GT. 400. ) ) .AND. & ( grid%landmask(i,j) .GT. 0.5 ) ) THEN IF ( ( model_config_rec%sf_surface_physics(grid%id) .NE. LSMSCHEME ) .AND. & ( model_config_rec%sf_surface_physics(grid%id) .NE. RUCLSMSCHEME ).AND. & ( model_config_rec%sf_surface_physics(grid%id) .NE. PXLSMSCHEME ) ) THEN print *,'error in the grid%tslb' print *,'i,j=',i,j print *,'grid%landmask=',grid%landmask(i,j) print *,'grid%tsk, grid%sst, grid%tmn=',grid%tsk(i,j),grid%sst(i,j),grid%tmn(i,j) print *,'grid%tslb = ',grid%tslb(i,:,j) print *,'old grid%smois = ',grid%smois(i,:,j) grid%smois(i,1,j) = 0.3 grid%smois(i,2,j) = 0.3 grid%smois(i,3,j) = 0.3 grid%smois(i,4,j) = 0.3 END IF IF ( (grid%tsk(i,j).GT.170. .AND. grid%tsk(i,j).LT.400.) .AND. & (grid%tmn(i,j).GT.170. .AND. grid%tmn(i,j).LT.400.) ) THEN fake_soil_temp : SELECT CASE ( model_config_rec%sf_surface_physics(grid%id) ) CASE ( SLABSCHEME ) DO ns = 1 , model_config_rec%num_soil_layers grid%tslb(i,ns,j) = ( grid%tsk(i,j)*(3.0 - grid%zs(ns)) + & grid%tmn(i,j)*(0.0 - grid%zs(ns)) ) /(3.0 - 0.0) END DO CASE ( LSMSCHEME , RUCLSMSCHEME, PXLSMSCHEME ) CALL wrf_error_fatal ( 'Assigning constant soil moisture, bad idea') DO ns = 1 , model_config_rec%num_soil_layers grid%tslb(i,ns,j) = ( grid%tsk(i,j)*(3.0 - grid%zs(ns)) + & grid%tmn(i,j)*(0.0 - grid%zs(ns)) ) /(3.0 - 0.0) END DO END SELECT fake_soil_temp else if(grid%tsk(i,j).gt.170. .and. grid%tsk(i,j).lt.400.)then CALL wrf_error_fatal ( 'grid%tslb unreasonable 1' ) DO ns = 1 , model_config_rec%num_soil_layers grid%tslb(i,ns,j)=grid%tsk(i,j) END DO else if(grid%sst(i,j).gt.170. .and. grid%sst(i,j).lt.400.)then CALL wrf_error_fatal ( 'grid%tslb unreasonable 2' ) DO ns = 1 , model_config_rec%num_soil_layers grid%tslb(i,ns,j)=grid%sst(i,j) END DO else if(grid%tmn(i,j).gt.170. .and. grid%tmn(i,j).lt.400.)then CALL wrf_error_fatal ( 'grid%tslb unreasonable 3' ) DO ns = 1 , model_config_rec%num_soil_layers grid%tslb(i,ns,j)=grid%tmn(i,j) END DO else CALL wrf_error_fatal ( 'grid%tslb unreasonable 4' ) endif END IF END DO END DO END IF ! Adjustments for the seaice field AFTER the grid%tslb computations. This is ! is for the Noah LSM scheme. num_veg_cat = SIZE ( grid%landusef , DIM=2 ) num_soil_top_cat = SIZE ( grid%soilctop , DIM=2 ) num_soil_bot_cat = SIZE ( grid%soilcbot , DIM=2 ) CALL nl_get_seaice_threshold ( grid%id , grid%seaice_threshold ) CALL nl_get_isice ( grid%id , grid%isice ) CALL nl_get_iswater ( grid%id , grid%iswater ) CALL adjust_for_seaice_post ( grid%xice , grid%landmask , grid%tsk , grid%tsk_save , & grid%ivgtyp , grid%vegcat , grid%lu_index , & grid%xland , grid%landusef , grid%isltyp , grid%soilcat , & grid%soilctop , & grid%soilcbot , grid%tmn , grid%vegfra , & grid%tslb , grid%smois , grid%sh2o , & grid%seaice_threshold , & num_veg_cat , num_soil_top_cat , num_soil_bot_cat , & model_config_rec%num_soil_layers , & grid%iswater , grid%isice , & model_config_rec%sf_surface_physics(grid%id) , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Let us make sure (again) that the grid%landmask and the veg/soil categories match. oops1=0 oops2=0 DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( ( ( grid%landmask(i,j) .LT. 0.5 ) .AND. & ( grid%ivgtyp(i,j) .NE. config_flags%iswater .OR. grid%isltyp(i,j) .NE. 14 ) ) .OR. & ( ( grid%landmask(i,j) .GT. 0.5 ) .AND. & ( grid%ivgtyp(i,j) .EQ. config_flags%iswater .OR. grid%isltyp(i,j) .EQ. 14 ) ) ) THEN IF ( grid%tslb(i,1,j) .GT. 1. ) THEN oops1=oops1+1 grid%ivgtyp(i,j) = 5 grid%isltyp(i,j) = 8 grid%landmask(i,j) = 1 grid%xland(i,j) = 1 ELSE IF ( grid%sst(i,j) .GT. 1. ) THEN oops2=oops2+1 grid%ivgtyp(i,j) = config_flags%iswater grid%isltyp(i,j) = 14 grid%landmask(i,j) = 0 grid%xland(i,j) = 2 ELSE print *,'the grid%landmask and soil/veg cats do not match' print *,'i,j=',i,j print *,'grid%landmask=',grid%landmask(i,j) print *,'grid%ivgtyp=',grid%ivgtyp(i,j) print *,'grid%isltyp=',grid%isltyp(i,j) print *,'iswater=', config_flags%iswater print *,'grid%tslb=',grid%tslb(i,:,j) print *,'grid%sst=',grid%sst(i,j) CALL wrf_error_fatal ( 'mismatch_landmask_ivgtyp' ) END IF END IF END DO END DO if (oops1.gt.0) then print *,'points artificially set to land : ',oops1 endif if(oops2.gt.0) then print *,'points artificially set to water: ',oops2 endif ! fill grid%sst array with grid%tsk if missing in real input (needed for time-varying grid%sst in wrf) DO j = jts, MIN(jde-1,jte) DO i = its, MIN(ide-1,ite) IF ( flag_sst .NE. 1 ) THEN grid%sst(i,j) = grid%tsk(i,j) ENDIF END DO END DO ! From the full level data, we can get the half levels, reciprocals, and layer ! thicknesses. These are all defined at half level locations, so one less level. ! We allow the vertical coordinate to *accidently* come in upside down. We want ! the first full level to be the ground surface. ! Check whether grid%znw (full level) data are truly full levels. If not, we need to adjust them ! to be full levels. ! in this test, we check if grid%znw(1) is neither 0 nor 1 (within a tolerance of 10**-5) were_bad = .false. IF ( ( (grid%znw(1).LT.(1-1.E-5) ) .OR. ( grid%znw(1).GT.(1+1.E-5) ) ).AND. & ( (grid%znw(1).LT.(0-1.E-5) ) .OR. ( grid%znw(1).GT.(0+1.E-5) ) ) ) THEN were_bad = .true. print *,'Your grid%znw input values are probably half-levels. ' print *,grid%znw print *,'WRF expects grid%znw values to be full levels. ' print *,'Adjusting now to full levels...' ! We want to ignore the first value if it's negative IF (grid%znw(1).LT.0) THEN grid%znw(1)=0 END IF DO k=2,kde grid%znw(k)=2*grid%znw(k)-grid%znw(k-1) END DO END IF ! Let's check our changes IF ( ( ( grid%znw(1) .LT. (1-1.E-5) ) .OR. ( grid%znw(1) .GT. (1+1.E-5) ) ).AND. & ( ( grid%znw(1) .LT. (0-1.E-5) ) .OR. ( grid%znw(1) .GT. (0+1.E-5) ) ) ) THEN print *,'The input grid%znw height values were half-levels or erroneous. ' print *,'Attempts to treat the values as half-levels and change them ' print *,'to valid full levels failed.' CALL wrf_error_fatal("bad grid%znw values from input files") ELSE IF ( were_bad ) THEN print *,'...adjusted. grid%znw array now contains full eta level values. ' ENDIF IF ( grid%znw(1) .LT. grid%znw(kde) ) THEN DO k=1, kde/2 hold_znw = grid%znw(k) grid%znw(k)=grid%znw(kde+1-k) grid%znw(kde+1-k)=hold_znw END DO END IF DO k=1, kde-1 grid%dnw(k) = grid%znw(k+1) - grid%znw(k) grid%rdnw(k) = 1./grid%dnw(k) grid%znu(k) = 0.5*(grid%znw(k+1)+grid%znw(k)) END DO ! Now the same sort of computations with the half eta levels, even ANOTHER ! level less than the one above. DO k=2, kde-1 grid%dn(k) = 0.5*(grid%dnw(k)+grid%dnw(k-1)) grid%rdn(k) = 1./grid%dn(k) grid%fnp(k) = .5* grid%dnw(k )/grid%dn(k) grid%fnm(k) = .5* grid%dnw(k-1)/grid%dn(k) END DO ! Scads of vertical coefficients. cof1 = (2.*grid%dn(2)+grid%dn(3))/(grid%dn(2)+grid%dn(3))*grid%dnw(1)/grid%dn(2) cof2 = grid%dn(2) /(grid%dn(2)+grid%dn(3))*grid%dnw(1)/grid%dn(3) grid%cf1 = grid%fnp(2) + cof1 grid%cf2 = grid%fnm(2) - cof1 - cof2 grid%cf3 = cof2 grid%cfn = (.5*grid%dnw(kde-1)+grid%dn(kde-1))/grid%dn(kde-1) grid%cfn1 = -.5*grid%dnw(kde-1)/grid%dn(kde-1) ! Inverse grid distances. grid%rdx = 1./config_flags%dx grid%rdy = 1./config_flags%dy ! Some of the many weird geopotential initializations that we'll see today: grid%ph0 is total, ! and grid%ph_2 is a perturbation from the base state geopotential. We set the base geopotential ! at the lowest level to terrain elevation * gravity. DO j=jts,jte DO i=its,ite grid%ph0(i,1,j) = grid%ht(i,j) * g grid%ph_2(i,1,j) = 0. END DO END DO ! Base state potential temperature and inverse density (alpha = 1/rho) from ! the half eta levels and the base-profile surface pressure. Compute 1/rho ! from equation of state. The potential temperature is a perturbation from t0. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) ! Base state pressure is a function of eta level and terrain, only, plus ! the hand full of constants: p00 (sea level pressure, Pa), t00 (sea level ! temperature, K), and A (temperature difference, from 1000 mb to 300 mb, K). p_surf = p00 * EXP ( -t00/a + ( (t00/a)**2 - 2.*g*grid%ht(i,j)/a/r_d ) **0.5 ) DO k = 1, kte-1 grid%php(i,k,j) = grid%znw(k)*(p_surf - grid%p_top) + grid%p_top ! temporary, full lev base pressure grid%pb(i,k,j) = grid%znu(k)*(p_surf - grid%p_top) + grid%p_top ! temp = MAX ( 200., t00 + A*LOG(grid%pb(i,k,j)/p00) ) temp = t00 + A*LOG(grid%pb(i,k,j)/p00) grid%t_init(i,k,j) = temp*(p00/grid%pb(i,k,j))**(r_d/cp) - t0 grid%alb(i,k,j) = (r_d/p1000mb)*(grid%t_init(i,k,j)+t0)*(grid%pb(i,k,j)/p1000mb)**cvpm END DO ! Base state mu is defined as base state surface pressure minus grid%p_top grid%mub(i,j) = p_surf - grid%p_top ! Dry surface pressure is defined as the following (this mu is from the input file ! computed from the dry pressure). Here the dry pressure is just reconstituted. pd_surf = grid%mu0(i,j) + grid%p_top ! Integrate base geopotential, starting at terrain elevation. This assures that ! the base state is in exact hydrostatic balance with respect to the model equations. ! This field is on full levels. grid%phb(i,1,j) = grid%ht(i,j) * g DO k = 2,kte grid%phb(i,k,j) = grid%phb(i,k-1,j) - grid%dnw(k-1)*grid%mub(i,j)*grid%alb(i,k-1,j) END DO END DO END DO ! Fill in the outer rows and columns to allow us to be sloppy. IF ( ite .EQ. ide ) THEN i = ide DO j = jts, MIN(jde-1,jte) grid%mub(i,j) = grid%mub(i-1,j) grid%mu_2(i,j) = grid%mu_2(i-1,j) DO k = 1, kte-1 grid%pb(i,k,j) = grid%pb(i-1,k,j) grid%t_init(i,k,j) = grid%t_init(i-1,k,j) grid%alb(i,k,j) = grid%alb(i-1,k,j) END DO DO k = 1, kte grid%phb(i,k,j) = grid%phb(i-1,k,j) END DO END DO END IF IF ( jte .EQ. jde ) THEN j = jde DO i = its, ite grid%mub(i,j) = grid%mub(i,j-1) grid%mu_2(i,j) = grid%mu_2(i,j-1) DO k = 1, kte-1 grid%pb(i,k,j) = grid%pb(i,k,j-1) grid%t_init(i,k,j) = grid%t_init(i,k,j-1) grid%alb(i,k,j) = grid%alb(i,k,j-1) END DO DO k = 1, kte grid%phb(i,k,j) = grid%phb(i,k,j-1) END DO END DO END IF ! Compute the perturbation dry pressure (grid%mub + grid%mu_2 + ptop = dry grid%psfc). DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) grid%mu_2(i,j) = grid%mu0(i,j) - grid%mub(i,j) END DO END DO ! Fill in the outer rows and columns to allow us to be sloppy. IF ( ite .EQ. ide ) THEN i = ide DO j = jts, MIN(jde-1,jte) grid%mu_2(i,j) = grid%mu_2(i-1,j) END DO END IF IF ( jte .EQ. jde ) THEN j = jde DO i = its, ite grid%mu_2(i,j) = grid%mu_2(i,j-1) END DO END IF lev500 = 0 DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) ! Assign the potential temperature (perturbation from t0) and qv on all the mass ! point locations. DO k = 1 , kde-1 grid%t_2(i,k,j) = grid%t_2(i,k,j) - t0 END DO dpmu = 10001. loop_count = 0 DO WHILE ( ( ABS(dpmu) .GT. 10. ) .AND. & ( loop_count .LT. 5 ) ) loop_count = loop_count + 1 ! Integrate the hydrostatic equation (from the RHS of the bigstep vertical momentum ! equation) down from the top to get the pressure perturbation. First get the pressure ! perturbation, moisture, and inverse density (total and perturbation) at the top-most level. k = kte-1 qvf1 = 0.5*(moist(i,k,j,P_QV)+moist(i,k,j,P_QV)) qvf2 = 1./(1.+qvf1) qvf1 = qvf1*qvf2 grid%p(i,k,j) = - 0.5*(grid%mu_2(i,j)+qvf1*grid%mub(i,j))/grid%rdnw(k)/qvf2 qvf = 1. + rvovrd*moist(i,k,j,P_QV) grid%alt(i,k,j) = (r_d/p1000mb)*(grid%t_2(i,k,j)+t0)*qvf& *(((grid%p(i,k,j)+grid%pb(i,k,j))/p1000mb)**cvpm) grid%al(i,k,j) = grid%alt(i,k,j) - grid%alb(i,k,j) ! Now, integrate down the column to compute the pressure perturbation, and diagnose the two ! inverse density fields (total and perturbation). DO k=kte-2,1,-1 qvf1 = 0.5*(moist(i,k,j,P_QV)+moist(i,k+1,j,P_QV)) qvf2 = 1./(1.+qvf1) qvf1 = qvf1*qvf2 grid%p(i,k,j) = grid%p(i,k+1,j) - (grid%mu_2(i,j) + qvf1*grid%mub(i,j))/qvf2/grid%rdn(k+1) qvf = 1. + rvovrd*moist(i,k,j,P_QV) grid%alt(i,k,j) = (r_d/p1000mb)*(grid%t_2(i,k,j)+t0)*qvf* & (((grid%p(i,k,j)+grid%pb(i,k,j))/p1000mb)**cvpm) grid%al(i,k,j) = grid%alt(i,k,j) - grid%alb(i,k,j) END DO #if 1 ! This is the hydrostatic equation used in the model after the small timesteps. In ! the model, grid%al (inverse density) is computed from the geopotential. DO k = 2,kte grid%ph_2(i,k,j) = grid%ph_2(i,k-1,j) - & grid%dnw(k-1) * ( (grid%mub(i,j)+grid%mu_2(i,j))*grid%al(i,k-1,j) & + grid%mu_2(i,j)*grid%alb(i,k-1,j) ) grid%ph0(i,k,j) = grid%ph_2(i,k,j) + grid%phb(i,k,j) END DO #else ! Get the perturbation geopotential from the 3d height array from WPS. DO k = 2,kte grid%ph_2(i,k,j) = grid%ph0(i,k,j)*g - grid%phb(i,k,j) END DO #endif ! Adjust the column pressure so that the computed 500 mb height is close to the ! input value (of course, not when we are doing hybrid input). IF ( ( flag_metgrid .EQ. 1 ) .AND. ( i .EQ. its ) .AND. ( j .EQ. jts ) ) THEN DO k = 1 , num_metgrid_levels IF ( ABS ( grid%p_gc(i,k,j) - 50000. ) .LT. 1. ) THEN lev500 = k EXIT END IF END DO END IF ! We only do the adjustment of height if we have the input data on pressure ! surfaces, and folks have asked to do this option. IF ( ( flag_metgrid .EQ. 1 ) .AND. & ( config_flags%adjust_heights ) .AND. & ( lev500 .NE. 0 ) ) THEN DO k = 2 , kte-1 ! Get the pressures on the full eta levels (grid%php is defined above as ! the full-lev base pressure, an easy array to use for 3d space). pl = grid%php(i,k ,j) + & ( grid%p(i,k-1 ,j) * ( grid%znw(k ) - grid%znu(k ) ) + & grid%p(i,k ,j) * ( grid%znu(k-1 ) - grid%znw(k ) ) ) / & ( grid%znu(k-1 ) - grid%znu(k ) ) pu = grid%php(i,k+1,j) + & ( grid%p(i,k-1+1,j) * ( grid%znw(k +1) - grid%znu(k+1) ) + & grid%p(i,k +1,j) * ( grid%znu(k-1+1) - grid%znw(k+1) ) ) / & ( grid%znu(k-1+1) - grid%znu(k+1) ) ! If these pressure levels trap 500 mb, use them to interpolate ! to the 500 mb level of the computed height. IF ( ( pl .GE. 50000. ) .AND. ( pu .LT. 50000. ) ) THEN zl = ( grid%ph_2(i,k ,j) + grid%phb(i,k ,j) ) / g zu = ( grid%ph_2(i,k+1,j) + grid%phb(i,k+1,j) ) / g z500 = ( zl * ( LOG(50000.) - LOG(pu ) ) + & zu * ( LOG(pl ) - LOG(50000.) ) ) / & ( LOG(pl) - LOG(pu) ) ! z500 = ( zl * ( (50000.) - (pu ) ) + & ! zu * ( (pl ) - (50000.) ) ) / & ! ( (pl) - (pu) ) ! Compute the difference of the 500 mb heights (computed minus input), and ! then the change in grid%mu_2. The grid%php is still full-levels, base pressure. dz500 = z500 - grid%ght_gc(i,lev500,j) tvsfc = ((grid%t_2(i,1,j)+t0)*((grid%p(i,1,j)+grid%php(i,1,j))/p1000mb)**(r_d/cp)) * & (1.+0.6*moist(i,1,j,P_QV)) dpmu = ( grid%php(i,1,j) + grid%p(i,1,j) ) * EXP ( g * dz500 / ( r_d * tvsfc ) ) dpmu = dpmu - ( grid%php(i,1,j) + grid%p(i,1,j) ) grid%mu_2(i,j) = grid%mu_2(i,j) - dpmu EXIT END IF END DO ELSE dpmu = 0. END IF END DO END DO END DO ! If this is data from the SI, then we probably do not have the original ! surface data laying around. Note that these are all the lowest levels ! of the respective 3d arrays. For surface pressure, we assume that the ! vertical gradient of grid%p prime is zilch. This is not all that important. ! These are filled in so that the various plotting routines have something ! to play with at the initial time for the model. IF ( flag_metgrid .NE. 1 ) THEN DO j = jts, min(jde-1,jte) DO i = its, min(ide,ite) grid%u10(i,j)=grid%u_2(i,1,j) END DO END DO DO j = jts, min(jde,jte) DO i = its, min(ide-1,ite) grid%v10(i,j)=grid%v_2(i,1,j) END DO END DO DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) p_surf = p00 * EXP ( -t00/a + ( (t00/a)**2 - 2.*g*grid%ht(i,j)/a/r_d ) **0.5 ) grid%psfc(i,j)=p_surf + grid%p(i,1,j) grid%q2(i,j)=moist(i,1,j,P_QV) grid%th2(i,j)=grid%t_2(i,1,j)+300. grid%t2(i,j)=grid%th2(i,j)*(((grid%p(i,1,j)+grid%pb(i,1,j))/p00)**(r_d/cp)) END DO END DO ! If this data is from WPS, then we have previously assigned the surface ! data for u, v, and t. If we have an input qv, welp, we assigned that one, ! too. Now we pick up the left overs, and if RH came in - we assign the ! mixing ratio. ELSE IF ( flag_metgrid .EQ. 1 ) THEN DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) p_surf = p00 * EXP ( -t00/a + ( (t00/a)**2 - 2.*g*grid%ht(i,j)/a/r_d ) **0.5 ) grid%psfc(i,j)=p_surf + grid%p(i,1,j) grid%th2(i,j)=grid%t2(i,j)*(p00/(grid%p(i,1,j)+grid%pb(i,1,j)))**(r_d/cp) END DO END DO IF ( flag_qv .NE. 1 ) THEN DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) grid%q2(i,j)=moist(i,1,j,P_QV) END DO END DO END IF END IF ips = its ; ipe = ite ; jps = jts ; jpe = jte ; kps = kts ; kpe = kte #ifdef DM_PARALLEL # include "HALO_EM_INIT_1.inc" # include "HALO_EM_INIT_2.inc" # include "HALO_EM_INIT_3.inc" # include "HALO_EM_INIT_4.inc" # include "HALO_EM_INIT_5.inc" #endif RETURN END SUBROUTINE init_domain_rk !--------------------------------------------------------------------- SUBROUTINE const_module_initialize ( p00 , t00 , a ) USE module_configure IMPLICIT NONE ! For the real-data-cases only. REAL , INTENT(OUT) :: p00 , t00 , a CALL nl_get_base_pres ( 1 , p00 ) CALL nl_get_base_temp ( 1 , t00 ) CALL nl_get_base_lapse ( 1 , a ) END SUBROUTINE const_module_initialize !------------------------------------------------------------------- SUBROUTINE rebalance_driver ( grid ) IMPLICIT NONE TYPE (domain) :: grid CALL rebalance( grid & ! #include "actual_new_args.inc" ! ) END SUBROUTINE rebalance_driver !--------------------------------------------------------------------- SUBROUTINE rebalance ( grid & ! #include "dummy_new_args.inc" ! ) IMPLICIT NONE TYPE (domain) :: grid #include "dummy_new_decl.inc" TYPE (grid_config_rec_type) :: config_flags REAL :: p_surf , pd_surf, p_surf_int , pb_int , ht_hold REAL :: qvf , qvf1 , qvf2 REAL :: p00 , t00 , a REAL , DIMENSION(:,:,:) , ALLOCATABLE :: t_init_int ! Local domain indices and counters. INTEGER :: num_veg_cat , num_soil_top_cat , num_soil_bot_cat INTEGER :: & ids, ide, jds, jde, kds, kde, & ims, ime, jms, jme, kms, kme, & its, ite, jts, jte, kts, kte, & ips, ipe, jps, jpe, kps, kpe, & i, j, k SELECT CASE ( model_data_order ) CASE ( DATA_ORDER_ZXY ) kds = grid%sd31 ; kde = grid%ed31 ; ids = grid%sd32 ; ide = grid%ed32 ; jds = grid%sd33 ; jde = grid%ed33 ; kms = grid%sm31 ; kme = grid%em31 ; ims = grid%sm32 ; ime = grid%em32 ; jms = grid%sm33 ; jme = grid%em33 ; kts = grid%sp31 ; kte = grid%ep31 ; ! note that tile is entire patch its = grid%sp32 ; ite = grid%ep32 ; ! note that tile is entire patch jts = grid%sp33 ; jte = grid%ep33 ; ! note that tile is entire patch CASE ( DATA_ORDER_XYZ ) ids = grid%sd31 ; ide = grid%ed31 ; jds = grid%sd32 ; jde = grid%ed32 ; kds = grid%sd33 ; kde = grid%ed33 ; ims = grid%sm31 ; ime = grid%em31 ; jms = grid%sm32 ; jme = grid%em32 ; kms = grid%sm33 ; kme = grid%em33 ; its = grid%sp31 ; ite = grid%ep31 ; ! note that tile is entire patch jts = grid%sp32 ; jte = grid%ep32 ; ! note that tile is entire patch kts = grid%sp33 ; kte = grid%ep33 ; ! note that tile is entire patch CASE ( DATA_ORDER_XZY ) ids = grid%sd31 ; ide = grid%ed31 ; kds = grid%sd32 ; kde = grid%ed32 ; jds = grid%sd33 ; jde = grid%ed33 ; ims = grid%sm31 ; ime = grid%em31 ; kms = grid%sm32 ; kme = grid%em32 ; jms = grid%sm33 ; jme = grid%em33 ; its = grid%sp31 ; ite = grid%ep31 ; ! note that tile is entire patch kts = grid%sp32 ; kte = grid%ep32 ; ! note that tile is entire patch jts = grid%sp33 ; jte = grid%ep33 ; ! note that tile is entire patch END SELECT ALLOCATE ( t_init_int(ims:ime,kms:kme,jms:jme) ) ! Some of the many weird geopotential initializations that we'll see today: grid%ph0 is total, ! and grid%ph_2 is a perturbation from the base state geopotential. We set the base geopotential ! at the lowest level to terrain elevation * gravity. DO j=jts,jte DO i=its,ite grid%ph0(i,1,j) = grid%ht_fine(i,j) * g grid%ph_2(i,1,j) = 0. END DO END DO ! To define the base state, we call a USER MODIFIED routine to set the three ! necessary constants: p00 (sea level pressure, Pa), t00 (sea level temperature, K), ! and A (temperature difference, from 1000 mb to 300 mb, K). CALL const_module_initialize ( p00 , t00 , a ) ! Base state potential temperature and inverse density (alpha = 1/rho) from ! the half eta levels and the base-profile surface pressure. Compute 1/rho ! from equation of state. The potential temperature is a perturbation from t0. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) ! Base state pressure is a function of eta level and terrain, only, plus ! the hand full of constants: p00 (sea level pressure, Pa), t00 (sea level ! temperature, K), and A (temperature difference, from 1000 mb to 300 mb, K). ! The fine grid terrain is ht_fine, the interpolated is grid%ht. p_surf = p00 * EXP ( -t00/a + ( (t00/a)**2 - 2.*g*grid%ht_fine(i,j)/a/r_d ) **0.5 ) p_surf_int = p00 * EXP ( -t00/a + ( (t00/a)**2 - 2.*g*grid%ht(i,j) /a/r_d ) **0.5 ) DO k = 1, kte-1 grid%pb(i,k,j) = grid%znu(k)*(p_surf - grid%p_top) + grid%p_top pb_int = grid%znu(k)*(p_surf_int - grid%p_top) + grid%p_top grid%t_init(i,k,j) = (t00 + A*LOG(grid%pb(i,k,j)/p00))*(p00/grid%pb(i,k,j))**(r_d/cp) - t0 t_init_int(i,k,j)= (t00 + A*LOG(pb_int /p00))*(p00/pb_int )**(r_d/cp) - t0 grid%alb(i,k,j) = (r_d/p1000mb)*(grid%t_init(i,k,j)+t0)*(grid%pb(i,k,j)/p1000mb)**cvpm END DO ! Base state mu is defined as base state surface pressure minus grid%p_top grid%mub(i,j) = p_surf - grid%p_top ! Dry surface pressure is defined as the following (this mu is from the input file ! computed from the dry pressure). Here the dry pressure is just reconstituted. pd_surf = ( grid%mub(i,j) + grid%mu_2(i,j) ) + grid%p_top ! Integrate base geopotential, starting at terrain elevation. This assures that ! the base state is in exact hydrostatic balance with respect to the model equations. ! This field is on full levels. grid%phb(i,1,j) = grid%ht_fine(i,j) * g DO k = 2,kte grid%phb(i,k,j) = grid%phb(i,k-1,j) - grid%dnw(k-1)*grid%mub(i,j)*grid%alb(i,k-1,j) END DO END DO END DO ! Replace interpolated terrain with fine grid values. DO j = jts, MIN(jte,jde-1) DO i = its, MIN(ite,ide-1) grid%ht(i,j) = grid%ht_fine(i,j) END DO END DO ! Perturbation fields. DO j = jts, min(jde-1,jte) DO i = its, min(ide-1,ite) ! The potential temperature is THETAnest = THETAinterp + ( TBARnest - TBARinterp) DO k = 1 , kde-1 grid%t_2(i,k,j) = grid%t_2(i,k,j) + ( grid%t_init(i,k,j) - t_init_int(i,k,j) ) END DO ! Integrate the hydrostatic equation (from the RHS of the bigstep vertical momentum ! equation) down from the top to get the pressure perturbation. First get the pressure ! perturbation, moisture, and inverse density (total and perturbation) at the top-most level. k = kte-1 qvf1 = 0.5*(moist(i,k,j,P_QV)+moist(i,k,j,P_QV)) qvf2 = 1./(1.+qvf1) qvf1 = qvf1*qvf2 grid%p(i,k,j) = - 0.5*(grid%mu_2(i,j)+qvf1*grid%mub(i,j))/grid%rdnw(k)/qvf2 qvf = 1. + rvovrd*moist(i,k,j,P_QV) grid%alt(i,k,j) = (r_d/p1000mb)*(grid%t_2(i,k,j)+t0)*qvf* & (((grid%p(i,k,j)+grid%pb(i,k,j))/p1000mb)**cvpm) grid%al(i,k,j) = grid%alt(i,k,j) - grid%alb(i,k,j) ! Now, integrate down the column to compute the pressure perturbation, and diagnose the two ! inverse density fields (total and perturbation). DO k=kte-2,1,-1 qvf1 = 0.5*(moist(i,k,j,P_QV)+moist(i,k+1,j,P_QV)) qvf2 = 1./(1.+qvf1) qvf1 = qvf1*qvf2 grid%p(i,k,j) = grid%p(i,k+1,j) - (grid%mu_2(i,j) + qvf1*grid%mub(i,j))/qvf2/grid%rdn(k+1) qvf = 1. + rvovrd*moist(i,k,j,P_QV) grid%alt(i,k,j) = (r_d/p1000mb)*(grid%t_2(i,k,j)+t0)*qvf* & (((grid%p(i,k,j)+grid%pb(i,k,j))/p1000mb)**cvpm) grid%al(i,k,j) = grid%alt(i,k,j) - grid%alb(i,k,j) END DO ! This is the hydrostatic equation used in the model after the small timesteps. In ! the model, grid%al (inverse density) is computed from the geopotential. DO k = 2,kte grid%ph_2(i,k,j) = grid%ph_2(i,k-1,j) - & grid%dnw(k-1) * ( (grid%mub(i,j)+grid%mu_2(i,j))*grid%al(i,k-1,j) & + grid%mu_2(i,j)*grid%alb(i,k-1,j) ) grid%ph0(i,k,j) = grid%ph_2(i,k,j) + grid%phb(i,k,j) END DO END DO END DO DEALLOCATE ( t_init_int ) ips = its ; ipe = ite ; jps = jts ; jpe = jte ; kps = kts ; kpe = kte #ifdef DM_PARALLEL # include "HALO_EM_INIT_1.inc" # include "HALO_EM_INIT_2.inc" # include "HALO_EM_INIT_3.inc" # include "HALO_EM_INIT_4.inc" # include "HALO_EM_INIT_5.inc" #endif END SUBROUTINE rebalance !--------------------------------------------------------------------- RECURSIVE SUBROUTINE find_my_parent ( grid_ptr_in , grid_ptr_out , id_i_am , id_wanted , found_the_id ) USE module_domain TYPE(domain) , POINTER :: grid_ptr_in , grid_ptr_out TYPE(domain) , POINTER :: grid_ptr_sibling INTEGER :: id_wanted , id_i_am LOGICAL :: found_the_id found_the_id = .FALSE. grid_ptr_sibling => grid_ptr_in DO WHILE ( ASSOCIATED ( grid_ptr_sibling ) ) IF ( grid_ptr_sibling%grid_id .EQ. id_wanted ) THEN found_the_id = .TRUE. grid_ptr_out => grid_ptr_sibling RETURN ELSE IF ( grid_ptr_sibling%num_nests .GT. 0 ) THEN grid_ptr_sibling => grid_ptr_sibling%nests(1)%ptr CALL find_my_parent ( grid_ptr_sibling , grid_ptr_out , id_i_am , id_wanted , found_the_id ) ELSE grid_ptr_sibling => grid_ptr_sibling%sibling END IF END DO END SUBROUTINE find_my_parent #endif !--------------------------------------------------------------------- #ifdef VERT_UNIT !This is a main program for a small unit test for the vertical interpolation. program vint implicit none integer , parameter :: ij = 3 integer , parameter :: keta = 30 integer , parameter :: kgen =20 integer :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte integer :: generic real , dimension(1:ij,kgen,1:ij) :: fo , po real , dimension(1:ij,1:keta,1:ij) :: fn_calc , fn_interp , pn integer, parameter :: interp_type = 1 ! 2 ! integer, parameter :: lagrange_order = 2 ! 1 integer :: lagrange_order logical, parameter :: lowest_lev_from_sfc = .FALSE. ! .TRUE. logical, parameter :: use_levels_below_ground = .FALSE. ! .TRUE. logical, parameter :: use_surface = .FALSE. ! .TRUE. real , parameter :: zap_close_levels = 500. ! 100. integer, parameter :: force_sfc_in_vinterp = 0 ! 6 integer :: k ids = 1 ; ide = ij ; jds = 1 ; jde = ij ; kds = 1 ; kde = keta ims = 1 ; ime = ij ; jms = 1 ; jme = ij ; kms = 1 ; kme = keta its = 1 ; ite = ij ; jts = 1 ; jte = ij ; kts = 1 ; kte = keta generic = kgen print *,' ' print *,'------------------------------------' print *,'UNIT TEST FOR VERTICAL INTERPOLATION' print *,'------------------------------------' print *,' ' do lagrange_order = 1 , 2 print *,' ' print *,'------------------------------------' print *,'Lagrange Order = ',lagrange_order print *,'------------------------------------' print *,' ' call fillitup ( fo , po , fn_calc , pn , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte , & generic , lagrange_order ) print *,' ' print *,'Level Pressure Field' print *,' (Pa) (generic)' print *,'------------------------------------' print *,' ' do k = 1 , generic write (*,fmt='(i2,2x,f12.3,1x,g15.8)' ) & k,po(2,k,2),fo(2,k,2) end do print *,' ' call vert_interp ( fo , po , fn_interp , pn , & generic , 'T' , & interp_type , lagrange_order , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) print *,'Multi-Order Interpolator' print *,'------------------------------------' print *,' ' print *,'Level Pressure Field Field Field' print *,' (Pa) Calc Interp Diff' print *,'------------------------------------' print *,' ' do k = kts , kte-1 write (*,fmt='(i2,2x,f12.3,1x,3(g15.7))' ) & k,pn(2,k,2),fn_calc(2,k,2),fn_interp(2,k,2),fn_calc(2,k,2)-fn_interp(2,k,2) end do call vert_interp_old ( fo , po , fn_interp , pn , & generic , 'T' , & interp_type , lagrange_order , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) print *,'Linear Interpolator' print *,'------------------------------------' print *,' ' print *,'Level Pressure Field Field Field' print *,' (Pa) Calc Interp Diff' print *,'------------------------------------' print *,' ' do k = kts , kte-1 write (*,fmt='(i2,2x,f12.3,1x,3(g15.7))' ) & k,pn(2,k,2),fn_calc(2,k,2),fn_interp(2,k,2),fn_calc(2,k,2)-fn_interp(2,k,2) end do end do end program vint subroutine wrf_error_fatal (string) character (len=*) :: string print *,string stop end subroutine wrf_error_fatal subroutine fillitup ( fo , po , fn , pn , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte , & generic , lagrange_order ) implicit none integer , intent(in) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte integer , intent(in) :: generic , lagrange_order real , dimension(ims:ime,generic,jms:jme) , intent(out) :: fo , po real , dimension(ims:ime,kms:kme,jms:jme) , intent(out) :: fn , pn integer :: i , j , k real , parameter :: piov2 = 3.14159265358 / 2. k = 1 do j = jts , jte do i = its , ite po(i,k,j) = 102000. end do end do do k = 2 , generic do j = jts , jte do i = its , ite po(i,k,j) = ( 5000. * ( 1 - (k-1) ) + 100000. * ( (k-1) - (generic-1) ) ) / (1. - real(generic-1) ) end do end do end do if ( lagrange_order .eq. 1 ) then do k = 1 , generic do j = jts , jte do i = its , ite fo(i,k,j) = po(i,k,j) ! fo(i,k,j) = sin(po(i,k,j) * piov2 / 102000. ) end do end do end do else if ( lagrange_order .eq. 2 ) then do k = 1 , generic do j = jts , jte do i = its , ite fo(i,k,j) = (((po(i,k,j)-5000.)/102000.)*((102000.-po(i,k,j))/102000.))*102000. ! fo(i,k,j) = sin(po(i,k,j) * piov2 / 102000. ) end do end do end do end if !!!!!!!!!!!! do k = kts , kte do j = jts , jte do i = its , ite pn(i,k,j) = ( 5000. * ( 0 - (k-1) ) + 102000. * ( (k-1) - (kte-1) ) ) / (-1. * real(kte-1) ) end do end do end do do k = kts , kte-1 do j = jts , jte do i = its , ite pn(i,k,j) = ( pn(i,k,j) + pn(i,k+1,j) ) /2. end do end do end do if ( lagrange_order .eq. 1 ) then do k = kts , kte-1 do j = jts , jte do i = its , ite fn(i,k,j) = pn(i,k,j) ! fn(i,k,j) = sin(pn(i,k,j) * piov2 / 102000. ) end do end do end do else if ( lagrange_order .eq. 2 ) then do k = kts , kte-1 do j = jts , jte do i = its , ite fn(i,k,j) = (((pn(i,k,j)-5000.)/102000.)*((102000.-pn(i,k,j))/102000.))*102000. ! fn(i,k,j) = sin(pn(i,k,j) * piov2 / 102000. ) end do end do end do end if end subroutine fillitup #endif !--------------------------------------------------------------------- SUBROUTINE vert_interp ( fo , po , fnew , pnu , & generic , var_type , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Vertically interpolate the new field. The original field on the original ! pressure levels is provided, and the new pressure surfaces to interpolate to. IMPLICIT NONE INTEGER , INTENT(IN) :: interp_type , lagrange_order , extrap_type LOGICAL , INTENT(IN) :: lowest_lev_from_sfc , use_levels_below_ground , use_surface REAL , INTENT(IN) :: zap_close_levels INTEGER , INTENT(IN) :: force_sfc_in_vinterp INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte INTEGER , INTENT(IN) :: generic CHARACTER (LEN=1) :: var_type REAL , DIMENSION(ims:ime,generic,jms:jme) , INTENT(IN) :: fo , po REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: pnu REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(OUT) :: fnew REAL , DIMENSION(ims:ime,generic,jms:jme) :: forig , porig REAL , DIMENSION(ims:ime,kms:kme,jms:jme) :: pnew ! Local vars INTEGER :: i , j , k , ko , kn , k1 , k2 , ko_1 , ko_2 , knext INTEGER :: istart , iend , jstart , jend , kstart , kend INTEGER , DIMENSION(ims:ime,kms:kme ) :: k_above , k_below INTEGER , DIMENSION(ims:ime ) :: ks INTEGER , DIMENSION(ims:ime ) :: ko_above_sfc INTEGER :: count , zap , zap_below , zap_above , kst , kcount INTEGER :: kinterp_start , kinterp_end , sfc_level LOGICAL :: any_below_ground REAL :: p1 , p2 , pn, hold REAL , DIMENSION(1:generic) :: ordered_porig , ordered_forig REAL , DIMENSION(kts:kte) :: ordered_pnew , ordered_fnew ! Horiontal loop bounds for different variable types. IF ( var_type .EQ. 'U' ) THEN istart = its iend = ite jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = MAX(ids+1,its) , MIN(ide-1,ite) porig(i,k,j) = ( po(i,k,j) + po(i-1,k,j) ) * 0.5 END DO END DO IF ( ids .EQ. its ) THEN DO k = 1,generic porig(its,k,j) = po(its,k,j) END DO END IF IF ( ide .EQ. ite ) THEN DO k = 1,generic porig(ite,k,j) = po(ite-1,k,j) END DO END IF DO k = kstart,kend DO i = MAX(ids+1,its) , MIN(ide-1,ite) pnew(i,k,j) = ( pnu(i,k,j) + pnu(i-1,k,j) ) * 0.5 END DO END DO IF ( ids .EQ. its ) THEN DO k = kstart,kend pnew(its,k,j) = pnu(its,k,j) END DO END IF IF ( ide .EQ. ite ) THEN DO k = kstart,kend pnew(ite,k,j) = pnu(ite-1,k,j) END DO END IF END DO ELSE IF ( var_type .EQ. 'V' ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = jte kstart = kts kend = kte-1 DO i = istart,iend DO k = 1,generic DO j = MAX(jds+1,jts) , MIN(jde-1,jte) porig(i,k,j) = ( po(i,k,j) + po(i,k,j-1) ) * 0.5 END DO END DO IF ( jds .EQ. jts ) THEN DO k = 1,generic porig(i,k,jts) = po(i,k,jts) END DO END IF IF ( jde .EQ. jte ) THEN DO k = 1,generic porig(i,k,jte) = po(i,k,jte-1) END DO END IF DO k = kstart,kend DO j = MAX(jds+1,jts) , MIN(jde-1,jte) pnew(i,k,j) = ( pnu(i,k,j) + pnu(i,k,j-1) ) * 0.5 END DO END DO IF ( jds .EQ. jts ) THEN DO k = kstart,kend pnew(i,k,jts) = pnu(i,k,jts) END DO END IF IF ( jde .EQ. jte ) THEN DO k = kstart,kend pnew(i,k,jte) = pnu(i,k,jte-1) END DO END IF END DO ELSE IF ( ( var_type .EQ. 'W' ) .OR. ( var_type .EQ. 'Z' ) ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO ELSE IF ( ( var_type .EQ. 'T' ) .OR. ( var_type .EQ. 'Q' ) ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO ELSE istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO END IF DO j = jstart , jend ! The lowest level is the surface. Levels 2 through "generic" are supposed to ! be "bottom-up". Flip if they are not. This is based on the input pressure ! array. IF ( porig(its,2,j) .LT. porig(its,generic,j) ) THEN DO kn = 2 , ( generic + 1 ) / 2 DO i = istart , iend hold = porig(i,kn,j) porig(i,kn,j) = porig(i,generic+2-kn,j) porig(i,generic+2-kn,j) = hold forig(i,kn,j) = fo (i,generic+2-kn,j) forig(i,generic+2-kn,j) = fo (i,kn,j) END DO DO i = istart , iend forig(i,1,j) = fo (i,1,j) END DO END DO ELSE DO kn = 1 , generic DO i = istart , iend forig(i,kn,j) = fo (i,kn,j) END DO END DO END IF ! Skip all of the levels below ground in the original data based upon the surface pressure. ! The ko_above_sfc is the index in the pressure array that is above the surface. If there ! are no levels underground, this is index = 2. The remaining levels are eligible for use ! in the vertical interpolation. DO i = istart , iend ko_above_sfc(i) = -1 END DO DO ko = kstart+1 , kend DO i = istart , iend IF ( ko_above_sfc(i) .EQ. -1 ) THEN IF ( porig(i,1,j) .GT. porig(i,ko,j) ) THEN ko_above_sfc(i) = ko END IF END IF END DO END DO ! Piece together columns of the original input data. Pass the vertical columns to ! the iterpolator. DO i = istart , iend ! If the surface value is in the middle of the array, three steps: 1) do the ! values below the ground (this is just to catch the occasional value that is ! inconsistently below the surface based on input data), 2) do the surface level, then ! 3) add in the levels that are above the surface. For the levels next to the surface, ! we check to remove any levels that are "too close". When building the column of input ! pressures, we also attend to the request for forcing the surface analysis to be used ! in a few lower eta-levels. ! Fill in the column from up to the level just below the surface with the input ! presssure and the input field (orig or old, which ever). For an isobaric input ! file, this data is isobaric. ! How many levels have we skipped in the input column. zap = 0 zap_below = 0 zap_above = 0 IF ( ko_above_sfc(i) .GT. 2 ) THEN count = 1 DO ko = 2 , ko_above_sfc(i)-1 ordered_porig(count) = porig(i,ko,j) ordered_forig(count) = forig(i,ko,j) count = count + 1 END DO ! Make sure the pressure just below the surface is not "too close", this ! will cause havoc with the higher order interpolators. In case of a "too close" ! instance, we toss out the offending level (NOT the surface one) by simply ! decrementing the accumulating loop counter. IF ( ordered_porig(count-1) - porig(i,1,j) .LT. zap_close_levels ) THEN count = count -1 zap = 1 zap_below = 1 END IF ! Add in the surface values. ordered_porig(count) = porig(i,1,j) ordered_forig(count) = forig(i,1,j) count = count + 1 ! A usual way to do the vertical interpolation is to pay more attention to the ! surface data. Why? Well it has about 20x the density as the upper air, so we ! hope the analysis is better there. We more strongly use this data by artificially ! tossing out levels above the surface that are beneath a certain number of prescribed ! eta levels at this (i,j). The "zap" value is how many levels of input we are ! removing, which is used to tell the interpolator how many valid values are in ! the column. The "count" value is the increment to the index of levels, and is ! only used for assignments. IF ( force_sfc_in_vinterp .GT. 0 ) THEN ! Get the pressure at the eta level. We want to remove all input pressure levels ! between the level above the surface to the pressure at this eta surface. That ! forces the surface value to be used through the selected eta level. Keep track ! of two things: the level to use above the eta levels, and how many levels we are ! skipping. knext = ko_above_sfc(i) find_level : DO ko = ko_above_sfc(i) , generic IF ( porig(i,ko,j) .LE. pnew(i,force_sfc_in_vinterp,j) ) THEN knext = ko exit find_level ELSE zap = zap + 1 zap_above = zap_above + 1 END IF END DO find_level ! No request for special interpolation, so we just assign the next level to use ! above the surface as, ta da, the first level above the surface. I know, wow. ELSE knext = ko_above_sfc(i) END IF ! One more time, make sure the pressure just above the surface is not "too close", this ! will cause havoc with the higher order interpolators. In case of a "too close" ! instance, we toss out the offending level above the surface (NOT the surface one) by simply ! incrementing the loop counter. Here, count-1 is the surface level and knext is either ! the next level up OR it is the level above the prescribed number of eta surfaces. IF ( ordered_porig(count-1) - porig(i,knext,j) .LT. zap_close_levels ) THEN kst = knext+1 zap = zap + 1 zap_above = zap_above + 1 ELSE kst = knext END IF DO ko = kst , generic ordered_porig(count) = porig(i,ko,j) ordered_forig(count) = forig(i,ko,j) count = count + 1 END DO ! This is easy, the surface is the lowest level, just stick them in, in this order. OK, ! there are a couple of subtleties. We have to check for that special interpolation that ! skips some input levels so that the surface is used for the lowest few eta levels. Also, ! we must macke sure that we still do not have levels that are "too close" together. ELSE ! Initialize no input levels have yet been removed from consideration. zap = 0 ! The surface is the lowest level, so it gets set right away to location 1. ordered_porig(1) = porig(i,1,j) ordered_forig(1) = forig(i,1,j) ! We start filling in the array at loc 2, as in just above the level we just stored. count = 2 ! Are we forcing the interpolator to skip valid input levels so that the ! surface data is used through more levels? Essentially as above. IF ( force_sfc_in_vinterp .GT. 0 ) THEN knext = 2 find_level2: DO ko = 2 , generic IF ( porig(i,ko,j) .LE. pnew(i,force_sfc_in_vinterp,j) ) THEN knext = ko exit find_level2 ELSE zap = zap + 1 zap_above = zap_above + 1 END IF END DO find_level2 ELSE knext = 2 END IF ! Fill in the data above the surface. The "knext" index is either the one ! just above the surface OR it is the index associated with the level that ! is just above the pressure at this (i,j) of the top eta level that is to ! be directly impacted with the surface level in interpolation. DO ko = knext , generic IF ( ordered_porig(count-1) - porig(i,ko,j) .LT. zap_close_levels ) THEN zap = zap + 1 zap_above = zap_above + 1 CYCLE END IF ordered_porig(count) = porig(i,ko,j) ordered_forig(count) = forig(i,ko,j) count = count + 1 END DO END IF ! Now get the column of the "new" pressure data. So, this one is easy. DO kn = kstart , kend ordered_pnew(kn) = pnew(i,kn,j) END DO ! How many levels (count) are we shipping to the Lagrange interpolator. IF ( ( use_levels_below_ground ) .AND. ( use_surface ) ) THEN ! Use all levels, including the input surface, and including the pressure ! levels below ground. We know to stop when we have reached the top of ! the input pressure data. count = 0 find_how_many_1 : DO ko = 1 , generic IF ( porig(i,generic,j) .EQ. ordered_porig(ko) ) THEN count = count + 1 EXIT find_how_many_1 ELSE count = count + 1 END IF END DO find_how_many_1 kinterp_start = 1 kinterp_end = kinterp_start + count - 1 ELSE IF ( ( use_levels_below_ground ) .AND. ( .NOT. use_surface ) ) THEN ! Use all levels (excluding the input surface) and including the pressure ! levels below ground. We know to stop when we have reached the top of ! the input pressure data. count = 0 find_sfc_2 : DO ko = 1 , generic IF ( porig(i,1,j) .EQ. ordered_porig(ko) ) THEN sfc_level = ko EXIT find_sfc_2 END IF END DO find_sfc_2 DO ko = sfc_level , generic-1 ordered_porig(ko) = ordered_porig(ko+1) ordered_forig(ko) = ordered_forig(ko+1) END DO ordered_porig(generic) = 1.E-5 ordered_forig(generic) = 1.E10 count = 0 find_how_many_2 : DO ko = 1 , generic IF ( porig(i,generic,j) .EQ. ordered_porig(ko) ) THEN count = count + 1 EXIT find_how_many_2 ELSE count = count + 1 END IF END DO find_how_many_2 kinterp_start = 1 kinterp_end = kinterp_start + count - 1 ELSE IF ( ( .NOT. use_levels_below_ground ) .AND. ( use_surface ) ) THEN ! Use all levels above the input surface pressure. kcount = ko_above_sfc(i)-1-zap_below count = 0 DO ko = 1 , generic IF ( porig(i,ko,j) .EQ. ordered_porig(kcount) ) THEN ! write (6,fmt='(f11.3,f11.3,g11.5)') porig(i,ko,j),ordered_porig(kcount),ordered_forig(kcount) kcount = kcount + 1 count = count + 1 ELSE ! write (6,fmt='(f11.3 )') porig(i,ko,j) END IF END DO kinterp_start = ko_above_sfc(i)-1-zap_below kinterp_end = kinterp_start + count - 1 END IF ! The polynomials are either in pressure or LOG(pressure). IF ( interp_type .EQ. 1 ) THEN CALL lagrange_setup ( var_type , & ordered_porig(kinterp_start:kinterp_end) , & ordered_forig(kinterp_start:kinterp_end) , & count , lagrange_order , extrap_type , & ordered_pnew(kstart:kend) , ordered_fnew , kend-kstart+1 ,i,j) ELSE CALL lagrange_setup ( var_type , & LOG(ordered_porig(kinterp_start:kinterp_end)) , & ordered_forig(kinterp_start:kinterp_end) , & count , lagrange_order , extrap_type , & LOG(ordered_pnew(kstart:kend)) , ordered_fnew , kend-kstart+1 ,i,j) END IF ! Save the computed data. DO kn = kstart , kend fnew(i,kn,j) = ordered_fnew(kn) END DO ! There may have been a request to have the surface data from the input field ! to be assigned as to the lowest eta level. This assumes thin layers (usually ! the isobaric original field has the surface from 2-m T and RH, and 10-m U and V). IF ( lowest_lev_from_sfc ) THEN fnew(i,1,j) = forig(i,ko_above_sfc(i)-1,j) END IF END DO END DO END SUBROUTINE vert_interp !--------------------------------------------------------------------- SUBROUTINE vert_interp_old ( forig , po , fnew , pnu , & generic , var_type , & interp_type , lagrange_order , extrap_type , & lowest_lev_from_sfc , use_levels_below_ground , use_surface , & zap_close_levels , force_sfc_in_vinterp , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Vertically interpolate the new field. The original field on the original ! pressure levels is provided, and the new pressure surfaces to interpolate to. IMPLICIT NONE INTEGER , INTENT(IN) :: interp_type , lagrange_order , extrap_type LOGICAL , INTENT(IN) :: lowest_lev_from_sfc , use_levels_below_ground , use_surface REAL , INTENT(IN) :: zap_close_levels INTEGER , INTENT(IN) :: force_sfc_in_vinterp INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte INTEGER , INTENT(IN) :: generic CHARACTER (LEN=1) :: var_type REAL , DIMENSION(ims:ime,generic,jms:jme) , INTENT(IN) :: forig , po REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: pnu REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(OUT) :: fnew REAL , DIMENSION(ims:ime,generic,jms:jme) :: porig REAL , DIMENSION(ims:ime,kms:kme,jms:jme) :: pnew ! Local vars INTEGER :: i , j , k , ko , kn , k1 , k2 , ko_1 , ko_2 INTEGER :: istart , iend , jstart , jend , kstart , kend INTEGER , DIMENSION(ims:ime,kms:kme ) :: k_above , k_below INTEGER , DIMENSION(ims:ime ) :: ks INTEGER , DIMENSION(ims:ime ) :: ko_above_sfc LOGICAL :: any_below_ground REAL :: p1 , p2 , pn integer vert_extrap vert_extrap = 0 ! Horiontal loop bounds for different variable types. IF ( var_type .EQ. 'U' ) THEN istart = its iend = ite jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = MAX(ids+1,its) , MIN(ide-1,ite) porig(i,k,j) = ( po(i,k,j) + po(i-1,k,j) ) * 0.5 END DO END DO IF ( ids .EQ. its ) THEN DO k = 1,generic porig(its,k,j) = po(its,k,j) END DO END IF IF ( ide .EQ. ite ) THEN DO k = 1,generic porig(ite,k,j) = po(ite-1,k,j) END DO END IF DO k = kstart,kend DO i = MAX(ids+1,its) , MIN(ide-1,ite) pnew(i,k,j) = ( pnu(i,k,j) + pnu(i-1,k,j) ) * 0.5 END DO END DO IF ( ids .EQ. its ) THEN DO k = kstart,kend pnew(its,k,j) = pnu(its,k,j) END DO END IF IF ( ide .EQ. ite ) THEN DO k = kstart,kend pnew(ite,k,j) = pnu(ite-1,k,j) END DO END IF END DO ELSE IF ( var_type .EQ. 'V' ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = jte kstart = kts kend = kte-1 DO i = istart,iend DO k = 1,generic DO j = MAX(jds+1,jts) , MIN(jde-1,jte) porig(i,k,j) = ( po(i,k,j) + po(i,k,j-1) ) * 0.5 END DO END DO IF ( jds .EQ. jts ) THEN DO k = 1,generic porig(i,k,jts) = po(i,k,jts) END DO END IF IF ( jde .EQ. jte ) THEN DO k = 1,generic porig(i,k,jte) = po(i,k,jte-1) END DO END IF DO k = kstart,kend DO j = MAX(jds+1,jts) , MIN(jde-1,jte) pnew(i,k,j) = ( pnu(i,k,j) + pnu(i,k,j-1) ) * 0.5 END DO END DO IF ( jds .EQ. jts ) THEN DO k = kstart,kend pnew(i,k,jts) = pnu(i,k,jts) END DO END IF IF ( jde .EQ. jte ) THEN DO k = kstart,kend pnew(i,k,jte) = pnu(i,k,jte-1) END DO END IF END DO ELSE IF ( ( var_type .EQ. 'W' ) .OR. ( var_type .EQ. 'Z' ) ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO ELSE IF ( ( var_type .EQ. 'T' ) .OR. ( var_type .EQ. 'Q' ) ) THEN istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO ELSE istart = its iend = MIN(ide-1,ite) jstart = jts jend = MIN(jde-1,jte) kstart = kts kend = kte-1 DO j = jstart,jend DO k = 1,generic DO i = istart,iend porig(i,k,j) = po(i,k,j) END DO END DO DO k = kstart,kend DO i = istart,iend pnew(i,k,j) = pnu(i,k,j) END DO END DO END DO END IF DO j = jstart , jend ! Skip all of the levels below ground in the original data based upon the surface pressure. ! The ko_above_sfc is the index in the pressure array that is above the surface. If there ! are no levels underground, this is index = 2. The remaining levels are eligible for use ! in the vertical interpolation. DO i = istart , iend ko_above_sfc(i) = -1 END DO DO ko = kstart+1 , kend DO i = istart , iend IF ( ko_above_sfc(i) .EQ. -1 ) THEN IF ( porig(i,1,j) .GT. porig(i,ko,j) ) THEN ko_above_sfc(i) = ko END IF END IF END DO END DO ! Initialize interpolation location. These are the levels in the original pressure ! data that are physically below and above the targeted new pressure level. DO kn = kts , kte DO i = its , ite k_above(i,kn) = -1 k_below(i,kn) = -2 END DO END DO ! Starting location is no lower than previous found location. This is for O(n logn) ! and not O(n^2), where n is the number of vertical levels to search. DO i = its , ite ks(i) = 1 END DO ! Find trapping layer for interpolation. The kn index runs through all of the "new" ! levels of data. DO kn = kstart , kend DO i = istart , iend ! For each "new" level (kn), we search to find the trapping levels in the "orig" ! data. Most of the time, the "new" levels are the eta surfaces, and the "orig" ! levels are the input pressure levels. found_trap_above : DO ko = ks(i) , generic-1 ! Because we can have levels in the interpolation that are not valid, ! let's toss out any candidate orig pressure values that are below ground ! based on the surface pressure. If the level =1, then this IS the surface ! level, so we HAVE to keep that one, but maybe not the ones above. If the ! level (ks) is NOT=1, then we have to just CYCLE our loop to find a legit ! below-pressure value. If we are not below ground, then we choose two ! neighboring levels to test whether they surround the new pressure level. ! The input trapping levels that we are trying is the surface and the first valid ! level above the surface. IF ( ( ko .LT. ko_above_sfc(i) ) .AND. ( ko .EQ. 1 ) ) THEN ko_1 = ko ko_2 = ko_above_sfc(i) ! The "below" level is underground, cycle until we get to a valid pressure ! above ground. ELSE IF ( ( ko .LT. ko_above_sfc(i) ) .AND. ( ko .NE. 1 ) ) THEN CYCLE found_trap_above ! The "below" level is above the surface, so we are in the clear to test these ! two levels out. ELSE ko_1 = ko ko_2 = ko+1 END IF ! The test of the candidate levels: "below" has to have a larger pressure, and ! "above" has to have a smaller pressure. ! OK, we found the correct two surrounding levels. The locations are saved for use in the ! interpolation. IF ( ( porig(i,ko_1,j) .GE. pnew(i,kn,j) ) .AND. & ( porig(i,ko_2,j) .LT. pnew(i,kn,j) ) ) THEN k_above(i,kn) = ko_2 k_below(i,kn) = ko_1 ks(i) = ko_1 EXIT found_trap_above ! What do we do is we need to extrapolate the data underground? This happens when the ! lowest pressure that we have is physically "above" the new target pressure. Our ! actions depend on the type of variable we are interpolating. ELSE IF ( porig(i,1,j) .LT. pnew(i,kn,j) ) THEN ! For horizontal winds and moisture, we keep a constant value under ground. IF ( ( var_type .EQ. 'U' ) .OR. & ( var_type .EQ. 'V' ) .OR. & ( var_type .EQ. 'Q' ) ) THEN k_above(i,kn) = 1 ks(i) = 1 ! For temperature and height, we extrapolate the data. Hopefully, we are not ! extrapolating too far. For pressure level input, the eta levels are always ! contained within the surface to p_top levels, so no extrapolation is ever ! required. ELSE IF ( ( var_type .EQ. 'Z' ) .OR. & ( var_type .EQ. 'T' ) ) THEN k_above(i,kn) = ko_above_sfc(i) k_below(i,kn) = 1 ks(i) = 1 ! Just a catch all right now. ELSE k_above(i,kn) = 1 ks(i) = 1 END IF EXIT found_trap_above ! The other extrapolation that might be required is when we are going above the ! top level of the input data. Usually this means we chose a P_PTOP value that ! was inappropriate, and we should stop and let someone fix this mess. ELSE IF ( porig(i,generic,j) .GT. pnew(i,kn,j) ) THEN print *,'data is too high, try a lower p_top' print *,'pnew=',pnew(i,kn,j) print *,'porig=',porig(i,:,j) CALL wrf_error_fatal ('requested p_top is higher than input data, lower p_top') END IF END DO found_trap_above END DO END DO ! Linear vertical interpolation. DO kn = kstart , kend DO i = istart , iend IF ( k_above(i,kn) .EQ. 1 ) THEN fnew(i,kn,j) = forig(i,1,j) ELSE k2 = MAX ( k_above(i,kn) , 2) k1 = MAX ( k_below(i,kn) , 1) IF ( k1 .EQ. k2 ) THEN CALL wrf_error_fatal ( 'identical values in the interp, bad for divisions' ) END IF IF ( interp_type .EQ. 1 ) THEN p1 = porig(i,k1,j) p2 = porig(i,k2,j) pn = pnew(i,kn,j) ELSE IF ( interp_type .EQ. 2 ) THEN p1 = ALOG(porig(i,k1,j)) p2 = ALOG(porig(i,k2,j)) pn = ALOG(pnew(i,kn,j)) END IF IF ( ( p1-pn) * (p2-pn) > 0. ) THEN ! CALL wrf_error_fatal ( 'both trapping pressures are on the same side of the new pressure' ) ! CALL wrf_debug ( 0 , 'both trapping pressures are on the same side of the new pressure' ) vert_extrap = vert_extrap + 1 END IF fnew(i,kn,j) = ( forig(i,k1,j) * ( p2 - pn ) + & forig(i,k2,j) * ( pn - p1 ) ) / & ( p2 - p1 ) END IF END DO END DO search_below_ground : DO kn = kstart , kend any_below_ground = .FALSE. DO i = istart , iend IF ( k_above(i,kn) .EQ. 1 ) THEN fnew(i,kn,j) = forig(i,1,j) any_below_ground = .TRUE. END IF END DO IF ( .NOT. any_below_ground ) THEN EXIT search_below_ground END IF END DO search_below_ground ! There may have been a request to have the surface data from the input field ! to be assigned as to the lowest eta level. This assumes thin layers (usually ! the isobaric original field has the surface from 2-m T and RH, and 10-m U and V). DO i = istart , iend IF ( lowest_lev_from_sfc ) THEN fnew(i,1,j) = forig(i,ko_above_sfc(i),j) END IF END DO END DO print *,'VERT EXTRAP = ', vert_extrap END SUBROUTINE vert_interp_old !--------------------------------------------------------------------- SUBROUTINE lagrange_setup ( var_type , all_x , all_y , all_dim , n , extrap_type , & target_x , target_y , target_dim ,i,j) ! We call a Lagrange polynomial interpolator. The parallel concerns are put off as this ! is initially set up for vertical use. The purpose is an input column of pressure (all_x), ! and the associated pressure level data (all_y). These are assumed to be sorted (ascending ! or descending, no matter). The locations to be interpolated to are the pressures in ! target_x, probably the new vertical coordinate values. The field that is output is the ! target_y, which is defined at the target_x location. Mostly we expect to be 2nd order ! overlapping polynomials, with only a single 2nd order method near the top and bottom. ! When n=1, this is linear; when n=2, this is a second order interpolator. IMPLICIT NONE CHARACTER (LEN=1) :: var_type INTEGER , INTENT(IN) :: all_dim , n , extrap_type , target_dim REAL, DIMENSION(all_dim) , INTENT(IN) :: all_x , all_y REAL , DIMENSION(target_dim) , INTENT(IN) :: target_x REAL , DIMENSION(target_dim) , INTENT(OUT) :: target_y ! Brought in for debug purposes, all of the computations are in a single column. INTEGER , INTENT(IN) :: i,j ! Local vars REAL , DIMENSION(n+1) :: x , y REAL :: a , b REAL :: target_y_1 , target_y_2 LOGICAL :: found_loc INTEGER :: loop , loc_center_left , loc_center_right , ist , iend , target_loop INTEGER :: vboundb , vboundt ! Local vars for the problem of extrapolating theta below ground. REAL :: temp_1 , temp_2 , temp_3 , temp_y REAL :: depth_of_extrap_in_p , avg_of_extrap_p , temp_extrap_starting_point , dhdp , dh , dt REAL , PARAMETER :: RovCp = 287. / 1004. REAL , PARAMETER :: CRC_const1 = 11880.516 ! m REAL , PARAMETER :: CRC_const2 = 0.1902632 ! REAL , PARAMETER :: CRC_const3 = 0.0065 ! K/km IF ( all_dim .LT. n+1 ) THEN print *,'all_dim = ',all_dim print *,'order = ',n print *,'i,j = ',i,j print *,'p array = ',all_x print *,'f array = ',all_y print *,'p target= ',target_x CALL wrf_error_fatal ( 'troubles, the interpolating order is too large for this few input values' ) END IF IF ( n .LT. 1 ) THEN CALL wrf_error_fatal ( 'pal, linear is about as low as we go' ) END IF ! We can pinch in the area of the higher order interpolation with vbound. If ! vbound = 0, no pinching. If vbound = m, then we make the lower "m" and upper ! "m" eta levels use a linear interpolation. vboundb = 4 vboundt = 0 ! Loop over the list of target x and y values. DO target_loop = 1 , target_dim ! Find the two trapping x values, and keep the indices. found_loc = .FALSE. find_trap : DO loop = 1 , all_dim -1 a = target_x(target_loop) - all_x(loop) b = target_x(target_loop) - all_x(loop+1) IF ( a*b .LE. 0.0 ) THEN loc_center_left = loop loc_center_right = loop+1 found_loc = .TRUE. EXIT find_trap END IF END DO find_trap IF ( ( .NOT. found_loc ) .AND. ( target_x(target_loop) .GT. all_x(1) ) ) THEN ! Isothermal extrapolation. IF ( ( extrap_type .EQ. 1 ) .AND. ( var_type .EQ. 'T' ) ) THEN temp_1 = all_y(1) * ( all_x(1) / 100000. ) ** RovCp target_y(target_loop) = temp_1 * ( 100000. / target_x(target_loop) ) ** RovCp ! Standard atmosphere -6.5 K/km lapse rate for the extrapolation. ELSE IF ( ( extrap_type .EQ. 2 ) .AND. ( var_type .EQ. 'T' ) ) THEN depth_of_extrap_in_p = target_x(target_loop) - all_x(1) avg_of_extrap_p = ( target_x(target_loop) + all_x(1) ) * 0.5 temp_extrap_starting_point = all_y(1) * ( all_x(1) / 100000. ) ** RovCp dhdp = CRC_const1 * CRC_const2 * ( avg_of_extrap_p / 100. ) ** ( CRC_const2 - 1. ) dh = dhdp * ( depth_of_extrap_in_p / 100. ) dt = dh * CRC_const3 target_y(target_loop) = ( temp_extrap_starting_point + dt ) * ( 100000. / target_x(target_loop) ) ** RovCp ! Adiabatic extrapolation for theta. ELSE IF ( ( extrap_type .EQ. 3 ) .AND. ( var_type .EQ. 'T' ) ) THEN target_y(target_loop) = all_y(1) ! Wild extrapolation for non-temperature vars. ELSE IF ( extrap_type .EQ. 1 ) THEN target_y(target_loop) = ( all_y(2) * ( target_x(target_loop) - all_x(3) ) + & all_y(3) * ( all_x(2) - target_x(target_loop) ) ) / & ( all_x(2) - all_x(3) ) ! Use a constant value below ground. ELSE IF ( extrap_type .EQ. 2 ) THEN target_y(target_loop) = all_y(1) ELSE IF ( extrap_type .EQ. 3 ) THEN CALL wrf_error_fatal ( 'You are not allowed to use extrap_option #3 for any var except for theta.' ) END IF CYCLE ELSE IF ( .NOT. found_loc ) THEN print *,'i,j = ',i,j print *,'target pressure and value = ',target_x(target_loop),target_y(target_loop) DO loop = 1 , all_dim print *,'column of pressure and value = ',all_x(loop),all_y(loop) END DO CALL wrf_error_fatal ( 'troubles, could not find trapping x locations' ) END IF ! Even or odd order? We can put the value in the middle if this is ! an odd order interpolator. For the even guys, we'll do it twice ! and shift the range one index, then get an average. IF ( MOD(n,2) .NE. 0 ) THEN IF ( ( loc_center_left -(((n+1)/2)-1) .GE. 1 ) .AND. & ( loc_center_right+(((n+1)/2)-1) .LE. all_dim ) ) THEN ist = loc_center_left -(((n+1)/2)-1) iend = ist + n CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , n , target_x(target_loop) , target_y(target_loop) ) ELSE IF ( .NOT. found_loc ) THEN CALL wrf_error_fatal ( 'I doubt this will happen, I will only do 2nd order for now' ) END IF END IF ELSE IF ( ( MOD(n,2) .EQ. 0 ) .AND. & ( ( target_loop .GE. 1 + vboundb ) .AND. ( target_loop .LE. target_dim - vboundt ) ) ) THEN IF ( ( loc_center_left -(((n )/2)-1) .GE. 1 ) .AND. & ( loc_center_right+(((n )/2) ) .LE. all_dim ) .AND. & ( loc_center_left -(((n )/2) ) .GE. 1 ) .AND. & ( loc_center_right+(((n )/2)-1) .LE. all_dim ) ) THEN ist = loc_center_left -(((n )/2)-1) iend = ist + n CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , n , target_x(target_loop) , target_y_1 ) ist = loc_center_left -(((n )/2) ) iend = ist + n CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , n , target_x(target_loop) , target_y_2 ) target_y(target_loop) = ( target_y_1 + target_y_2 ) * 0.5 ELSE IF ( ( loc_center_left -(((n )/2)-1) .GE. 1 ) .AND. & ( loc_center_right+(((n )/2) ) .LE. all_dim ) ) THEN ist = loc_center_left -(((n )/2)-1) iend = ist + n CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , n , target_x(target_loop) , target_y(target_loop) ) ELSE IF ( ( loc_center_left -(((n )/2) ) .GE. 1 ) .AND. & ( loc_center_right+(((n )/2)-1) .LE. all_dim ) ) THEN ist = loc_center_left -(((n )/2) ) iend = ist + n CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , n , target_x(target_loop) , target_y(target_loop) ) ELSE CALL wrf_error_fatal ( 'unauthorized area, you should not be here' ) END IF ELSE IF ( MOD(n,2) .EQ. 0 ) THEN ist = loc_center_left iend = loc_center_right CALL lagrange_interp ( all_x(ist:iend) , all_y(ist:iend) , 1 , target_x(target_loop) , target_y(target_loop) ) END IF END DO END SUBROUTINE lagrange_setup !--------------------------------------------------------------------- SUBROUTINE lagrange_interp ( x , y , n , target_x , target_y ) ! Interpolation using Lagrange polynomials. ! P(x) = f(x0)Ln0(x) + ... + f(xn)Lnn(x) ! where Lnk(x) = (x -x0)(x -x1)...(x -xk-1)(x -xk+1)...(x -xn) ! --------------------------------------------- ! (xk-x0)(xk-x1)...(xk-xk-1)(xk-xk+1)...(xk-xn) IMPLICIT NONE INTEGER , INTENT(IN) :: n REAL , DIMENSION(0:n) , INTENT(IN) :: x , y REAL , INTENT(IN) :: target_x REAL , INTENT(OUT) :: target_y ! Local vars INTEGER :: i , k REAL :: numer , denom , Px REAL , DIMENSION(0:n) :: Ln Px = 0. DO i = 0 , n numer = 1. denom = 1. DO k = 0 , n IF ( k .EQ. i ) CYCLE numer = numer * ( target_x - x(k) ) denom = denom * ( x(i) - x(k) ) END DO Ln(i) = y(i) * numer / denom Px = Px + Ln(i) END DO target_y = Px END SUBROUTINE lagrange_interp #ifndef VERT_UNIT !--------------------------------------------------------------------- SUBROUTINE p_dry ( mu0 , eta , pdht , pdry , full_levs , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute reference pressure and the reference mu. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte LOGICAL :: full_levs REAL , DIMENSION(ims:ime, jms:jme) , INTENT(IN) :: mu0 REAL , DIMENSION( kms:kme ) , INTENT(IN) :: eta REAL :: pdht REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(OUT) :: pdry ! Local vars INTEGER :: i , j , k REAL , DIMENSION( kms:kme ) :: eta_h IF ( full_levs ) THEN DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte DO i = its , MIN (ide-1 , ite ) pdry(i,k,j) = eta(k) * mu0(i,j) + pdht END DO END DO END DO ELSE DO k = kts , kte-1 eta_h(k) = ( eta(k) + eta(k+1) ) * 0.5 END DO DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte-1 DO i = its , MIN (ide-1 , ite ) pdry(i,k,j) = eta_h(k) * mu0(i,j) + pdht END DO END DO END DO END IF END SUBROUTINE p_dry !--------------------------------------------------------------------- SUBROUTINE p_dts ( pdts , intq , psfc , p_top , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute difference between the dry, total surface pressure and the top pressure. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , INTENT(IN) :: p_top REAL , DIMENSION(ims:ime,jms:jme) , INTENT(IN) :: psfc REAL , DIMENSION(ims:ime,jms:jme) , INTENT(IN) :: intq REAL , DIMENSION(ims:ime,jms:jme) , INTENT(OUT) :: pdts ! Local vars INTEGER :: i , j , k DO j = jts , MIN ( jde-1 , jte ) DO i = its , MIN (ide-1 , ite ) pdts(i,j) = psfc(i,j) - intq(i,j) - p_top END DO END DO END SUBROUTINE p_dts !--------------------------------------------------------------------- SUBROUTINE p_dhs ( pdhs , ht , p0 , t0 , a , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute dry, hydrostatic surface pressure. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , DIMENSION(ims:ime, jms:jme) , INTENT(IN) :: ht REAL , DIMENSION(ims:ime, jms:jme) , INTENT(OUT) :: pdhs REAL , INTENT(IN) :: p0 , t0 , a ! Local vars INTEGER :: i , j , k REAL , PARAMETER :: Rd = 287. REAL , PARAMETER :: g = 9.8 DO j = jts , MIN ( jde-1 , jte ) DO i = its , MIN (ide-1 , ite ) pdhs(i,j) = p0 * EXP ( -t0/a + SQRT ( (t0/a)**2 - 2. * g * ht(i,j)/(a * Rd) ) ) END DO END DO END SUBROUTINE p_dhs !--------------------------------------------------------------------- SUBROUTINE find_p_top ( p , p_top , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Find the largest pressure in the top level. This is our p_top. We are ! assuming that the top level is the location where the pressure is a minimum ! for each column. In cases where the top surface is not isobaric, a ! communicated value must be shared in the calling routine. Also in cases ! where the top surface is not isobaric, care must be taken that the new ! maximum pressure is not greater than the previous value. This test is ! also handled in the calling routine. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL :: p_top REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: p ! Local vars INTEGER :: i , j , k, min_lev i = its j = jts p_top = p(i,2,j) min_lev = 2 DO k = 2 , kte IF ( p_top .GT. p(i,k,j) ) THEN p_top = p(i,k,j) min_lev = k END IF END DO k = min_lev p_top = p(its,k,jts) DO j = jts , MIN ( jde-1 , jte ) DO i = its , MIN (ide-1 , ite ) p_top = MAX ( p_top , p(i,k,j) ) END DO END DO END SUBROUTINE find_p_top !--------------------------------------------------------------------- SUBROUTINE t_to_theta ( t , p , p00 , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute dry, hydrostatic surface pressure. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , INTENT(IN) :: p00 REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: p REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(INOUT) :: t ! Local vars INTEGER :: i , j , k REAL , PARAMETER :: Rd = 287. REAL , PARAMETER :: Cp = 1004. DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte DO i = its , MIN (ide-1 , ite ) t(i,k,j) = t(i,k,j) * ( p00 / p(i,k,j) ) ** (Rd / Cp) END DO END DO END DO END SUBROUTINE t_to_theta !--------------------------------------------------------------------- SUBROUTINE integ_moist ( q_in , p_in , pd_out , t_in , ght_in , intq , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Integrate the moisture field vertically. Mostly used to get the total ! vapor pressure, which can be subtracted from the total pressure to get ! the dry pressure. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: q_in , p_in , t_in , ght_in REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(OUT) :: pd_out REAL , DIMENSION(ims:ime, jms:jme) , INTENT(OUT) :: intq ! Local vars INTEGER :: i , j , k INTEGER , DIMENSION(ims:ime) :: level_above_sfc REAL , DIMENSION(ims:ime,jms:jme) :: psfc , tsfc , qsfc, zsfc REAL , DIMENSION(ims:ime,kms:kme) :: q , p , t , ght, pd REAL :: rhobar , qbar , dz REAL :: p1 , p2 , t1 , t2 , q1 , q2 , z1, z2 LOGICAL :: upside_down REAL , PARAMETER :: Rd = 287. REAL , PARAMETER :: g = 9.8 ! Get a surface value, always the first level of a 3d field. DO j = jts , MIN ( jde-1 , jte ) DO i = its , MIN (ide-1 , ite ) psfc(i,j) = p_in(i,kts,j) tsfc(i,j) = t_in(i,kts,j) qsfc(i,j) = q_in(i,kts,j) zsfc(i,j) = ght_in(i,kts,j) END DO END DO IF ( p_in(its,kts+1,jts) .LT. p_in(its,kte,jts) ) THEN upside_down = .TRUE. ELSE upside_down = .FALSE. END IF DO j = jts , MIN ( jde-1 , jte ) ! Initialize the integrated quantity of moisture to zero. DO i = its , MIN (ide-1 , ite ) intq(i,j) = 0. END DO IF ( upside_down ) THEN DO i = its , MIN (ide-1 , ite ) p(i,kts) = p_in(i,kts,j) t(i,kts) = t_in(i,kts,j) q(i,kts) = q_in(i,kts,j) ght(i,kts) = ght_in(i,kts,j) DO k = kts+1,kte p(i,k) = p_in(i,kte+2-k,j) t(i,k) = t_in(i,kte+2-k,j) q(i,k) = q_in(i,kte+2-k,j) ght(i,k) = ght_in(i,kte+2-k,j) END DO END DO ELSE DO i = its , MIN (ide-1 , ite ) DO k = kts,kte p(i,k) = p_in(i,k ,j) t(i,k) = t_in(i,k ,j) q(i,k) = q_in(i,k ,j) ght(i,k) = ght_in(i,k ,j) END DO END DO END IF ! Find the first level above the ground. If all of the levels are above ground, such as ! a terrain following lower coordinate, then the first level above ground is index #2. DO i = its , MIN (ide-1 , ite ) level_above_sfc(i) = -1 IF ( p(i,kts+1) .LT. psfc(i,j) ) THEN level_above_sfc(i) = kts+1 ELSE find_k : DO k = kts+1,kte-1 IF ( ( p(i,k )-psfc(i,j) .GE. 0. ) .AND. & ( p(i,k+1)-psfc(i,j) .LT. 0. ) ) THEN level_above_sfc(i) = k+1 EXIT find_k END IF END DO find_k IF ( level_above_sfc(i) .EQ. -1 ) THEN print *,'i,j = ',i,j print *,'p = ',p(i,:) print *,'p sfc = ',psfc(i,j) CALL wrf_error_fatal ( 'Could not find level above ground') END IF END IF END DO DO i = its , MIN (ide-1 , ite ) ! Account for the moisture above the ground. pd(i,kte) = p(i,kte) DO k = kte-1,level_above_sfc(i),-1 rhobar = ( p(i,k ) / ( Rd * t(i,k ) ) + & p(i,k+1) / ( Rd * t(i,k+1) ) ) * 0.5 qbar = ( q(i,k ) + q(i,k+1) ) * 0.5 dz = ght(i,k+1) - ght(i,k) intq(i,j) = intq(i,j) + g * qbar * rhobar / (1. + qbar) * dz pd(i,k) = p(i,k) - intq(i,j) END DO ! Account for the moisture between the surface and the first level up. IF ( ( p(i,level_above_sfc(i)-1)-psfc(i,j) .GE. 0. ) .AND. & ( p(i,level_above_sfc(i) )-psfc(i,j) .LT. 0. ) .AND. & ( level_above_sfc(i) .GT. kts ) ) THEN p1 = psfc(i,j) p2 = p(i,level_above_sfc(i)) t1 = tsfc(i,j) t2 = t(i,level_above_sfc(i)) q1 = qsfc(i,j) q2 = q(i,level_above_sfc(i)) z1 = zsfc(i,j) z2 = ght(i,level_above_sfc(i)) rhobar = ( p1 / ( Rd * t1 ) + & p2 / ( Rd * t2 ) ) * 0.5 qbar = ( q1 + q2 ) * 0.5 dz = z2 - z1 IF ( dz .GT. 0.1 ) THEN intq(i,j) = intq(i,j) + g * qbar * rhobar / (1. + qbar) * dz END IF ! Fix the underground values. DO k = level_above_sfc(i)-1,kts+1,-1 pd(i,k) = p(i,k) - intq(i,j) END DO END IF pd(i,kts) = psfc(i,j) - intq(i,j) END DO IF ( upside_down ) THEN DO i = its , MIN (ide-1 , ite ) pd_out(i,kts,j) = pd(i,kts) DO k = kts+1,kte pd_out(i,kte+2-k,j) = pd(i,k) END DO END DO ELSE DO i = its , MIN (ide-1 , ite ) DO k = kts,kte pd_out(i,k,j) = pd(i,k) END DO END DO END IF END DO END SUBROUTINE integ_moist !--------------------------------------------------------------------- SUBROUTINE rh_to_mxrat (rh, t, p, q , wrt_liquid , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte LOGICAL , INTENT(IN) :: wrt_liquid REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(IN) :: p , t REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(INOUT) :: rh REAL , DIMENSION(ims:ime,kms:kme,jms:jme) , INTENT(OUT) :: q ! Local vars INTEGER :: i , j , k REAL :: ew , q1 , t1 REAL, PARAMETER :: T_REF = 0.0 REAL, PARAMETER :: MW_AIR = 28.966 REAL, PARAMETER :: MW_VAP = 18.0152 REAL, PARAMETER :: A0 = 6.107799961 REAL, PARAMETER :: A1 = 4.436518521e-01 REAL, PARAMETER :: A2 = 1.428945805e-02 REAL, PARAMETER :: A3 = 2.650648471e-04 REAL, PARAMETER :: A4 = 3.031240396e-06 REAL, PARAMETER :: A5 = 2.034080948e-08 REAL, PARAMETER :: A6 = 6.136820929e-11 REAL, PARAMETER :: ES0 = 6.1121 REAL, PARAMETER :: C1 = 9.09718 REAL, PARAMETER :: C2 = 3.56654 REAL, PARAMETER :: C3 = 0.876793 REAL, PARAMETER :: EIS = 6.1071 REAL :: RHS REAL, PARAMETER :: TF = 273.16 REAL :: TK REAL :: ES REAL :: QS REAL, PARAMETER :: EPS = 0.622 REAL, PARAMETER :: SVP1 = 0.6112 REAL, PARAMETER :: SVP2 = 17.67 REAL, PARAMETER :: SVP3 = 29.65 REAL, PARAMETER :: SVPT0 = 273.15 ! This subroutine computes mixing ratio (q, kg/kg) from basic variables ! pressure (p, Pa), temperature (t, K) and relative humidity (rh, 1-100%). ! The reference temperature (t_ref, C) is used to describe the temperature ! at which the liquid and ice phase change occurs. DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte DO i = its , MIN (ide-1 , ite ) rh(i,k,j) = MIN ( MAX ( rh(i,k,j) , 0. ) , 100. ) END DO END DO END DO IF ( wrt_liquid ) THEN DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte DO i = its , MIN (ide-1 , ite ) ! es is reduced by RH here to avoid problems in low-pressure cases es=.01*rh(i,k,j)*svp1*10.*EXP(svp2*(t(i,k,j)-svpt0)/(t(i,k,j)-svp3)) IF (es .ge. p(i,k,j)/100.)THEN q(i,k,j)=1.0 print *,'warning: vapor pressure exceeds total pressure ' print *,'setting mixing ratio to 1' ELSE q(i,k,j)=eps*es/(p(i,k,j)/100.-es) ENDIF END DO END DO END DO ELSE DO j = jts , MIN ( jde-1 , jte ) DO k = kts , kte DO i = its , MIN (ide-1 , ite ) t1 = t(i,k,j) - 273.16 ! Obviously dry. IF ( t1 .lt. -200. ) THEN q(i,k,j) = 0 ELSE ! First compute the ambient vapor pressure of water IF ( ( t1 .GE. t_ref ) .AND. ( t1 .GE. -47.) ) THEN ! liq phase ESLO ew = a0 + t1 * (a1 + t1 * (a2 + t1 * (a3 + t1 * (a4 + t1 * (a5 + t1 * a6))))) ELSE IF ( ( t1 .GE. t_ref ) .AND. ( t1 .LT. -47. ) ) then !liq phas poor ES ew = es0 * exp(17.67 * t1 / ( t1 + 243.5)) ELSE tk = t(i,k,j) rhs = -c1 * (tf / tk - 1.) - c2 * alog10(tf / tk) + & c3 * (1. - tk / tf) + alog10(eis) ew = 10. ** rhs END IF ! Now sat vap pres obtained compute local vapor pressure ew = MAX ( ew , 0. ) * rh(i,k,j) * 0.01 ! Now compute the specific humidity using the partial vapor ! pressures of water vapor (ew) and dry air (p-ew). The ! constants assume that the pressure is in hPa, so we divide ! the pressures by 100. q1 = mw_vap * ew q1 = q1 / (q1 + mw_air * (p(i,k,j)/100. - ew)) q(i,k,j) = q1 / (1. - q1 ) END IF END DO END DO END DO END IF END SUBROUTINE rh_to_mxrat !--------------------------------------------------------------------- SUBROUTINE compute_eta ( znw , & eta_levels , max_eta , max_dz , & p_top , g , p00 , cvpm , a , r_d , cp , t00 , p1000mb , t0 , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Compute eta levels, either using given values from the namelist (hardly ! a computation, yep, I know), or assuming a constant dz above the PBL, ! knowing p_top and the number of eta levels. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , INTENT(IN) :: max_dz REAL , INTENT(IN) :: p_top , g , p00 , cvpm , a , r_d , cp , t00 , p1000mb , t0 INTEGER , INTENT(IN) :: max_eta REAL , DIMENSION (max_eta) , INTENT(IN) :: eta_levels REAL , DIMENSION (kts:kte) , INTENT(OUT) :: znw ! Local vars INTEGER :: k REAL :: mub , t_init , p_surf , pb, ztop, ztop_pbl , dz , temp REAL , DIMENSION(kts:kte) :: dnw INTEGER , PARAMETER :: prac_levels = 17 INTEGER :: loop , loop1 REAL , DIMENSION(prac_levels) :: znw_prac , znu_prac , dnw_prac REAL , DIMENSION(kts:kte) :: alb , phb ! Gee, do the eta levels come in from the namelist? IF ( ABS(eta_levels(1)+1.) .GT. 0.0000001 ) THEN ! Check to see if the array is oriented OK, we can easily fix an upside down oops. IF ( ( ABS(eta_levels(1 )-1.) .LT. 0.0000001 ) .AND. & ( ABS(eta_levels(kde)-0.) .LT. 0.0000001 ) ) THEN DO k = kds+1 , kde-1 znw(k) = eta_levels(k) END DO znw( 1) = 1. znw(kde) = 0. ELSE IF ( ( ABS(eta_levels(kde)-1.) .LT. 0.0000001 ) .AND. & ( ABS(eta_levels(1 )-0.) .LT. 0.0000001 ) ) THEN DO k = kds+1 , kde-1 znw(k) = eta_levels(kde+1-k) END DO znw( 1) = 1. znw(kde) = 0. ELSE CALL wrf_error_fatal ( 'First eta level should be 1.0 and the last 0.0 in namelist' ) END IF ! Check to see if the input full-level eta array is monotonic. DO k = kds , kde-1 IF ( znw(k) .LE. znw(k+1) ) THEN PRINT *,'eta on full levels is not monotonic' PRINT *,'eta (',k,') = ',znw(k) PRINT *,'eta (',k+1,') = ',znw(k+1) CALL wrf_error_fatal ( 'Fix non-monotonic "eta_levels" in the namelist.input file' ) END IF END DO ! Compute eta levels assuming a constant delta z above the PBL. ELSE ! Compute top of the atmosphere with some silly levels. We just want to ! integrate to get a reasonable value for ztop. We use the planned PBL-esque ! levels, and then just coarse resolution above that. We know p_top, and we ! have the base state vars. p_surf = p00 znw_prac = (/ 1.000 , 0.993 , 0.983 , 0.970 , 0.954 , 0.934 , 0.909 , & 0.88 , 0.8 , 0.7 , 0.6 , 0.5 , 0.4 , 0.3 , 0.2 , 0.1 , 0.0 /) DO k = 1 , prac_levels - 1 znu_prac(k) = ( znw_prac(k) + znw_prac(k+1) ) * 0.5 dnw_prac(k) = znw_prac(k+1) - znw_prac(k) END DO DO k = 1, prac_levels-1 pb = znu_prac(k)*(p_surf - p_top) + p_top ! temp = MAX ( 200., t00 + A*LOG(pb/p00) ) temp = t00 + A*LOG(pb/p00) t_init = temp*(p00/pb)**(r_d/cp) - t0 alb(k) = (r_d/p1000mb)*(t_init+t0)*(pb/p1000mb)**cvpm END DO ! Base state mu is defined as base state surface pressure minus p_top mub = p_surf - p_top ! Integrate base geopotential, starting at terrain elevation. phb(1) = 0. DO k = 2,prac_levels phb(k) = phb(k-1) - dnw_prac(k-1)*mub*alb(k-1) END DO ! So, now we know the model top in meters. Get the average depth above the PBL ! of each of the remaining levels. We are going for a constant delta z thickness. ztop = phb(prac_levels) / g ztop_pbl = phb(8 ) / g dz = ( ztop - ztop_pbl ) / REAL ( kde - 8 ) ! Standard levels near the surface so no one gets in trouble. DO k = 1 , 8 znw(k) = znw_prac(k) END DO ! Using d phb(k)/ d eta(k) = -mub * alb(k), eqn 2.9 ! Skamarock et al, NCAR TN 468. Use full levels, so ! use twice the thickness. DO k = 8, kte-1-2 pb = znw(k) * (p_surf - p_top) + p_top ! temp = MAX ( 200., t00 + A*LOG(pb/p00) ) temp = t00 + A*LOG(pb/p00) t_init = temp*(p00/pb)**(r_d/cp) - t0 alb(k) = (r_d/p1000mb)*(t_init+t0)*(pb/p1000mb)**cvpm znw(k+1) = znw(k) - dz*g / ( mub*alb(k) ) END DO znw(kte-2) = 0.000 ! There is some iteration. We want the top level, ztop, to be ! consistent with the delta z, and we want the half level values ! to be consistent with the eta levels. The inner loop to 10 gets ! the eta levels very accurately, but has a residual at the top, due ! to dz changing. We reset dz five times, and then things seem OK. DO loop1 = 1 , 5 DO loop = 1 , 10 DO k = 8, kte-1-2 pb = (znw(k)+znw(k+1))*0.5 * (p_surf - p_top) + p_top ! temp = MAX ( 200., t00 + A*LOG(pb/p00) ) temp = t00 + A*LOG(pb/p00) t_init = temp*(p00/pb)**(r_d/cp) - t0 alb(k) = (r_d/p1000mb)*(t_init+t0)*(pb/p1000mb)**cvpm znw(k+1) = znw(k) - dz*g / ( mub*alb(k) ) END DO IF ( ( loop1 .EQ. 5 ) .AND. ( loop .EQ. 10 ) ) THEN print *,'Converged znw(kte) should be about 0.0 = ',znw(kte-2) END IF znw(kte-2) = 0.000 END DO ! Here is where we check the eta levels values we just computed. DO k = 1, kde-1-2 pb = (znw(k)+znw(k+1))*0.5 * (p_surf - p_top) + p_top ! temp = MAX ( 200., t00 + A*LOG(pb/p00) ) temp = t00 + A*LOG(pb/p00) t_init = temp*(p00/pb)**(r_d/cp) - t0 alb(k) = (r_d/p1000mb)*(t_init+t0)*(pb/p1000mb)**cvpm END DO phb(1) = 0. DO k = 2,kde-2 phb(k) = phb(k-1) - (znw(k)-znw(k-1)) * mub*alb(k-1) END DO ! Reset the model top and the dz, and iterate. ztop = phb(kde-2)/g ztop_pbl = phb(8)/g dz = ( ztop - ztop_pbl ) / REAL ( (kde-2) - 8 ) END DO IF ( dz .GT. max_dz ) THEN print *,'z (m) = ',phb(1)/g do k = 2 ,kte-2 print *,'z (m) and dz (m) = ',phb(k)/g,(phb(k)-phb(k-1))/g end do print *,'dz (m) above fixed eta levels = ',dz print *,'namelist max_dz (m) = ',max_dz print *,'namelist p_top (Pa) = ',p_top CALL wrf_debug ( 0, 'You need one of three things:' ) CALL wrf_debug ( 0, '1) More eta levels to reduce the dz: e_vert' ) CALL wrf_debug ( 0, '2) A lower p_top so your total height is reduced: p_top_requested') CALL wrf_debug ( 0, '3) Increase the maximum allowable eta thickness: max_dz') CALL wrf_debug ( 0, 'All are namelist options') CALL wrf_error_fatal ( 'dz above fixed eta levels is too large') END IF ! Add those 2 levels back into the middle, just above the 8 levels ! that semi define a boundary layer. After we open up the levels, ! then we just linearly interpolate in znw. So now levels 1-8 are ! specified as the fixed boundary layer levels given in this routine. ! The top levels, 12 through kte are those computed. The middle ! levels 9, 10, and 11 are equi-spaced in znw, and are each 1/2 the ! the znw thickness of levels 11 through 12. DO k = kte-2 , 9 , -1 znw(k+2) = znw(k) END DO znw( 9) = 0.75 * znw( 8) + 0.25 * znw(12) znw(10) = 0.50 * znw( 8) + 0.50 * znw(12) znw(11) = 0.25 * znw( 8) + 0.75 * znw(12) END IF END SUBROUTINE compute_eta !--------------------------------------------------------------------- SUBROUTINE monthly_min_max ( field_in , field_min , field_max , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Plow through each month, find the max, min values for each i,j. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , DIMENSION(ims:ime,12,jms:jme) , INTENT(IN) :: field_in REAL , DIMENSION(ims:ime, jms:jme) , INTENT(OUT) :: field_min , field_max ! Local vars INTEGER :: i , j , l REAL :: minner , maxxer DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) minner = field_in(i,1,j) maxxer = field_in(i,1,j) DO l = 2 , 12 IF ( field_in(i,l,j) .LT. minner ) THEN minner = field_in(i,l,j) END IF IF ( field_in(i,l,j) .GT. maxxer ) THEN maxxer = field_in(i,l,j) END IF END DO field_min(i,j) = minner field_max(i,j) = maxxer END DO END DO END SUBROUTINE monthly_min_max !--------------------------------------------------------------------- SUBROUTINE monthly_interp_to_date ( field_in , date_str , field_out , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Linrarly in time interpolate data to a current valid time. The data is ! assumed to come in "monthly", valid at the 15th of every month. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte CHARACTER (LEN=24) , INTENT(IN) :: date_str REAL , DIMENSION(ims:ime,12,jms:jme) , INTENT(IN) :: field_in REAL , DIMENSION(ims:ime, jms:jme) , INTENT(OUT) :: field_out ! Local vars INTEGER :: i , j , l INTEGER , DIMENSION(0:13) :: middle INTEGER :: target_julyr , target_julday , target_date INTEGER :: julyr , julday , int_month , month1 , month2 REAL :: gmt CHARACTER (LEN=4) :: yr CHARACTER (LEN=2) :: mon , day15 WRITE(day15,FMT='(I2.2)') 15 DO l = 1 , 12 WRITE(mon,FMT='(I2.2)') l CALL get_julgmt ( date_str(1:4)//'-'//mon//'-'//day15//'_'//'00:00:00.0000' , julyr , julday , gmt ) middle(l) = julyr*1000 + julday END DO l = 0 middle(l) = middle( 1) - 31 l = 13 middle(l) = middle(12) + 31 CALL get_julgmt ( date_str , target_julyr , target_julday , gmt ) target_date = target_julyr * 1000 + target_julday find_month : DO l = 0 , 12 IF ( ( middle(l) .LT. target_date ) .AND. ( middle(l+1) .GE. target_date ) ) THEN DO j = jts , MIN ( jde-1 , jte ) DO i = its , MIN (ide-1 , ite ) int_month = l IF ( ( int_month .EQ. 0 ) .OR. ( int_month .EQ. 12 ) ) THEN month1 = 12 month2 = 1 ELSE month1 = int_month month2 = month1 + 1 END IF field_out(i,j) = ( field_in(i,month2,j) * ( target_date - middle(l) ) + & field_in(i,month1,j) * ( middle(l+1) - target_date ) ) / & ( middle(l+1) - middle(l) ) END DO END DO EXIT find_month END IF END DO find_month END SUBROUTINE monthly_interp_to_date !--------------------------------------------------------------------- SUBROUTINE sfcprs (t, q, height, pslv, ter, avgsfct, p, & psfc, ez_method, & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Computes the surface pressure using the input height, ! temperature and q (already computed from relative ! humidity) on p surfaces. Sea level pressure is used ! to extrapolate a first guess. IMPLICIT NONE REAL, PARAMETER :: g = 9.8 REAL, PARAMETER :: gamma = 6.5E-3 REAL, PARAMETER :: pconst = 10000.0 REAL, PARAMETER :: Rd = 287. REAL, PARAMETER :: TC = 273.15 + 17.5 REAL, PARAMETER :: gammarg = gamma * Rd / g REAL, PARAMETER :: rov2 = Rd / 2. INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte LOGICAL , INTENT ( IN ) :: ez_method REAL , DIMENSION (ims:ime,kms:kme,jms:jme) , INTENT(IN ):: t, q, height, p REAL , DIMENSION (ims:ime, jms:jme) , INTENT(IN ):: pslv , ter, avgsfct REAL , DIMENSION (ims:ime, jms:jme) , INTENT(OUT):: psfc INTEGER :: i INTEGER :: j INTEGER :: k INTEGER , DIMENSION (its:ite,jts:jte) :: k500 , k700 , k850 LOGICAL :: l1 LOGICAL :: l2 LOGICAL :: l3 LOGICAL :: OK REAL :: gamma78 ( its:ite,jts:jte ) REAL :: gamma57 ( its:ite,jts:jte ) REAL :: ht ( its:ite,jts:jte ) REAL :: p1 ( its:ite,jts:jte ) REAL :: t1 ( its:ite,jts:jte ) REAL :: t500 ( its:ite,jts:jte ) REAL :: t700 ( its:ite,jts:jte ) REAL :: t850 ( its:ite,jts:jte ) REAL :: tfixed ( its:ite,jts:jte ) REAL :: tsfc ( its:ite,jts:jte ) REAL :: tslv ( its:ite,jts:jte ) ! We either compute the surface pressure from a time averaged surface temperature ! (what we will call the "easy way"), or we try to remove the diurnal impact on the ! surface temperature (what we will call the "other way"). Both are essentially ! corrections to a sea level pressure with a high-resolution topography field. IF ( ez_method ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) psfc(i,j) = pslv(i,j) * ( 1.0 + gamma * ter(i,j) / avgsfct(i,j) ) ** ( - g / ( Rd * gamma ) ) END DO END DO ELSE ! Find the locations of the 850, 700 and 500 mb levels. k850 = 0 ! find k at: P=850 k700 = 0 ! P=700 k500 = 0 ! P=500 i = its j = jts DO k = kts+1 , kte IF (NINT(p(i,k,j)) .EQ. 85000) THEN k850(i,j) = k ELSE IF (NINT(p(i,k,j)) .EQ. 70000) THEN k700(i,j) = k ELSE IF (NINT(p(i,k,j)) .EQ. 50000) THEN k500(i,j) = k END IF END DO IF ( ( k850(i,j) .EQ. 0 ) .OR. ( k700(i,j) .EQ. 0 ) .OR. ( k500(i,j) .EQ. 0 ) ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) psfc(i,j) = pslv(i,j) * ( 1.0 + gamma * ter(i,j) / t(i,1,j) ) ** ( - g / ( Rd * gamma ) ) END DO END DO RETURN #if 0 ! Possibly it is just that we have a generalized vertical coord, so we do not ! have the values exactly. Do a simple assignment to a close vertical level. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) DO k = kts+1 , kte-1 IF ( ( p(i,k,j) - 85000. ) * ( p(i,k+1,j) - 85000. ) .LE. 0.0 ) THEN k850(i,j) = k END IF IF ( ( p(i,k,j) - 70000. ) * ( p(i,k+1,j) - 70000. ) .LE. 0.0 ) THEN k700(i,j) = k END IF IF ( ( p(i,k,j) - 50000. ) * ( p(i,k+1,j) - 50000. ) .LE. 0.0 ) THEN k500(i,j) = k END IF END DO END DO END DO ! If we *still* do not have the k levels, punt. I mean, we did try. OK = .TRUE. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) IF ( ( k850(i,j) .EQ. 0 ) .OR. ( k700(i,j) .EQ. 0 ) .OR. ( k500(i,j) .EQ. 0 ) ) THEN OK = .FALSE. PRINT '(A)','(i,j) = ',i,j,' Error in finding p level for 850, 700 or 500 hPa.' DO K = kts+1 , kte PRINT '(A,I3,A,F10.2,A)','K = ',k,' PRESSURE = ',p(i,k,j),' Pa' END DO PRINT '(A)','Expected 850, 700, and 500 mb values, at least.' END IF END DO END DO IF ( .NOT. OK ) THEN CALL wrf_error_fatal ( 'wrong pressure levels' ) END IF #endif ! We are here if the data is isobaric and we found the levels for 850, 700, ! and 500 mb right off the bat. ELSE DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) k850(i,j) = k850(its,jts) k700(i,j) = k700(its,jts) k500(i,j) = k500(its,jts) END DO END DO END IF ! The 850 hPa level of geopotential height is called something special. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) ht(i,j) = height(i,k850(i,j),j) END DO END DO ! The variable ht is now -ter/ht(850 hPa). The plot thickens. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) ht(i,j) = -ter(i,j) / ht(i,j) END DO END DO ! Make an isothermal assumption to get a first guess at the surface ! pressure. This is to tell us which levels to use for the lapse ! rates in a bit. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) psfc(i,j) = pslv(i,j) * (pslv(i,j) / p(i,k850(i,j),j)) ** ht(i,j) END DO END DO ! Get a pressure more than pconst Pa above the surface - p1. The ! p1 is the top of the level that we will use for our lapse rate ! computations. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) IF ( ( psfc(i,j) - 95000. ) .GE. 0. ) THEN p1(i,j) = 85000. ELSE IF ( ( psfc(i,j) - 70000. ) .GE. 0. ) THEN p1(i,j) = psfc(i,j) - pconst ELSE p1(i,j) = 50000. END IF END DO END DO ! Compute virtual temperatures for k850, k700, and k500 layers. Now ! you see why we wanted Q on pressure levels, it all is beginning ! to make sense. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) t850(i,j) = t(i,k850(i,j),j) * (1. + 0.608 * q(i,k850(i,j),j)) t700(i,j) = t(i,k700(i,j),j) * (1. + 0.608 * q(i,k700(i,j),j)) t500(i,j) = t(i,k500(i,j),j) * (1. + 0.608 * q(i,k500(i,j),j)) END DO END DO ! Compute lapse rates between these three levels. These are ! environmental values for each (i,j). DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) gamma78(i,j) = ALOG(t850(i,j) / t700(i,j)) / ALOG (p(i,k850(i,j),j) / p(i,k700(i,j),j) ) gamma57(i,j) = ALOG(t700(i,j) / t500(i,j)) / ALOG (p(i,k700(i,j),j) / p(i,k500(i,j),j) ) END DO END DO DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) IF ( ( psfc(i,j) - 95000. ) .GE. 0. ) THEN t1(i,j) = t850(i,j) ELSE IF ( ( psfc(i,j) - 85000. ) .GE. 0. ) THEN t1(i,j) = t700(i,j) * (p1(i,j) / (p(i,k700(i,j),j))) ** gamma78(i,j) ELSE IF ( ( psfc(i,j) - 70000. ) .GE. 0.) THEN t1(i,j) = t500(i,j) * (p1(i,j) / (p(i,k500(i,j),j))) ** gamma57(i,j) ELSE t1(i,j) = t500(i,j) ENDIF END DO END DO ! From our temperature way up in the air, we extrapolate down to ! the sea level to get a guess at the sea level temperature. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) tslv(i,j) = t1(i,j) * (pslv(i,j) / p1(i,j)) ** gammarg END DO END DO ! The new surface temperature is computed from the with new sea level ! temperature, just using the elevation and a lapse rate. This lapse ! rate is -6.5 K/km. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) tsfc(i,j) = tslv(i,j) - gamma * ter(i,j) END DO END DO ! A small correction to the sea-level temperature, in case it is too warm. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) tfixed(i,j) = tc - 0.005 * (tsfc(i,j) - tc) ** 2 END DO END DO DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) l1 = tslv(i,j) .LT. tc l2 = tsfc(i,j) .LE. tc l3 = .NOT. l1 IF ( l2 .AND. l3 ) THEN tslv(i,j) = tc ELSE IF ( ( .NOT. l2 ) .AND. l3 ) THEN tslv(i,j) = tfixed(i,j) END IF END DO END DO ! Finally, we can get to the surface pressure. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) p1(i,j) = - ter(i,j) * g / ( rov2 * ( tsfc(i,j) + tslv(i,j) ) ) psfc(i,j) = pslv(i,j) * EXP ( p1(i,j) ) END DO END DO END IF ! Surface pressure and sea-level pressure are the same at sea level. ! DO j = jts , MIN(jde-1,jte) ! DO i = its , MIN(ide-1,ite) ! IF ( ABS ( ter(i,j) ) .LT. 0.1 ) THEN ! psfc(i,j) = pslv(i,j) ! END IF ! END DO ! END DO END SUBROUTINE sfcprs !--------------------------------------------------------------------- SUBROUTINE sfcprs2(t, q, height, psfc_in, ter, avgsfct, p, & psfc, ez_method, & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Computes the surface pressure using the input height, ! temperature and q (already computed from relative ! humidity) on p surfaces. Sea level pressure is used ! to extrapolate a first guess. IMPLICIT NONE REAL, PARAMETER :: g = 9.8 REAL, PARAMETER :: Rd = 287. INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte LOGICAL , INTENT ( IN ) :: ez_method REAL , DIMENSION (ims:ime,kms:kme,jms:jme) , INTENT(IN ):: t, q, height, p REAL , DIMENSION (ims:ime, jms:jme) , INTENT(IN ):: psfc_in , ter, avgsfct REAL , DIMENSION (ims:ime, jms:jme) , INTENT(OUT):: psfc INTEGER :: i INTEGER :: j INTEGER :: k REAL :: tv_sfc_avg , tv_sfc , del_z ! Compute the new surface pressure from the old surface pressure, and a ! known change in elevation at the surface. ! del_z = diff in surface topo, lo-res vs hi-res ! psfc = psfc_in * exp ( g del_z / (Rd Tv_sfc ) ) IF ( ez_method ) THEN DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) tv_sfc_avg = avgsfct(i,j) * (1. + 0.608 * q(i,1,j)) del_z = height(i,1,j) - ter(i,j) psfc(i,j) = psfc_in(i,j) * EXP ( g * del_z / ( Rd * tv_sfc_avg ) ) END DO END DO ELSE DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) tv_sfc = t(i,1,j) * (1. + 0.608 * q(i,1,j)) del_z = height(i,1,j) - ter(i,j) psfc(i,j) = psfc_in(i,j) * EXP ( g * del_z / ( Rd * tv_sfc ) ) END DO END DO END IF END SUBROUTINE sfcprs2 !--------------------------------------------------------------------- SUBROUTINE sfcprs3( height , p , ter , slp , psfc , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) ! Computes the surface pressure by vertically interpolating ! linearly (or log) in z the pressure, to the targeted topography. IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , DIMENSION (ims:ime,kms:kme,jms:jme) , INTENT(IN ):: height, p REAL , DIMENSION (ims:ime, jms:jme) , INTENT(IN ):: ter , slp REAL , DIMENSION (ims:ime, jms:jme) , INTENT(OUT):: psfc INTEGER :: i INTEGER :: j INTEGER :: k LOGICAL :: found_loc REAL :: zl , zu , pl , pu , zm ! Loop over each grid point DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) ! Find the trapping levels found_loc = .FALSE. ! Normal sort of scenario - the model topography is somewhere between ! the height values of 1000 mb and the top of the model. found_k_loc : DO k = kts+1 , kte-2 IF ( ( height(i,k ,j) .LE. ter(i,j) ) .AND. & ( height(i,k+1,j) .GT. ter(i,j) ) ) THEN zl = height(i,k ,j) zu = height(i,k+1,j) zm = ter(i,j) pl = p(i,k ,j) pu = p(i,k+1,j) psfc(i,j) = EXP ( ( LOG(pl) * ( zm - zu ) + LOG(pu) * ( zl - zm ) ) / ( zl - zu ) ) found_loc = .TRUE. EXIT found_k_loc END IF END DO found_k_loc ! Interpolate betwixt slp and the first isobaric level above - this is probably the ! usual thing over the ocean. IF ( .NOT. found_loc ) THEN IF ( slp(i,j) .GE. p(i,2,j) ) THEN zl = 0. zu = height(i,2 ,j) zm = ter(i,j) pl = slp(i,j) pu = p(i,2 ,j) psfc(i,j) = EXP ( ( LOG(pl) * ( zm - zu ) + LOG(pu) * ( zl - zm ) ) / ( zl - zu ) ) found_loc = .TRUE. ELSE found_slp_loc : DO k = kts+1 , kte-2 IF ( ( slp(i,j) .GE. p(i,k+1,j) ) .AND. & ( slp(i,j) .LT. p(i,k ,j) ) ) THEN zl = 0. zu = height(i,k+1,j) zm = ter(i,j) pl = slp(i,j) pu = p(i,k+1,j) psfc(i,j) = EXP ( ( LOG(pl) * ( zm - zu ) + LOG(pu) * ( zl - zm ) ) / ( zl - zu ) ) found_loc = .TRUE. EXIT found_slp_loc END IF END DO found_slp_loc END IF END IF ! Did we do what we wanted done. IF ( .NOT. found_loc ) THEN print *,'i,j = ',i,j print *,'p column = ',p(i,2:,j) print *,'z column = ',height(i,2:,j) print *,'model topo = ',ter(i,j) CALL wrf_error_fatal ( ' probs with sfc p computation ' ) END IF END DO END DO END SUBROUTINE sfcprs3 !--------------------------------------------------------------------- SUBROUTINE filter_topo ( ht_in , xlat , msftx , fft_filter_lat , & ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte ) IMPLICIT NONE INTEGER , INTENT(IN) :: ids , ide , jds , jde , kds , kde , & ims , ime , jms , jme , kms , kme , & its , ite , jts , jte , kts , kte REAL , INTENT(IN) :: fft_filter_lat REAL , DIMENSION(ims:ime,jms:jme) , INTENT(INOUT) :: ht_in REAL , DIMENSION(ims:ime,jms:jme) , INTENT(IN) :: xlat , msftx ! Local vars INTEGER :: i , j , j_lat_pos , j_lat_neg INTEGER :: i_kicker , ik , i1, i2, i3, i4 REAL :: length_scale , sum REAL , DIMENSION(its:ite,jts:jte) :: ht_out ! The filtering is a simple average on a latitude loop. Possibly a LONG list of ! numbers. We assume that ALL of the 2d arrays have been transposed so that ! each patch has the entire domain size of the i-dim local. IF ( ( its .NE. ids ) .OR. ( ite .NE. ide ) ) THEN CALL wrf_error_fatal ( 'filtering assumes all values on X' ) END IF ! Starting at the south pole, we find where the ! grid distance is big enough, then go back a point. Continuing to the ! north pole, we find the first small grid distance. These are the ! computational latitude loops and the associated computational poles. j_lat_neg = 0 j_lat_pos = jde + 1 loop_neg : DO j = jts , MIN(jde-1,jte) IF ( xlat(its,j) .LT. 0.0 ) THEN IF ( ABS(xlat(its,j)) .LT. fft_filter_lat ) THEN j_lat_neg = j - 1 EXIT loop_neg END IF END IF END DO loop_neg loop_pos : DO j = jts , MIN(jde-1,jte) IF ( xlat(its,j) .GT. 0.0 ) THEN IF ( xlat(its,j) .GE. fft_filter_lat ) THEN j_lat_pos = j EXIT loop_pos END IF END IF END DO loop_pos ! Set output values to initial input topo values for whole patch. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) ht_out(i,j) = ht_in(i,j) END DO END DO ! Filter the topo at the negative lats. DO j = j_lat_neg , jts , -1 i_kicker = MIN( MAX ( NINT(msftx(its,j)) , 1 ) , (ide - ids) / 2 ) print *,'j = ' , j, ', kicker = ',i_kicker DO i = its , MIN(ide-1,ite) IF ( ( i - i_kicker .GE. its ) .AND. ( i + i_kicker .LE. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i+ik,j) + ht_in(i-ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) ELSE IF ( ( i - i_kicker .LT. its ) .AND. ( i + i_kicker .LE. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i+ik,j) END DO i1 = i - i_kicker + ide -1 i2 = ide-1 i3 = ids i4 = i-1 DO ik = i1 , i2 sum = sum + ht_in(ik,j) END DO DO ik = i3 , i4 sum = sum + ht_in(ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) ELSE IF ( ( i - i_kicker .GE. its ) .AND. ( i + i_kicker .GT. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i-ik,j) END DO i1 = i+1 i2 = ide-1 i3 = ids i4 = ids + ( i_kicker+i ) - ide DO ik = i1 , i2 sum = sum + ht_in(ik,j) END DO DO ik = i3 , i4 sum = sum + ht_in(ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) END IF END DO END DO ! Filter the topo at the positive lats. DO j = j_lat_pos , MIN(jde-1,jte) i_kicker = MIN( MAX ( NINT(msftx(its,j)) , 1 ) , (ide - ids) / 2 ) print *,'j = ' , j, ', kicker = ',i_kicker DO i = its , MIN(ide-1,ite) IF ( ( i - i_kicker .GE. its ) .AND. ( i + i_kicker .LE. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i+ik,j) + ht_in(i-ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) ELSE IF ( ( i - i_kicker .LT. its ) .AND. ( i + i_kicker .LE. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i+ik,j) END DO i1 = i - i_kicker + ide -1 i2 = ide-1 i3 = ids i4 = i-1 DO ik = i1 , i2 sum = sum + ht_in(ik,j) END DO DO ik = i3 , i4 sum = sum + ht_in(ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) ELSE IF ( ( i - i_kicker .GE. its ) .AND. ( i + i_kicker .GT. ide-1 ) ) THEN sum = 0.0 DO ik = 1 , i_kicker sum = sum + ht_in(i-ik,j) END DO i1 = i+1 i2 = ide-1 i3 = ids i4 = ids + ( i_kicker+i ) - ide DO ik = i1 , i2 sum = sum + ht_in(ik,j) END DO DO ik = i3 , i4 sum = sum + ht_in(ik,j) END DO ht_out(i,j) = ( ht_in(i,j) + sum ) / REAL ( 2 * i_kicker + 1 ) END IF END DO END DO ! Set output values to initial input topo values for whole patch. DO j = jts , MIN(jde-1,jte) DO i = its , MIN(ide-1,ite) ht_in(i,j) = ht_out(i,j) END DO END DO END SUBROUTINE filter_topo !--------------------------------------------------------------------- SUBROUTINE init_module_initialize END SUBROUTINE init_module_initialize !--------------------------------------------------------------------- END MODULE module_initialize_real #endif