ERF
Energy Research and Forecasting: An Atmospheric Modeling Code
Functions
ERF_Substep_MT.cpp File Reference
#include <ERF_TI_fast_headers.H>
Include dependency graph for ERF_Substep_MT.cpp:

Functions

void erf_substep_MT (int step, int, int level, int finest_level, Vector< MultiFab > &S_slow_rhs, const Vector< MultiFab > &S_prev, Vector< MultiFab > &S_stg_data, const MultiFab &S_stg_prim, const MultiFab &qt, const MultiFab &pi_stage, const MultiFab &fast_coeffs, Vector< MultiFab > &S_data, MultiFab &lagged_delta_rt, MultiFab &avg_xmom, MultiFab &avg_ymom, MultiFab &avg_zmom, const MultiFab &cc_src, const MultiFab &xmom_src, const MultiFab &ymom_src, const MultiFab &zmom_src, const Geometry geom, const Real gravity, const bool use_lagged_delta_rt, std::unique_ptr< MultiFab > &z_t_rk, const MultiFab *z_t_pert, std::unique_ptr< MultiFab > &z_phys_nd_old, std::unique_ptr< MultiFab > &z_phys_nd_new, std::unique_ptr< MultiFab > &z_phys_nd_stg, std::unique_ptr< MultiFab > &detJ_cc_old, std::unique_ptr< MultiFab > &detJ_cc_new, std::unique_ptr< MultiFab > &detJ_cc_stg, const double dtau_d, const Real beta_s, const Real facinv, Vector< std::unique_ptr< MultiFab >> &mapfac, YAFluxRegister *fr_as_crse, YAFluxRegister *fr_as_fine, bool l_use_moisture, bool l_reflux, bool, const Real *sinesq_stag_d, const Real l_damp_coef)
 

Function Documentation

◆ erf_substep_MT()

void erf_substep_MT ( int  step,
int  ,
int  level,
int  finest_level,
Vector< MultiFab > &  S_slow_rhs,
const Vector< MultiFab > &  S_prev,
Vector< MultiFab > &  S_stg_data,
const MultiFab &  S_stg_prim,
const MultiFab &  qt,
const MultiFab &  pi_stage,
const MultiFab &  fast_coeffs,
Vector< MultiFab > &  S_data,
MultiFab &  lagged_delta_rt,
MultiFab &  avg_xmom,
MultiFab &  avg_ymom,
MultiFab &  avg_zmom,
const MultiFab &  cc_src,
const MultiFab &  xmom_src,
const MultiFab &  ymom_src,
const MultiFab &  zmom_src,
const Geometry  geom,
const Real  gravity,
const bool  use_lagged_delta_rt,
std::unique_ptr< MultiFab > &  z_t_rk,
const MultiFab *  z_t_pert,
std::unique_ptr< MultiFab > &  z_phys_nd_old,
std::unique_ptr< MultiFab > &  z_phys_nd_new,
std::unique_ptr< MultiFab > &  z_phys_nd_stg,
std::unique_ptr< MultiFab > &  detJ_cc_old,
std::unique_ptr< MultiFab > &  detJ_cc_new,
std::unique_ptr< MultiFab > &  detJ_cc_stg,
const double  dtau_d,
const Real  beta_s,
const Real  facinv,
Vector< std::unique_ptr< MultiFab >> &  mapfac,
YAFluxRegister *  fr_as_crse,
YAFluxRegister *  fr_as_fine,
bool  l_use_moisture,
bool  l_reflux,
bool  ,
const Real *  sinesq_stag_d,
const Real  l_damp_coef 
)

Function for computing the fast RHS with moving terrain

Parameters
[in]stepwhich fast time step within each Runge-Kutta step
[in]nrkwhich Runge-Kutta step
[in]levellevel of resolution
[in]finest_levelfinest level of resolution
[in]S_slow_rhsslow RHS computed in erf_slow_rhs_pre
[in]S_prevprevious solution
[in]S_stg_datasolution at previous RK stage
[in]S_stg_primprimitive variables at previous RK stage
[in]pi_stageExner function at previous RK stage
[in]fast_coeffscoefficients for the tridiagonal solve used in the fast integrator
[out]S_datacurrent solution
[in,out]lagged_delta_rt
[in,out]avg_xmomtime-averaged x-momentum to be used for updating slow variables
[in,out]avg_ymomtime-averaged y-momentum to be used for updating slow variables
[in,out]avg_zmomtime-averaged z-momentum to be used for updating slow variables
[in]cc_srcsource terms for conserved variables
[in]xmom_srcsource terms for x-momentum
[in]ymom_srcsource terms for y-momentum
[in]zmom_srcsource terms for z-momentum
[in]geomcontainer for geometric information
[in]gravityMagnitude of gravity
[in]use_lagged_delta_rtdefine lagged_delta_rt for our next step
[in]z_t_rkrate of change of grid height – only relevant for moving terrain
[in]z_t_pertrate of change of grid height – interpolated between RK stages
[in]z_phys_nd_oldheight coordinate at nodes at old time
[in]z_phys_nd_newheight coordinate at nodes at new time
[in]z_phys_nd_stgheight coordinate at nodes at previous stage
[in]detJ_cc_oldJacobian of the metric transformation at old time
[in]detJ_cc_newJacobian of the metric transformation at new time
[in]detJ_cc_stgJacobian of the metric transformation at previous stage
[in]dtaufast time step
[in]beta_sCoefficient which determines how implicit vs explicit the solve is
[in]facinvinverse factor for time-averaging the momenta
[in]mapfacvector of map factors
[in,out]fr_as_crseYAFluxRegister at level l at level l / l+1 interface
[in,out]fr_as_fineYAFluxRegister at level l at level l-1 / l interface
[in]l_use_moisture
[in]l_refluxshould we add fluxes to the FluxRegisters?
[in]l_damp_coef
89 {
90  BL_PROFILE_REGION("erf_substep_MT()");
91 
92  Real dtau = static_cast<Real>(dtau_d);
93 
94  Real beta_1 = myhalf * (one - beta_s); // multiplies explicit terms
95  Real beta_2 = myhalf * (one + beta_s); // multiplies implicit terms
96 
97  // How much do we project forward the (rho theta) that is used in the horizontal momentum equations
98  Real beta_d = Real(0.1);
99 
100  Real RvOverRd = R_v / R_d;
101 
102  bool l_rayleigh_impl_for_w = (sinesq_stag_d != nullptr);
103 
104  const Real* dx = geom.CellSize();
105  const GpuArray<Real, AMREX_SPACEDIM> dxInv = geom.InvCellSizeArray();
106 
107  Real dxi = dxInv[0];
108  Real dyi = dxInv[1];
109  Real dzi = dxInv[2];
110 
111  MultiFab coeff_A_mf(fast_coeffs, make_alias, 0, 1);
112  MultiFab inv_coeff_B_mf(fast_coeffs, make_alias, 1, 1);
113  MultiFab coeff_C_mf(fast_coeffs, make_alias, 2, 1);
114  MultiFab coeff_P_mf(fast_coeffs, make_alias, 3, 1);
115  MultiFab coeff_Q_mf(fast_coeffs, make_alias, 4, 1);
116 
117  // *************************************************************************
118  // Set gravity as a vector
119  const Array<Real,AMREX_SPACEDIM> grav{zero, zero, -gravity};
120  const GpuArray<Real,AMREX_SPACEDIM> grav_gpu{grav[0], grav[1], grav[2]};
121 
122  MultiFab extrap(S_data[IntVars::cons].boxArray(),S_data[IntVars::cons].DistributionMap(),1,1);
123 
124  MultiFab Omega(S_data[IntVars::zmom].boxArray(), S_data[IntVars::zmom].DistributionMap(), 1, 1);
125 
126  // *************************************************************************
127  // Define updates in the current RK stg
128  // *************************************************************************
129 #ifdef _OPENMP
130 #pragma omp parallel if (Gpu::notInLaunchRegion())
131 #endif
132  {
133  FArrayBox temp_rhs_fab;
134 
135  FArrayBox RHS_fab;
136  FArrayBox soln_fab;
137 
138  std::array<FArrayBox,AMREX_SPACEDIM> flux;
139 
140  // NOTE: we leave tiling off here for efficiency -- to make this loop work with tiling
141  // will require additional changes
142  for ( MFIter mfi(S_stg_data[IntVars::cons],false); mfi.isValid(); ++mfi)
143  {
144  Box bx = mfi.tilebox();
145  Box tbx = surroundingNodes(bx,0);
146  Box tby = surroundingNodes(bx,1);
147  Box tbz = surroundingNodes(bx,2);
148 
149  Box vbx = mfi.validbox();
150  const auto& vbx_hi = ubound(vbx);
151 
152  const Array4<Real const>& xmom_src_arr = xmom_src.const_array(mfi);
153  const Array4<Real const>& ymom_src_arr = ymom_src.const_array(mfi);
154  const Array4<Real const>& zmom_src_arr = zmom_src.const_array(mfi);
155  const Array4<Real const>& cc_src_arr = cc_src.const_array(mfi);
156 
157  const Array4<const Real> & stg_cons = S_stg_data[IntVars::cons].const_array(mfi);
158  const Array4<const Real> & stg_xmom = S_stg_data[IntVars::xmom].const_array(mfi);
159  const Array4<const Real> & stg_ymom = S_stg_data[IntVars::ymom].const_array(mfi);
160  const Array4<const Real> & stg_zmom = S_stg_data[IntVars::zmom].const_array(mfi);
161  const Array4<const Real> & prim = S_stg_prim.const_array(mfi);
162  const Array4<const Real> & qt_arr = qt.const_array(mfi);
163 
164  const Array4<const Real>& slow_rhs_cons = S_slow_rhs[IntVars::cons].const_array(mfi);
165  const Array4<const Real>& slow_rhs_rho_u = S_slow_rhs[IntVars::xmom].const_array(mfi);
166  const Array4<const Real>& slow_rhs_rho_v = S_slow_rhs[IntVars::ymom].const_array(mfi);
167  const Array4<const Real>& slow_rhs_rho_w = S_slow_rhs[IntVars::zmom].const_array(mfi);
168 
169  const Array4<Real>& cur_cons = S_data[IntVars::cons].array(mfi);
170  const Array4<Real>& cur_xmom = S_data[IntVars::xmom].array(mfi);
171  const Array4<Real>& cur_ymom = S_data[IntVars::ymom].array(mfi);
172  const Array4<Real>& cur_zmom = S_data[IntVars::zmom].array(mfi);
173 
174  const Array4<Real>& lagged = lagged_delta_rt.array(mfi);
175 
176  const Array4<const Real>& prev_cons = S_prev[IntVars::cons].const_array(mfi);
177  const Array4<const Real>& prev_xmom = S_prev[IntVars::xmom].const_array(mfi);
178  const Array4<const Real>& prev_ymom = S_prev[IntVars::ymom].const_array(mfi);
179  const Array4<const Real>& prev_zmom = S_prev[IntVars::zmom].const_array(mfi);
180 
181  // These store the advection momenta which we will use to update the slow variables
182  const Array4<Real>& avg_xmom_arr = avg_xmom.array(mfi);
183  const Array4<Real>& avg_ymom_arr = avg_ymom.array(mfi);
184  const Array4<Real>& avg_zmom_arr = avg_zmom.array(mfi);
185 
186  const Array4<const Real>& z_nd_old = z_phys_nd_old->const_array(mfi);
187  const Array4<const Real>& z_nd_new = z_phys_nd_new->const_array(mfi);
188  const Array4<const Real>& z_nd_stg = z_phys_nd_stg->const_array(mfi);
189  const Array4<const Real>& detJ_old = detJ_cc_old->const_array(mfi);
190  const Array4<const Real>& detJ_new = detJ_cc_new->const_array(mfi);
191  const Array4<const Real>& detJ_stg = detJ_cc_stg->const_array(mfi);
192 
193  const Array4<const Real>& z_t_arr = z_t_rk->const_array(mfi);
194  const Array4<const Real>& zp_t_arr = z_t_pert->const_array(mfi);
195 
196  const Array4< Real>& omega_arr = Omega.array(mfi);
197 
198  // Map factors
199  const Array4<const Real>& mf_mx = mapfac[MapFacType::m_x]->const_array(mfi);
200  const Array4<const Real>& mf_my = mapfac[MapFacType::m_y]->const_array(mfi);
201  const Array4<const Real>& mf_ux = mapfac[MapFacType::u_x]->const_array(mfi);
202  const Array4<const Real>& mf_vy = mapfac[MapFacType::v_y]->const_array(mfi);
203 
204  // *********************************************************************
205  // This must be done before we set cur_xmom and cur_ymom, since those
206  // in fact point to the same array as prev_xmom and prev_ymom
207  // *********************************************************************
208  Box gbxo = mfi.nodaltilebox(2);
209  {
210  BL_PROFILE("fast_MT_making_omega");
211  Box gbxo_lo = gbxo; gbxo_lo.setBig(2,0);
212  ParallelFor(gbxo_lo, [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
213  omega_arr(i,j,k) = zero;
214  });
215  Box gbxo_hi = gbxo; gbxo_hi.setSmall(2,gbxo.bigEnd(2));
216  ParallelFor(gbxo_hi, [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
217  omega_arr(i,j,k) = prev_zmom(i,j,k) - stg_zmom(i,j,k) - zp_t_arr(i,j,k);
218  });
219  Box gbxo_mid = gbxo; gbxo_mid.setSmall(2,1); gbxo_mid.setBig(2,gbxo.bigEnd(2)-1);
220  ParallelFor(gbxo_mid, [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
221  omega_arr(i,j,k) =
222  ( OmegaFromW(i,j,k,prev_zmom(i,j,k),prev_xmom,prev_ymom,mf_ux,mf_vy,z_nd_old,dxInv)
223  -OmegaFromW(i,j,k, stg_zmom(i,j,k), stg_xmom, stg_ymom,mf_ux,mf_vy,z_nd_old,dxInv) )
224  - zp_t_arr(i,j,k);
225  });
226  } // end profile
227  // *********************************************************************
228 
229  const Array4<const Real>& pi_stage_ca = pi_stage.const_array(mfi);
230 
231  const Array4<Real>& theta_extrap = extrap.array(mfi);
232 
233  // Note: it is important to grow the tilebox rather than use growntilebox because
234  // we need to fill the ghost cells of the tilebox so we can use them below
235  Box gbx = mfi.tilebox(); gbx.grow(1);
236  Box gtbx = mfi.nodaltilebox(0); gtbx.grow(1); gtbx.setSmall(2,0);
237  Box gtby = mfi.nodaltilebox(1); gtby.grow(1); gtby.setSmall(2,0);
238 
239  if (step == 0) {
240  ParallelFor(gbx,
241  [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
242  cur_cons(i,j,k,Rho_comp) = prev_cons(i,j,k,Rho_comp);
243  cur_cons(i,j,k,RhoTheta_comp) = prev_cons(i,j,k,RhoTheta_comp);
244 
245  Real delta_rt = cur_cons(i,j,k,RhoTheta_comp) - stg_cons(i,j,k,RhoTheta_comp);
246  theta_extrap(i,j,k) = delta_rt;
247 
248  // NOTE: qv is not changing over the fast steps so we use the stage data
249  Real qv = (l_use_moisture) ? prim(i,j,k,PrimQ1_comp) : zero;
250  theta_extrap(i,j,k) *= (one + RvOverRd*qv);
251 
252  // We define lagged_delta_rt for our next step as the current delta_rt
253  lagged(i,j,k) = delta_rt;
254  });
255  } else if (use_lagged_delta_rt) {
256  // This is the default for cases with no or static terrain
257  ParallelFor(gbx,
258  [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
259  Real delta_rt = cur_cons(i,j,k,RhoTheta_comp) - stg_cons(i,j,k,RhoTheta_comp);
260  theta_extrap(i,j,k) = delta_rt + beta_d * (delta_rt - lagged(i,j,k));
261 
262  // NOTE: qv is not changing over the fast steps so we use the stage data
263  Real qv = (l_use_moisture) ? prim(i,j,k,PrimQ1_comp) : zero;
264  theta_extrap(i,j,k) *= (one + RvOverRd*qv);
265 
266  // We define lagged_delta_rt for our next step as the current delta_rt
267  lagged(i,j,k) = delta_rt;
268  });
269  } else {
270  // For the moving wave problem, this choice seems more robust
271  ParallelFor(gbx,
272  [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept {
273  theta_extrap(i,j,k) = cur_cons(i,j,k,RhoTheta_comp) - stg_cons(i,j,k,RhoTheta_comp);
274 
275  // NOTE: qv is not changing over the fast steps so we use the stage data
276  Real qv = (l_use_moisture) ? prim(i,j,k,PrimQ1_comp) : zero;
277  theta_extrap(i,j,k) *= (one + RvOverRd*qv);
278  });
279  } // if step
280 
281  RHS_fab.resize (tbz,1, The_Async_Arena());
282  soln_fab.resize (tbz,1, The_Async_Arena());
283  temp_rhs_fab.resize(tbz,2, The_Async_Arena());
284 
285  auto const& RHS_a = RHS_fab.array();
286  auto const& soln_a = soln_fab.array();
287  auto const& temp_rhs_arr = temp_rhs_fab.array();
288 
289  auto const& coeffA_a = coeff_A_mf.array(mfi);
290  auto const& inv_coeffB_a = inv_coeff_B_mf.array(mfi);
291  auto const& coeffC_a = coeff_C_mf.array(mfi);
292  auto const& coeffP_a = coeff_P_mf.array(mfi);
293  auto const& coeffQ_a = coeff_Q_mf.array(mfi);
294 
295  // *********************************************************************
296  // Define updates in the RHS of {x, y, z}-momentum equations
297  // *********************************************************************
298  {
299  BL_PROFILE("substep_xymom_T");
300  ParallelFor(tbx, tby,
301  [=] AMREX_GPU_DEVICE (int i, int j, int k)
302  {
303  // Add (negative) gradient of (rho theta) multiplied by lagged "pi"
304  Real h_xi_old = Compute_h_xi_AtIface(i, j, k, dxInv, z_nd_old);
305  Real h_zeta_old = Compute_h_zeta_AtIface(i, j, k, dxInv, z_nd_old);
306  Real gp_xi = (theta_extrap(i,j,k) - theta_extrap(i-1,j,k)) * dxi;
307  Real gp_zeta_on_iface = (k == 0) ?
308  myhalf * dzi * ( theta_extrap(i-1,j,k+1) + theta_extrap(i,j,k+1)
309  - theta_extrap(i-1,j,k ) - theta_extrap(i,j,k ) ) :
310  fourth * dzi * ( theta_extrap(i-1,j,k+1) + theta_extrap(i,j,k+1)
311  - theta_extrap(i-1,j,k-1) - theta_extrap(i,j,k-1) );
312  Real gpx = h_zeta_old * gp_xi - h_xi_old * gp_zeta_on_iface;
313  gpx *= mf_ux(i,j,0);
314 
315  Real q = (l_use_moisture) ? myhalf * (qt_arr(i-1,j,k) + qt_arr(i,j,k)) : zero;
316 
317  Real pi_c = myhalf * (pi_stage_ca(i-1,j,k) + pi_stage_ca(i ,j,k));
318  Real fast_rhs_rho_u = -Gamma * R_d * pi_c * gpx / (one + q);
319 
320  // We have already scaled the source terms to have the extra factor of dJ
321  cur_xmom(i,j,k) = h_zeta_old * prev_xmom(i,j,k) + dtau * fast_rhs_rho_u
322  + dtau * slow_rhs_rho_u(i,j,k)
323  + dtau * xmom_src_arr(i,j,k);
324  },
325  [=] AMREX_GPU_DEVICE (int i, int j, int k)
326  {
327  // Add (negative) gradient of (rho theta) multiplied by lagged "pi"
328  Real h_eta_old = Compute_h_eta_AtJface(i, j, k, dxInv, z_nd_old);
329  Real h_zeta_old = Compute_h_zeta_AtJface(i, j, k, dxInv, z_nd_old);
330  Real gp_eta = (theta_extrap(i,j,k) -theta_extrap(i,j-1,k)) * dyi;
331  Real gp_zeta_on_jface = (k == 0) ?
332  myhalf * dzi * ( theta_extrap(i,j,k+1) + theta_extrap(i,j-1,k+1)
333  - theta_extrap(i,j,k ) - theta_extrap(i,j-1,k ) ) :
334  fourth * dzi * ( theta_extrap(i,j,k+1) + theta_extrap(i,j-1,k+1)
335  - theta_extrap(i,j,k-1) - theta_extrap(i,j-1,k-1) );
336  Real gpy = h_zeta_old * gp_eta - h_eta_old * gp_zeta_on_jface;
337  gpy *= mf_vy(i,j,0);
338 
339  Real q = (l_use_moisture) ? myhalf * (qt_arr(i,j-1,k) + qt_arr(i,j,k)) : zero;
340 
341  Real pi_c = myhalf * (pi_stage_ca(i,j-1,k) + pi_stage_ca(i,j ,k));
342  Real fast_rhs_rho_v = -Gamma * R_d * pi_c * gpy / (one + q);
343 
344  // We have already scaled the source terms to have the extra factor of dJ
345  cur_ymom(i, j, k) = h_zeta_old * prev_ymom(i,j,k) + dtau * fast_rhs_rho_v
346  + dtau * slow_rhs_rho_v(i,j,k)
347  + dtau * ymom_src_arr(i,j,k);
348  });
349  } // end profile
350 
351  // *************************************************************************
352  // Define flux arrays for use in advection
353  // *************************************************************************
354  for (int dir = 0; dir < AMREX_SPACEDIM; ++dir) {
355  flux[dir].resize(surroundingNodes(bx,dir),2,The_Async_Arena());
356  flux[dir].setVal<RunOn::Device>(0);
357  }
358  const GpuArray<const Array4<Real>, AMREX_SPACEDIM>
359  flx_arr{{AMREX_D_DECL(flux[0].array(), flux[1].array(), flux[2].array())}};
360 
361  // *********************************************************************
362  {
363  BL_PROFILE("fast_T_making_rho_rhs");
364  ParallelFor(bx, [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept
365  {
366  Real h_zeta_stg_xlo = Compute_h_zeta_AtIface(i, j , k, dxInv, z_nd_stg);
367  Real h_zeta_stg_xhi = Compute_h_zeta_AtIface(i+1,j , k, dxInv, z_nd_stg);
368  Real xflux_lo = cur_xmom(i ,j,k) - stg_xmom(i ,j,k)*h_zeta_stg_xlo;
369  Real xflux_hi = cur_xmom(i+1,j,k) - stg_xmom(i+1,j,k)*h_zeta_stg_xhi;
370 
371  Real h_zeta_stg_yhi = Compute_h_zeta_AtJface(i, j+1, k, dxInv, z_nd_stg);
372  Real h_zeta_stg_ylo = Compute_h_zeta_AtJface(i, j , k, dxInv, z_nd_stg);
373  Real yflux_lo = cur_ymom(i,j ,k) - stg_ymom(i,j ,k)*h_zeta_stg_ylo;
374  Real yflux_hi = cur_ymom(i,j+1,k) - stg_ymom(i,j+1,k)*h_zeta_stg_yhi;
375 
376  // NOTE: we are saving the (1/J) weighting for later when we add this to rho and theta
377  temp_rhs_arr(i,j,k,0) = ( xflux_hi - xflux_lo ) * dxi + ( yflux_hi - yflux_lo ) * dyi;
378  temp_rhs_arr(i,j,k,1) = (( xflux_hi * (prim(i,j,k,0) + prim(i+1,j,k,0)) -
379  xflux_lo * (prim(i,j,k,0) + prim(i-1,j,k,0)) ) * dxi +
380  ( yflux_hi * (prim(i,j,k,0) + prim(i,j+1,k,0)) -
381  yflux_lo * (prim(i,j,k,0) + prim(i,j-1,k,0)) ) * dyi) * myhalf;
382 
383  if (l_reflux) {
384  (flx_arr[0])(i,j,k,0) = xflux_lo;
385  (flx_arr[0])(i,j,k,1) = (flx_arr[0])(i ,j,k,0) * myhalf * (prim(i,j,k,0) + prim(i-1,j,k,0));
386 
387  (flx_arr[1])(i,j,k,0) = yflux_lo;
388  (flx_arr[1])(i,j,k,1) = (flx_arr[1])(i,j ,k,0) * myhalf * (prim(i,j,k,0) + prim(i,j-1,k,0));
389 
390  if (i == vbx_hi.x) {
391  (flx_arr[0])(i+1,j,k,0) = xflux_hi;
392  (flx_arr[0])(i+1,j,k,1) = (flx_arr[0])(i+1,j,k,0) * myhalf * (prim(i,j,k,0) + prim(i+1,j,k,0));
393  }
394  if (j == vbx_hi.y) {
395  (flx_arr[1])(i,j+1,k,0) = yflux_hi;
396  (flx_arr[1])(i,j+1,k,1) = (flx_arr[1])(i,j+1,k,0) * myhalf * (prim(i,j,k,0) + prim(i,j+1,k,0));
397  }
398  }
399  });
400  } // end profile
401 
402  ParallelFor(tbx, tby,
403  [=] AMREX_GPU_DEVICE (int i, int j, int k)
404  {
405  Real h_zeta_new = Compute_h_zeta_AtIface(i, j, k, dxInv, z_nd_new);
406  cur_xmom(i, j, k) /= h_zeta_new;
407  avg_xmom_arr(i,j,k) += facinv*(cur_xmom(i,j,k) - stg_xmom(i,j,k));
408  },
409  [=] AMREX_GPU_DEVICE (int i, int j, int k)
410  {
411  Real h_zeta_new = Compute_h_zeta_AtJface(i, j, k, dxInv, z_nd_new);
412  cur_ymom(i, j, k) /= h_zeta_new;
413  avg_ymom_arr(i,j,k) += facinv*(cur_ymom(i,j,k) - stg_ymom(i,j,k));
414  });
415 
416  Box bx_shrunk_in_k = bx;
417  int klo = tbz.smallEnd(2);
418  int khi = tbz.bigEnd(2);
419  bx_shrunk_in_k.setSmall(2,klo+1);
420  bx_shrunk_in_k.setBig(2,khi-1);
421 
422  // Note that the notes use "g" to mean the magnitude of gravity, so it is positive
423  // We set grav_gpu[2] to be the vector component which is negative
424  // We define halfg to match the notes (which is why we take the absolute value)
425  Real halfg = std::abs(myhalf * grav_gpu[2]);
426 
427  {
428  BL_PROFILE("fast_loop_on_shrunk_t");
429  //Note we don't act on the bottom or top boundaries of the domain
430  ParallelFor(bx_shrunk_in_k, [=] AMREX_GPU_DEVICE (int i, int j, int k)
431  {
432  Real dJ_old_kface = myhalf * (detJ_old(i,j,k) + detJ_old(i,j,k-1));
433  Real dJ_new_kface = myhalf * (detJ_new(i,j,k) + detJ_new(i,j,k-1));
434  Real dJ_stg_kface = myhalf * (detJ_stg(i,j,k) + detJ_stg(i,j,k-1));
435 
436  Real coeff_P = coeffP_a(i,j,k);
437  Real coeff_Q = coeffQ_a(i,j,k);
438 
439  Real theta_t_lo = myhalf * ( prim(i,j,k-2,PrimTheta_comp) + prim(i,j,k-1,PrimTheta_comp) );
440  Real theta_t_mid = myhalf * ( prim(i,j,k-1,PrimTheta_comp) + prim(i,j,k ,PrimTheta_comp) );
441  Real theta_t_hi = myhalf * ( prim(i,j,k ,PrimTheta_comp) + prim(i,j,k+1,PrimTheta_comp) );
442 
443  // line 2 last two terms (order dtau)
444  Real R0_tmp = coeff_P * cur_cons(i,j,k ,RhoTheta_comp)
445  + coeff_Q * cur_cons(i,j,k-1,RhoTheta_comp)
446  - coeff_P * stg_cons(i,j,k ,RhoTheta_comp) * (dJ_stg_kface/dJ_old_kface)
447  - coeff_Q * stg_cons(i,j,k-1,RhoTheta_comp) * (dJ_stg_kface/dJ_old_kface)
448  - halfg * ( cur_cons(i,j,k,Rho_comp) + cur_cons(i,j,k-1,Rho_comp) )
449  + halfg * ( stg_cons(i,j,k,Rho_comp) + stg_cons(i,j,k-1,Rho_comp) ) * (dJ_stg_kface/dJ_old_kface);
450 
451  // line 3 residuals (order dtau^2) one <-> beta_2
452  Real R1_tmp = - halfg * ( slow_rhs_cons(i,j,k ,Rho_comp) + slow_rhs_cons(i,j,k-1,Rho_comp))
453  + coeff_P * slow_rhs_cons(i,j,k ,RhoTheta_comp) + coeff_Q * slow_rhs_cons(i,j,k-1,RhoTheta_comp);
454 
455  Real Omega_kp1 = omega_arr(i,j,k+1);
456  Real Omega_k = omega_arr(i,j,k );
457  Real Omega_km1 = omega_arr(i,j,k-1);
458 
459  Real detJdiff = (detJ_old(i,j,k) - detJ_old(i,j,k-1)) / (detJ_old(i,j,k)*detJ_old(i,j,k-1));
460 
461  // consolidate lines 4&5 (order dtau^2)
462  R1_tmp += halfg * ( beta_1 * dzi * (Omega_kp1/detJ_old(i,j,k) + detJdiff*Omega_k - Omega_km1/detJ_old(i,j,k-1))
463  + temp_rhs_arr(i,j,k,Rho_comp)/detJ_old(i,j,k) + temp_rhs_arr(i,j,k-1,Rho_comp)/detJ_old(i,j,k-1) );
464 
465  // consolidate lines 6&7 (order dtau^2)
466  R1_tmp += -(
467  coeff_P/detJ_old(i,j,k ) * ( beta_1 * dzi * (Omega_kp1*theta_t_hi - Omega_k*theta_t_mid)
468  +temp_rhs_arr(i,j,k ,RhoTheta_comp) ) +
469  coeff_Q/detJ_old(i,j,k-1) * ( beta_1 * dzi * (Omega_k*theta_t_mid - Omega_km1*theta_t_lo)
470  +temp_rhs_arr(i,j,k-1,RhoTheta_comp) ) );
471 
472  // line 1
473  RHS_a(i,j,k) = prev_zmom(i,j,k) - (dJ_stg_kface/dJ_old_kface) * stg_zmom(i,j,k)
474  + dtau * slow_rhs_rho_w(i,j,k) / dJ_stg_kface
475  + dtau * zmom_src_arr(i,j,k);
476 
477  RHS_a(i,j,k) += dtau * R0_tmp;
478 
479  RHS_a(i,j,k) += dtau * dtau*beta_2*R1_tmp;
480 
481  // We cannot use omega_arr here since that was built with old_rho_u and old_rho_v ...
482  Real UppVpp = (dJ_new_kface/dJ_old_kface) * OmegaFromW(i,j,k,0.,cur_xmom,cur_ymom,mf_ux,mf_vy,z_nd_new,dxInv)
483  -(dJ_stg_kface/dJ_old_kface) * OmegaFromW(i,j,k,0.,stg_xmom,stg_ymom,mf_ux,mf_vy,z_nd_stg,dxInv);
484  RHS_a(i,j,k) += UppVpp;
485  });
486  } // end profile
487 
488  Box b2d = tbz; // Copy constructor
489  b2d.setRange(2,0);
490 
491  auto const lo = lbound(bx);
492  auto const hi = ubound(bx);
493 
494  {
495  BL_PROFILE("substep_b2d_loop_t");
496 
497 #ifdef AMREX_USE_GPU
498  ParallelFor(b2d, [=] AMREX_GPU_DEVICE (int i, int j, int)
499  {
500  // Moving terrain
501  Real rho_on_bdy = myhalf * ( prev_cons(i,j,lo.z) + prev_cons(i,j,lo.z-1) );
502  RHS_a(i,j,lo.z) = rho_on_bdy * zp_t_arr(i,j,lo.z);
503 
504  soln_a(i,j,lo.z) = RHS_a(i,j,lo.z) * inv_coeffB_a(i,j,lo.z);
505 
506  RHS_a(i,j,hi.z+1) = dtau * (slow_rhs_rho_w(i,j,hi.z+1) + zmom_src_arr(i,j,hi.z+1));
507 
508  for (int k = lo.z+1; k <= hi.z+1; k++) {
509  soln_a(i,j,k) = (RHS_a(i,j,k)-coeffA_a(i,j,k)*soln_a(i,j,k-1)) * inv_coeffB_a(i,j,k);
510  }
511 
512  for (int k = hi.z; k >= lo.z; k--) {
513  soln_a(i,j,k) -= ( coeffC_a(i,j,k) * inv_coeffB_a(i,j,k) ) * soln_a(i,j,k+1);
514  }
515 
516  // We assume that Omega == w at the top boundary and that changes in J there are irrelevant
517  cur_zmom(i,j,hi.z+1) = stg_zmom(i,j,hi.z+1) + soln_a(i,j,hi.z+1);
518  });
519 #else
520  for (int j = lo.y; j <= hi.y; ++j) {
521  AMREX_PRAGMA_SIMD
522  for (int i = lo.x; i <= hi.x; ++i) {
523 
524  Real rho_on_bdy = myhalf * ( prev_cons(i,j,lo.z) + prev_cons(i,j,lo.z-1) );
525  RHS_a(i,j,lo.z) = rho_on_bdy * zp_t_arr(i,j,lo.z);
526 
527  soln_a(i,j,lo.z) = RHS_a(i,j,lo.z) * inv_coeffB_a(i,j,lo.z);
528  }
529  }
530 
531  for (int j = lo.y; j <= hi.y; ++j) {
532  AMREX_PRAGMA_SIMD
533  for (int i = lo.x; i <= hi.x; ++i) {
534  RHS_a(i,j,hi.z+1) = dtau * (slow_rhs_rho_w(i,j,hi.z+1) + zmom_src_arr(i,j,hi.z+1));
535  }
536  }
537  for (int k = lo.z+1; k <= hi.z+1; ++k) {
538  for (int j = lo.y; j <= hi.y; ++j) {
539  AMREX_PRAGMA_SIMD
540  for (int i = lo.x; i <= hi.x; ++i) {
541  soln_a(i,j,k) = (RHS_a(i,j,k)-coeffA_a(i,j,k)*soln_a(i,j,k-1)) * inv_coeffB_a(i,j,k);
542  }
543  }
544  }
545  for (int k = hi.z; k >= lo.z; --k) {
546  for (int j = lo.y; j <= hi.y; ++j) {
547  AMREX_PRAGMA_SIMD
548  for (int i = lo.x; i <= hi.x; ++i) {
549  soln_a(i,j,k) -= ( coeffC_a(i,j,k) * inv_coeffB_a(i,j,k) ) * soln_a(i,j,k+1);
550  }
551  }
552  }
553 
554  // We assume that Omega == w at the top boundary and that changes in J there are irrelevant
555  for (int j = lo.y; j <= hi.y; ++j) {
556  AMREX_PRAGMA_SIMD
557  for (int i = lo.x; i <= hi.x; ++i) {
558  cur_zmom(i,j,hi.z+1) = stg_zmom(i,j,hi.z+1) + soln_a(i,j,hi.z+1);
559  }
560  }
561 #endif
562  } // end profile
563 
564  {
565  BL_PROFILE("substep_new_drhow");
566  tbz.setBig(2,hi.z);
567  ParallelFor(tbz, [=] AMREX_GPU_DEVICE (int i, int j, int k)
568  {
569  Real rho_on_face = myhalf * (cur_cons(i,j,k,Rho_comp) + cur_cons(i,j,k-1,Rho_comp));
570 
571  if (k == lo.z) {
572  cur_zmom(i,j,k) = WFromOmega(i,j,k,rho_on_face*(z_t_arr(i,j,k)+zp_t_arr(i,j,k)),
573  cur_xmom,cur_ymom,mf_ux,mf_vy,z_nd_new,dxInv);
574 
575  // We need to set this here because it is used to define zflux_lo below
576  soln_a(i,j,k) = zero;
577 
578  } else {
579 
580  Real UppVpp = WFromOmega(i,j,k,zero,cur_xmom,cur_ymom,mf_ux,mf_vy,z_nd_new,dxInv)
581  - WFromOmega(i,j,k,zero,stg_xmom,stg_ymom,mf_ux,mf_vy,z_nd_stg,dxInv);
582  Real wpp = soln_a(i,j,k) + UppVpp;
583  Real dJ_old_kface = myhalf * (detJ_old(i,j,k) + detJ_old(i,j,k-1));
584  Real dJ_new_kface = myhalf * (detJ_new(i,j,k) + detJ_new(i,j,k-1));
585 
586  cur_zmom(i,j,k) = dJ_old_kface * (stg_zmom(i,j,k) + wpp);
587  cur_zmom(i,j,k) /= dJ_new_kface;
588 
589  soln_a(i,j,k) = OmegaFromW(i,j,k,cur_zmom(i,j,k),cur_xmom,cur_ymom,mf_ux,mf_vy,z_nd_new,dxInv)
590  - OmegaFromW(i,j,k,stg_zmom(i,j,k),stg_xmom,stg_ymom,mf_ux,mf_vy,z_nd_stg,dxInv);
591  soln_a(i,j,k) -= rho_on_face * zp_t_arr(i,j,k);
592  }
593 
594  if (l_rayleigh_impl_for_w && k > 0) {
595  Real damping_coeff = l_damp_coef * dtau * sinesq_stag_d[k];
596  cur_zmom(i,j,k) /= (one + damping_coeff);
597  }
598  });
599  } // end profile
600 
601  // **************************************************************************
602  // Define updates in the RHS of rho and (rho theta)
603  // **************************************************************************
604  {
605  BL_PROFILE("fast_rho_final_update");
606  ParallelFor(bx, [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept
607  {
608  Real zflux_lo = beta_2 * soln_a(i,j,k ) + beta_1 * omega_arr(i,j,k);
609  Real zflux_hi = beta_2 * soln_a(i,j,k+1) + beta_1 * omega_arr(i,j,k+1);
610 
611  // Note that in the solve we effectively impose new_drho_w(i,j,vbx_hi.z+1)=0
612  // so we don't update avg_zmom at k=vbx_hi.z+1
613  avg_zmom_arr(i,j,k) += facinv*zflux_lo / (mf_mx(i,j,0) * mf_my(i,j,0));
614  if (l_reflux) {
615  (flx_arr[2])(i,j,k,0) = zflux_lo / (mf_mx(i,j,0) * mf_my(i,j,0));
616  }
617 
618  // Note that the factor of (1/J) in the fast source term is canceled
619  // when we multiply old and new by detJ_old and detJ_new , respectively
620  // We have already scaled the slow source term to have the extra factor of dJ
621  Real fast_rhs_rho = -(temp_rhs_arr(i,j,k,0) + ( zflux_hi - zflux_lo ) * dzi);
622  Real fast_rhs_rhotheta = -( temp_rhs_arr(i,j,k,1) + myhalf *
623  ( zflux_hi * (prim(i,j,k) + prim(i,j,k+1))
624  - zflux_lo * (prim(i,j,k) + prim(i,j,k-1)) ) * dzi );
625 
626  cur_cons(i,j,k,0) *= (detJ_old(i,j,k)/detJ_new(i,j,k));
627  cur_cons(i,j,k,1) *= (detJ_old(i,j,k)/detJ_new(i,j,k));
628 
629  cur_cons(i,j,k,0) += dtau * ( slow_rhs_cons(i,j,k,0) + fast_rhs_rho / detJ_new(i,j,k));
630  cur_cons(i,j,k,1) += dtau * ( slow_rhs_cons(i,j,k,1) + fast_rhs_rhotheta / detJ_new(i,j,k));
631 
632  if (l_reflux) {
633  (flx_arr[2])(i,j,k,1) = (flx_arr[2])(i,j,k,0) * myhalf * (prim(i,j,k) + prim(i,j,k-1));
634  }
635 
636  if (k == vbx_hi.z) {
637  avg_zmom_arr(i,j,k+1) += facinv * zflux_hi / (mf_mx(i,j,0) * mf_my(i,j,0));
638  if (l_reflux) {
639  (flx_arr[2])(i,j,k+1,0) = zflux_hi / (mf_mx(i,j,0) * mf_my(i,j,0));
640  (flx_arr[2])(i,j,k+1,1) = (flx_arr[2])(i,j,k+1,0) * myhalf * (prim(i,j,k) + prim(i,j,k+1));
641  }
642  }
643 
644  // add in source terms for cell-centered conserved variables
645  cur_cons(i,j,k,Rho_comp) += dtau * cc_src_arr(i,j,k,Rho_comp);
646  cur_cons(i,j,k,RhoTheta_comp) += dtau * cc_src_arr(i,j,k,RhoTheta_comp);
647  });
648  } // end profile
649 
650  // We only add to the flux registers in the final RK step
651  if (l_reflux) {
652  int strt_comp_reflux = 0;
653  // For now we don't reflux (rho theta) because it seems to create issues at c/f boundaries
654  int num_comp_reflux = 1;
655  if (level < finest_level) {
656  fr_as_crse->CrseAdd(mfi,
657  {{AMREX_D_DECL(&(flux[0]), &(flux[1]), &(flux[2]))}},
658  dx, dtau, strt_comp_reflux, strt_comp_reflux, num_comp_reflux, RunOn::Device);
659  }
660  if (level > 0) {
661  fr_as_fine->FineAdd(mfi,
662  {{AMREX_D_DECL(&(flux[0]), &(flux[1]), &(flux[2]))}},
663  dx, dtau, strt_comp_reflux, strt_comp_reflux, num_comp_reflux, RunOn::Device);
664  }
665 
666  // This is necessary here so we don't go on to the next FArrayBox without
667  // having finished copying the fluxes into the FluxRegisters (since the fluxes
668  // are stored in temporary FArrayBox's)
669  Gpu::streamSynchronize();
670 
671  } // two-way coupling
672 
673  } // mfi
674  } // OMP
675 }
R_v
constexpr amrex::Real R_v
Definition: ERF_Constants.H:48
one
constexpr amrex::Real one
Definition: ERF_Constants.H:9
fourth
constexpr amrex::Real fourth
Definition: ERF_Constants.H:14
zero
constexpr amrex::Real zero
Definition: ERF_Constants.H:8
myhalf
constexpr amrex::Real myhalf
Definition: ERF_Constants.H:13
R_d
constexpr amrex::Real R_d
Definition: ERF_Constants.H:47
Gamma
constexpr amrex::Real Gamma
Definition: ERF_Constants.H:62
v_y
@ v_y
Definition: ERF_DataStruct.H:25
m_y
@ m_y
Definition: ERF_DataStruct.H:25
u_x
@ u_x
Definition: ERF_DataStruct.H:24
m_x
@ m_x
Definition: ERF_DataStruct.H:24
PrimQ1_comp
#define PrimQ1_comp
Definition: ERF_IndexDefines.H:58
Rho_comp
#define Rho_comp
Definition: ERF_IndexDefines.H:36
RhoTheta_comp
#define RhoTheta_comp
Definition: ERF_IndexDefines.H:37
PrimTheta_comp
#define PrimTheta_comp
Definition: ERF_IndexDefines.H:55
dxInv
amrex::GpuArray< Real, AMREX_SPACEDIM > dxInv
Definition: ERF_InitCustomPertVels_ParticleTests.H:17
dx
const Real dx
Definition: ERF_InitCustomPert_ABL.H:23
khi
const int khi
Definition: ERF_InitCustomPert_Bubble.H:21
ParallelFor
ParallelFor(grown_box, [=] AMREX_GPU_DEVICE(int i, int j, int k) { qrcuten_arr(i, j, k)=Real(0);qscuten_arr(i, j, k)=Real(0);qicuten_arr(i, j, k)=Real(0);})
Real
amrex::Real Real
Definition: ERF_ShocInterface.H:19
OmegaFromW
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real OmegaFromW(int &i, int &j, int &k, amrex::Real w, const amrex::Array4< const amrex::Real > &u_arr, const amrex::Array4< const amrex::Real > &v_arr, const amrex::Array4< const amrex::Real > &mf_u, const amrex::Array4< const amrex::Real > &mf_v, const amrex::Array4< const amrex::Real > &z_nd, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &dxInv)
Definition: ERF_TerrainMetrics.H:414
Compute_h_xi_AtIface
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real Compute_h_xi_AtIface(const int &i, const int &j, const int &k, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &cellSizeInv, const amrex::Array4< const amrex::Real > &z_nd)
Definition: ERF_TerrainMetrics.H:117
Compute_h_zeta_AtIface
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real Compute_h_zeta_AtIface(const int &i, const int &j, const int &k, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &cellSizeInv, const amrex::Array4< const amrex::Real > &z_nd)
Definition: ERF_TerrainMetrics.H:104
Compute_h_zeta_AtJface
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real Compute_h_zeta_AtJface(const int &i, const int &j, const int &k, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &cellSizeInv, const amrex::Array4< const amrex::Real > &z_nd)
Definition: ERF_TerrainMetrics.H:144
Compute_h_eta_AtJface
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real Compute_h_eta_AtJface(const int &i, const int &j, const int &k, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &cellSizeInv, const amrex::Array4< const amrex::Real > &z_nd)
Definition: ERF_TerrainMetrics.H:170
WFromOmega
AMREX_GPU_DEVICE AMREX_FORCE_INLINE amrex::Real WFromOmega(int &i, int &j, int &k, amrex::Real omega, const amrex::Array4< const amrex::Real > &u_arr, const amrex::Array4< const amrex::Real > &v_arr, const amrex::Array4< const amrex::Real > &mf_u, const amrex::Array4< const amrex::Real > &mf_v, const amrex::Array4< const amrex::Real > &z_nd, const amrex::GpuArray< amrex::Real, AMREX_SPACEDIM > &dxInv)
Definition: ERF_TerrainMetrics.H:464
GpVars::gpy
@ gpy
Definition: ERF_IndexDefines.H:186
GpVars::gpx
@ gpx
Definition: ERF_IndexDefines.H:185
IntVars::ymom
@ ymom
Definition: ERF_IndexDefines.H:195
IntVars::cons
@ cons
Definition: ERF_IndexDefines.H:193
IntVars::zmom
@ zmom
Definition: ERF_IndexDefines.H:196
IntVars::xmom
@ xmom
Definition: ERF_IndexDefines.H:194
MicVar_Kess::qt
@ qt
Definition: ERF_Kessler.H:29
MicVar_Kess::qv
@ qv
Definition: ERF_Kessler.H:30
WSM6Ind::q
@ q
Definition: ERF_WSM6.H:182
Here is the call graph for this function: