Update CATKE parameters for new implicit time-stepping (#2461)

* Bump version * Updates CATKE defaults
CliMA · May 9, 2022 · d4a462b · d4a462b
1 parent 920abd1
commit d4a462b
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 22 deletions.
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "Oceananigans"
 uuid = "9e8cae18-63c1-5223-a75c-80ca9d6e9a09"
 authors = ["Climate Modeling Alliance and contributors"]
-version = "0.76.1"
+version = "0.76.2"
 
 [deps]
 Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"

diff --git a/...rbulence_closure_implementations/CATKEVerticalDiffusivities/CATKEVerticalDiffusivities.jl b/...rbulence_closure_implementations/CATKEVerticalDiffusivities/CATKEVerticalDiffusivities.jl
@@ -105,7 +105,7 @@ Turbulent Kinetic Energy (TKE).
 CATKEVerticalDiffusivity(FT::DataType; kw...) = CATKEVerticalDiffusivity(VerticallyImplicitTimeDiscretization(), FT; kw...)
 
 function CATKEVerticalDiffusivity(time_discretization = VerticallyImplicitTimeDiscretization(), FT=Float64;
-                                  Cᴰ = 2.91,
+                                  Cᴰ = 0.81,
                                   mixing_length = MixingLength{FT}(),
                                   surface_TKE_flux = SurfaceTKEFlux{FT}(),
                                   warning = true)

diff --git a/...ceClosures/turbulence_closure_implementations/CATKEVerticalDiffusivities/mixing_length.jl b/...ceClosures/turbulence_closure_implementations/CATKEVerticalDiffusivities/mixing_length.jl
@@ -65,29 +65,29 @@ annealing and noisy Ensemble Kalman Inversion methods.
 Base.@kwdef struct MixingLength{FT}
     Cᵇ    :: FT = Inf # Inf is the "inert" value for this parameter
     Cˢ    :: FT = Inf # Inf is the "inert" value for this parameter
-    Cᵇu   :: FT = 0.596
-    Cᵇc   :: FT = 0.0723
-    Cᵇe   :: FT = 0.637
-    Cˢu   :: FT = 0.628
-    Cˢc   :: FT = 0.426
-    Cˢe   :: FT = 0.711
+    Cᵇu   :: FT = 1.55
+    Cᵇc   :: FT = 0.01
+    Cᵇe   :: FT = 0.60
+    Cˢu   :: FT = 5.1
+    Cˢc   :: FT = 4.3
+    Cˢe   :: FT = 1.49
     Cᵟu   :: FT = 0.5
     Cᵟc   :: FT = 0.5
     Cᵟe   :: FT = 0.5
     Cᴬu   :: FT = 0.0
-    Cᴬc   :: FT = 1.41
+    Cᴬc   :: FT = 0.0
     Cᴬe   :: FT = 0.0
     Cᴬˢu  :: FT = 0.0
-    Cᴬˢc  :: FT = 0.812
+    Cᴬˢc  :: FT = 0.0
     Cᴬˢe  :: FT = 0.0
-    Cᴷu⁻  :: FT = 0.343
-    Cᴷuʳ  :: FT = -0.721
-    Cᴷc⁻  :: FT = 0.343
-    Cᴷcʳ  :: FT = -0.891
-    Cᴷe⁻  :: FT = 1.42
-    Cᴷeʳ  :: FT = 1.50
-    CᴷRiʷ :: FT = 0.101
-    CᴷRiᶜ :: FT = 2.03
+    Cᴷu⁻  :: FT = 0.14
+    Cᴷuʳ  :: FT = 0.1
+    Cᴷc⁻  :: FT = 0.35
+    Cᴷcʳ  :: FT = 0.05
+    Cᴷe⁻  :: FT = 0.49
+    Cᴷeʳ  :: FT = 17.0
+    CᴷRiʷ :: FT = 30.0
+    CᴷRiᶜ :: FT = 1.1
 end
 
 #####
@@ -114,7 +114,7 @@ end
     FT = eltype(grid)
     N⁺ = sqrt_∂z_b(i, j, k, grid, buoyancy, tracers)
     e⁺ = ℑzᵃᵃᶠ(i, j, k, grid, ψ⁺, e)
-    return @inbounds ifelse(N⁺ == 0, FT(Inf), sqrt(e⁺) / N⁺)
+    return ifelse(N⁺ == 0, FT(Inf), sqrt(e⁺) / N⁺)
 end
 
 @inline function shear_mixing_lengthᶜᶜᶠ(i, j, k, grid, e, velocities, tracers, buoyancy)
@@ -223,7 +223,9 @@ end
 
     σu = momentum_stable_mixing_scale(i, j, k, grid, closure, velocities, tracers, buoyancy)
 
-    return max(ℓʰ, σu * max(Cᵟu * ℓᵟ, ℓ★))
+    return σu * max(Cᵟu * ℓᵟ, ℓ★)
+
+    #return max(ℓʰ, σu * max(Cᵟu * ℓᵟ, ℓ★))
 end
 
 @inline function tracer_mixing_lengthᶜᶜᶠ(i, j, k, grid, closure, velocities, tracers, buoyancy, clock, tracer_bcs)

diff --git a/...losures/turbulence_closure_implementations/CATKEVerticalDiffusivities/surface_TKE_flux.jl b/...losures/turbulence_closure_implementations/CATKEVerticalDiffusivities/surface_TKE_flux.jl
@@ -23,8 +23,8 @@ The calibration was performed using a combination of Markov Chain Monte Carlo (M
 annealing and noisy Ensemble Kalman Inversion methods.
 """
 Base.@kwdef struct SurfaceTKEFlux{FT}
-    Cᵂu★ :: FT = 4.56
-    CᵂwΔ :: FT = 4.46
+    Cᵂu★ :: FT = 0.01
+    CᵂwΔ :: FT = 40.0
 end
 
 #####