Skip to content

Commit

Permalink
Fix for rare instability in (probabilistic) seabed stress (#810)
Browse files Browse the repository at this point in the history
* Modified doc for Dupont et al ref

* Minor modif to v_i calculation in seabed_stress_factor_prob to prevent rare instability
  • Loading branch information
JFLemieux73 committed Jan 19, 2023
1 parent 0bf0fdc commit 28fdbeb
Show file tree
Hide file tree
Showing 5 changed files with 19 additions and 8 deletions.
11 changes: 6 additions & 5 deletions cicecore/cicedyn/dynamics/ice_dyn_shared.F90
Original file line number Diff line number Diff line change
Expand Up @@ -59,7 +59,7 @@ module ice_dyn_shared
yield_curve , & ! 'ellipse' ('teardrop' needs further testing)
visc_method , & ! method for viscosity calc at U points (C, CD grids)
seabed_stress_method ! method for seabed stress calculation
! LKD: Lemieux et al. 2015, probabilistic: Dupont et al. in prep.
! LKD: Lemieux et al. 2015, probabilistic: Dupont et al. 2022

real (kind=dbl_kind), parameter, public :: &
u0 = 5e-5_dbl_kind, & ! residual velocity for seabed stress (m/s)
Expand Down Expand Up @@ -1347,8 +1347,9 @@ end subroutine seabed_stress_factor_LKD
! a normal distribution with sigma_b = 2.5d0. An improvement would
! be to provide the distribution based on high resolution data.
!
! Dupont, F. Dumont, D., Lemieux, J.F., Dumas-Lefebvre, E., Caya, A.
! in prep.
! Dupont, F., D. Dumont, J.F. Lemieux, E. Dumas-Lefebvre, A. Caya (2022).
! A probabilistic seabed-ice keel interaction model, The Cryosphere, 16,
! 1963-1977.
!
! authors: D. Dumont, J.F. Lemieux, E. Dumas-Lefebvre, F. Dupont
!
Expand Down Expand Up @@ -1481,13 +1482,13 @@ subroutine seabed_stress_factor_prob (nx_block, ny_block, &
do n =1, ncat
v_i = v_i + vcat(n)**2 / (max(acat(n), puny))
enddo
v_i = v_i - m_i**2
v_i = max((v_i - m_i**2), puny)

mu_i = log(m_i/sqrt(c1 + v_i/m_i**2)) ! parameters for the log-normal
sigma_i = sqrt(log(c1 + v_i/m_i**2))

! max thickness associated with percentile of log-normal PDF
! x_kmax=x997 was obtained from an optimization procedure (Dupont et al.)
! x_kmax=x997 was obtained from an optimization procedure (Dupont et al. 2022)

x_kmax = exp(mu_i + sqrt(c2*sigma_i)*1.9430d0)

Expand Down
2 changes: 1 addition & 1 deletion cicecore/cicedyn/general/ice_init.F90
Original file line number Diff line number Diff line change
Expand Up @@ -392,7 +392,7 @@ subroutine input_data
dyscale = 1.0_dbl_kind ! user defined rectgrid y-grid scale factor (e.g., 1.02)
close_boundaries = .false. ! true = set land on edges of grid
seabed_stress= .false. ! if true, seabed stress for landfast is on
seabed_stress_method = 'LKD'! LKD = Lemieux et al 2015, probabilistic = Dupont et al. in prep
seabed_stress_method = 'LKD'! LKD = Lemieux et al 2015, probabilistic = Dupont et al. 2022
k1 = 7.5_dbl_kind ! 1st free parameter for landfast parameterization
k2 = 15.0_dbl_kind ! 2nd free parameter (N/m^3) for landfast parametrization
alphab = 20.0_dbl_kind ! alphab=Cb factor in Lemieux et al 2015
Expand Down
10 changes: 10 additions & 0 deletions doc/source/master_list.bib
Original file line number Diff line number Diff line change
Expand Up @@ -1082,6 +1082,16 @@ @article{Bouchat22
year = {2022}
}

@Article{Dupont22,
author = {F. Dupont and D. Dumont and J.F. Lemieux and E. Dumas-Lefebvre and A. Caya},
title = "{A probabilistic seabed-ice keel interaction model}",
journal = TC,
year = {2022},
volume = {16},
pages = {1963-1977},
url = {https://doi.org/10.5194/tc-16-1963-2022}
}

@Article{Tsujino18,
author = "H. Tsujino and S. Urakawa and R.J. Small and W.M. Kim and S.G. Yeager and et al.",
title = "{JRA‐55 based surface dataset for driving ocean–sea‐ice models (JRA55‐do)}",
Expand Down
2 changes: 1 addition & 1 deletion doc/source/science_guide/sg_dynamics.rst
Original file line number Diff line number Diff line change
Expand Up @@ -382,7 +382,7 @@ The value of :math:`k_1` can be changed at runtime using the namelist variable `

This more sophisticated grounding parameterization computes the seabed stress based
on the probability of contact between the ice thickness distribution
(ITD) and the seabed. Multi-thickness category models such as CICE typically use a
(ITD) and the seabed :cite:`Dupont22`. Multi-thickness category models such as CICE typically use a
few thickness categories (5-10). This crude representation of the ITD
does not resolve the tail of the ITD, which is crucial for grounding
events.
Expand Down
2 changes: 1 addition & 1 deletion doc/source/user_guide/ug_case_settings.rst
Original file line number Diff line number Diff line change
Expand Up @@ -485,7 +485,7 @@ dynamics_nml
"``revised_evp``", "logical", "use revised EVP formulation", "``.false.``"
"``seabed_stress``", "logical", "use seabed stress parameterization for landfast ice", "``.false.``"
"``seabed_stress_method``", "``LKD``", "linear keel draft method :cite:`Lemieux16`", "``LKD``"
"", "``probabilistic``", "probability of contact method (Dupont et al., in prep)", ""
"", "``probabilistic``", "probability of contact method :cite:`Dupont22`", ""
"``ssh_stress``", "``coupled``", "computed from coupled sea surface height gradient", "``geostrophic``"
"", "``geostropic``", "computed from ocean velocity", ""
"``threshold_hw``", "real", "Max water depth for grounding (see :cite:`Amundrud04`)", "30."
Expand Down

0 comments on commit 28fdbeb

Please sign in to comment.