Introduction
This page explains the workflow for deriving default settings for a new version of XBeach. Default settings are important for users with limited experience with XBeach. For these users the number of exposed settings should be limited, preferably to case related input, like waterlevels, bathymetry and wave conditions, only. The occasion of setting up this workflow is the preparation of XBeach as assessment model for the Dutch coast. In line with the current WTI project leading this preparation, this workflow is limited to application for the Dutch coast and 1D. Different default settings for different applications may be derived by using this workflow.
Overview
The workflow consists of the following steps:
- Create long list of XBeach settings
- Conduct marginal sensitivity analysis on erosion volume above SSL on long list
- Submit long list including sensitivity results to expert panel
- Create short list of XBeach settings based on feedback expert panel
- Compute variation matrix of relevant ranges for settings on short list on relevant cases
- Pick the combination of settings with smallest RMS error with respect to erosion volume above SSL
Long list
The long list is constructed based on parsing the params.F90 file for settings and removing any parameters that are either case-dependent input (e.g. Hm0) or irrelevant (ships, groundwater flow, etc.). The following parameters are excluded from the long list:
D50 # D50 grain size per grain type
D90 # D90 grain size per grain type
Hrms # Hrms wave height for instat = 0,1,2,3
Tlong # Wave group period for case instat = 1
Tm01switch # Turn off or on Tm01 or Tm-10 switch
Topt # Absolute period to optimize coefficient
Trep # Representative wave period for instat = 0,1,2,3
alfa # Angle of x-axis from East
aquiferbot # Level of uniform aquifer bottom
aquiferbotfile # Name of the aquifer bottom file
avalanching # Include avalanching (1) or exclude (0)
back # Switch for boundary at bay side, 0 = radiating boundary (Ad), 1 = reflective boundary; uu=0
bcfile # Name of spectrum file
bchwiz # Use beachwizard, 0 = beachwizard off, 1 beachwizard on, bed update off, 2 beachwiz on, bed update on. (also requires morphology == 1)
bedfricfile # Bed friction file (only valid with values of C)
bedfriction # Bed friction formulation: 'chezy','white-colebrook'
compi # Imaginary unit
depfile # Name of the input bathymetry file
dir0 # Mean wave direction (Nautical convention) for instat = 0,1,2,3
disch_loc_file # Name of discharge locations file
disch_timeseries_file # Name of discharge timeseries file
drifterfile # Name of drifter data file
dt # Computational time step, in hydrodynamic time
dthetaS_XB # The (counter-clockwise) angle in the degrees needed to rotate from the x-axis in SWAN to the x-axis pointing East
dx # Regular grid spacing in x-direction
dy # Regular grid spacing in y-direction
dzg1 # Thickness of top sediment class layers
dzg2 # Nominal thickness of variable sediment class layer
dzg3 # Thickness of bottom sediment class layers
flow # Include flow calculation (1), or exclude (0)
freewave # Switch for free wave propagation 0 = use cg (default); 1 = use sqrt(gh) in instat = 3
front # Switch for seaward flow boundary: 0 = radiating boundary(Ad), 1 = Van Dongeren, 1997
g # Gravitational acceleration
globalvars # Mnems of global output variables, not per se the same sice as nglobalvar (invalid variables, defaults)
gridform # Swicth to read in grid bathy files with 'XBeach' or 'Delft3D' format respectively
gw0 # Level initial groundwater level
gw0file # Name of initial groundwater level file
gwReturb # Reynolds number for start of turbulent flow in case of gwscheme = turbulent
gwfastsolve # Reduce full 2D non-hydrostatic solution to quasi-explicit in longshore direction
gwflow # Turn on (1) or off (0) groundwater flow module
gwheadmodel # Model to use for vertical groundwater head: 'parabolic' (default), or 'exponential'
gwhorinfil # switch to include horizontal infiltration from surface water to groundwater (default = 0)
gwnonh # Switch to turn on or off non-hydrostatic pressure for groundwater
gwscheme # Scheme for momentum equation (laminar, turbulent)
hotstartflow # Switch to hotstart flow conditions with pressure gradient balanced by wind and bed stress
instat # Wave boundary condtion type
kmax # Number of sigma layers in Quasi-3D model; kmax = 1 (default) is without vertical structure of flow and suspensions
kx # Darcy-flow permeability coefficient in x-direction [m/s]
ky # Darcy-flow permeability coefficient in y-direction [m/s]
kz # Darcy-flow permeability coefficient in z-direction [m/s]
lat # Latitude at model location for computing Coriolis
left # Switch for lateral boundary at ny+1, 'neumann' = vv computed from NSWE, 'wall' = reflective wall; vv=0
leftwave # old name for lateralwave
lwave # Include short wave forcing on NLSW equations and boundary conditions (1), or exclude (0)
maxiter # Maximum number of iterations in wave stationary
meanvars # Mnems of mean output variables (by variables)
mmpi # Number of domains in m direction (cross-shore) when manually specifying mpi domains
morfac # Morphological acceleration factor
morfacopt # Option indicating whether times should be adjusted (1) or not(0) for morfac
morphology # Include morphology (1) or exclude (0)
morstart # Start time morphology, in morphological time
morstop # Stop time morphology, in morphological time
mpiboundary # Fix mpi boundaries along y-lines ('y'), x-lines ('x'), use manual defined domains ('man') or find shortest boundary ('auto')
ncfilename # xbeach netcdf output file name
ncross # Number of output cross sections
nd # Number of computational layers in the bed
nd_var # Index of layer with variable thickness
ndischarge # Number of discharge locations
ndrifter # Number of drifers
ne_layer # Name of file containing depth of hard structure
ngd # Number of sediment classes
nglobalvar # Number of global output variables (as specified by user)
nhbreaker # Turn on or off nonhydrostatic breaker model
nmeanvar # Number of mean,min,max,var output variables
nmpi # Number of domains in n direction (alongshore) when manually specifying mpi domains
nonh # Non-hydrostatic pressure option: 0 = NSWE, 1 = NSW + non-hydrostatic pressure compensation Stelling & Zijlema, 2003
nonhspectrum # Switch between spectrum format for wave action balance of nonhydrostatic waves
npoints # Number of output point locations
npointvar # Number of point output variables
nrugauge # Number of output runup gauge locations
nship # Number of ships
nspectrumloc # Number of input spectrum locations
nspr # nspr = 1 long wave direction forced into centres of short wave bins, nspr = 0 regular long wave spreadin
ntdischarge # Length of discharge time series
nveg # Number of vegetation species
nx # Number of computiation cell corners in x-direction
ny # Number of computiation cell corners in y-direction
oldhu # Turn on / off old hu calculation
oldnyq # Turn off or on old nyquist switch
outputformat # Choice of output file format: 'netcdf', 'fortran', or 'debug'
paulrevere # Specifies tide on sea and land ('land') or two sea points ('sea') if tideloc = 2
pointtypes # Point types (0 = point, 1=rugauge)
pointvars # Mnems of point output variables (by variables)
posdwn # Bathymetry is specified positive down (1) or positive up (-1)
projection # projection string
px # Pi
q3d # Turn on (1) or off (0) quasi-3D sediment transport module
random # Random seed on (1) or off (0) for instat = 4,5,6 boundary conditions
rho # Density of water
rhoa # Air density
rhog8 # 1/8*rho*g
rhos # Solid sediment density (no pores)
right # Switch for lateral boundary at right, 0 = vv computed from NSWE, 1 = reflective wall; vv=0
rightwave # old name for lateralwave
rotate # Rotate (1) postprocess output with the rotate function.
rt # Duration of wave spectrum at offshore boundary, in morphological time
sedtrans # Include sediment transport (1) or exclude (0)
shipfile # Name of ship data file
ships # Turn on (1) or off (0) ship waves
solver # Solver used to solve the linear system, sip, or tridiag (only for 1d)
solver_maxit # Maximum number of iterations in the linear sip solver
struct # Switch for hard structures
swave # Include short waves (1), exclude short waves (0)
swrunup # Turn on (1) or off (0) short wave runup
t # Computational time, in hydrodynamic time
taper # Spin-up time of wave boundary conditions, in morphological time
thetamax # Higher directional limit (angle w.r.t computational x-axis)
thetamin # Lower directional limit (angle w.r.t computational x-axis)
thetanaut # Thetamin,thetamax in cartesian (0) or nautical convention (1)
thetanum # Coefficient determining whether upwind (1) or central scheme (0.5) is used.
tideloc # Number of corner points on which a tide time series is specified
tidetype # Switch for offfshore boundary, velocity boundary or instant water level boundary (default)
timings # Switch to turn on (1) or off (0) progress output to screen
tintc # Interval time of cross section output
tintg # Interval time of global output
tintm # Interval time of mean,var,max,min output
tintp # Interval time of point and runup gauge output
tnext # Next time point for output or wave stationary calculation, in hydrodynamic time
tscross # Name of file containing timings of cross section output
tsglobal # Name of file containing timings of global output
tsmean # Name of file containing timings of mean, max, min and var output
tspoints # Name of file containing timings of point output
tstart # Start time of output, in morphological time
tstop # Stop time of simulation, in morphological time
tunits # Units can be defined in udunits format (seconds since 1970-01-01 00:00:00.00 +1:00)
vardx # Switch for variable grid spacing: 1 = irregular spacing, 0 = regular grid spacing
vegetation # Turn on (1) or off (0) interaction of waves and flow with vegetation
vegiefile # Name of vegie species list file
vegiemapfile # Name of vegie species map file
wbcversion # Version of wave boundary conditions
wearth # Angular velocity of earth calculated as: 1/rotation_time (in hours), later changed in calculation code to rad/s
windfile # Name of file with non-stationary wind data
windth # Nautical wind direction, in case of stationary wind
windv # Wind velocity, in case of stationary wind
xfile # Name of the file containing x-coordinates of the calculation grid
xori # X-coordinate of origin of axis
xyfile # Name of the file containing (Delft3D) xy-coordinates of the calculation grid
yfile # Name of the file containing y-coordinates of the calculation grid
yori # Y-coordinate of origin of axis
zs0 # Inital water level
zs0file # Name of tide boundary condition series
zsinitfile # Name of inital condition file zs
Marginal sensitivity analysis
The marginal sensitivity analysis is performed using a default dune profile and representative settings for the Dutch coast (normative conditions). Using this model, a single parameter is varied over the applicable range of values. The default values are always included. The sensitivity is subsequently determined by looking at the variation in erosion volume depending on that parameter setting.
Preliminary results are as follows. The first plot shows for each varied parameter the variation in erosion volume above SSL. The black dot indicates the default value. The red line indicates a linear fit excluding any outliers. The second plot shows the minimum and maximum erosion volumes found when varying a parameter over it's valid range of values. The plot is ordered based on the difference between the minimum and maximum value found. Some parameters show no influence at all. This might be due to the fact that a parameter is only applicable to a part of XBeach that is not enabled by default. 
Expert panel
- Dano Roelvink
- Ad Reniers
- Jaap van Thiel de Vries
- Robert McCall
- Ap van Dongeren
Short list
Variation matrix
Resolution vs. dimensions
We can choose either to vary a large number of parameters (dimensions) using a limited resolution or vary a small number of parameters using a high resolution.
Cases
- Deltaflume ?
- Deltaflume 2006
- Boers ?
- 1953 storm surge
Result
Discussion
Several topics within this workflow are still being discussed. These discussions are briefly mentioned below.
XBeach allows a variety of performance indicators to be used to compare the performance of one set of settings against another. For example, a profile comparison with measured data may be used by computing a BSS score. The BSS score itself, however, is subject to discussion for this purpose. Moreover, comparison data is often obtained from flume experiments on scale. These experiments use scaling rules that are derived by comparing erosion profiles rather than profile shape. The profile shape for these tests can thus not be used. For now, we use the erosion volume above SSL as performance indicator. Another advantage of this indicator is that it relates well to other calibration studies for dune assessment models in the Netherlands.
Scaling
Comparison with data is often done using data obtained from flume experiments. These experiments are performed on scale. XBeach allows us to simulate the experiment at the scale of the experiment rather than on a 1:1 scale. As for now, this is not done for two reasons. First, the scaling within XBeach is still subject to discussion. Second, other calibration studies for dune assessment models also first scaled the experiment results and then performed the calibration.
