Property:Describe numerical limitations

From CSDMS

This is a property of type Text.

Showing 250 pages using this property.
S
'''Limitations''' The DIA approximation for the quadruplet wave-wave interactions depends on the width of the directional distribution of the wave spectrum. It seems to work reasonably in many cases but it is a poor approximation for long-crested waves (narrow directional distribution). It also depends on the frequency resolution. It seems to work reasonably in many cases but it is a poor approximation for frequency resolutions with ratios very different from 10% (see command CGRID). This is a fundamental problem that SWAN shares with other third-generation wave models such as WAM and WAVEWATCH III. The LTA approximation for the triad wave-wave interactions depends on the width of the directional distribution of the wave spectrum. The present tuning in SWAN (the default settings, see command TRIAD) seems to work reasonably in many cases but it has been obtained from observations in a narrow wave flume (long-crested waves). As an option SWAN computes wave-induced set-up. In 1D cases the computations are based on exact equations. In 2D cases, the computations are based on approximate equations. This approximation in SWAN can only be applied to open coast (unlimited supply of water from outside the domain, e.g. nearshore coasts and estuaries) in contrast to closed basin, e.g. lakes, where this option should not be used. The effects of wave-induced currents are always ignored. SWAN does not calculate wave-induced currents. If relevant, such currents should be provided as input to SWAN, e.g. from a circulation model which can be driven by waves from SWAN in an iteration procedure. In areas where variations in wave height are large within a horizontal scale of a few wave lengths, diffraction should be used. However, the computation of diffraction in arbitrary geophysical conditions is rather complicated and requires considerable computing effort. To avoid this, a phase-decoupled approach is employed in SWAN so that same qualitative behaviour of spatial redistribution and changes in wave direction is obtained. This approach, however, does not properly handle diffraction in harbors or in front of reflecting obstacles.  
G
(see papers)  +
H
--  +
S
--  +
A
--  +
B
--  +
C
--  +
D
--  +
--  +
E
--  +
--  +
F
--  +
--  +
--  +
--  +
I
--  +
L
--  +
M
--  +
--  +
S
--  +
--  +
T
--  +
--  +
E
K
S
--  +
--  +
F
--  +
L
--  +
W
S
--  +
C
--  +
F
--  +
S
--  +
--  +
--  +
--  +
2
--  +
B
--  +
--  +
C
--  +
D
--  +
--  +
W
--  +
--  +
S
--  +
A
--  +
S
--  +
D
--  +
M
--  +
S
--  +
T
--  +
P
F
--  +
--  +
G
--  +
C
--  +
--  +
F
I
A
--  +
--  +
--  +
--  +
B
D
--  +
--  +
F
G
R
--  +
S
W
G
W
--  +
Q
--  +
S
--  +
R
--  +
--  +
W
--  +
--  +
Y
--  +
U
--  +
W
--  +
R
--  +
1
S
--  +
O
--  +
A
--  +
E
--  +
S
--  +
D
L
--  +
U
A
--  +
H
--  +
M
--  +
D
--  +
O
--  +
H
--  +
M
--  +
H
--  +
P
--  +
L
--  +
D
--  +
L
--  +
--  +
--  +
S
--  +
L
--  +
I
H
--  +
S
--  +
O
B
A compilation of examples and documentation can be found in this Badlands-doc GitHub repository (http://github.com/badlands-model/Badlands-doc).  +
C
A documented numerical stability criterium must be adhered to for the solution to remain stable.  +
G
Analytical solution produces unrealistic results with low elastic thickness and/or very large cells due to the Green's function approximation.  +
Any of the processes modeling can be obviously improved.  +
A
Artificially limit the alpha values between 2 and -2 and the constant values between 0 and 2. Done so to avoid extreme outliers.  +
H
As the size of the DEM increases, processing time will increase. An 8000*8000 DEM will take several days to process. Documentation is available to help users select appropriate input parameters.  +
B
Assumes Bingham rheology  +
S
Assumes constant flexural rigidity  +
C
Assumes mixed-grain sediment compacts linearly  +
F
At the present I am trying to improve the wetting and drying algorithm, unphysical large velocities are sometime produced at the wet/dry front.  +
B
Becomes unstable at timesteps much greater than 10 years.  +
S
Becomes very slow if run on domains with pits. Options include filling pits before running the model, or using the Landlab DepressionFinderAndRouter.  +
G
Can be very slow in complex systems.  +
Catchment must be small enough that it can be approximated as being flat.  +
A
Code is research grade  +
P
Code is research grade  +
G
Concerns about extrapolations beyond range of data used for model calibration and parameterization.  +
W
Covergency and instabiliy may occur depending the stiffness of the problems.  +
Q
Currently it is not possible to model transgression followed by regression.  +
G
Currently only shallow water equations are solved, using explicit finite volume methods. High order Boussinesq equations to better model dispersive waves (e.g. for short wavelength submarine landslide generated tsunamis) would require implicit time stepping and is still in the experimental phase.  +
E
D8 flow codes are used to compute contributing areas. Would be better to use D-Infinity or the Mass-Flux method.  +
L
Depends on application/process  +
P
Depends only on computer resources.  +
C
Diffusion and other sediment transport routines require short time steps  +
G
Documentation provides a description of the available processes and associated limitations: https://gospl.readthedocs.io/en/latest/tech_guide/index.html  +
Does not consider influence of cross-shore sand transport, not intended for short-term storm-induced shoreline change. No wave reflection from structures. No direct provision for changing tide level.  +
H
Dpeending on the size of the DEM this can be quite a slow process, particularly the third step.  +
W
Dynamics, chemistry, RT, microphysical, and slab ocean model are all coupled, thus the model runs slowly.  +
G
Equation set is still and uses robust but relatively inefficient solvers. Recommend testing with coarse grid resolutions (say, 20x20 cells) before attempting larger/finer grids.  +
Explicit finite volume routing formulations are time-step limited.  +
Explicit forward in time finite volume method limits maximum timestep/resolution combination. This can be managed using the adaptive timestep solver that is included.  +
W
Explicit schemes make high resolution runs expensive.  +
D
Explicit solver can go unstable, and it is not always obvious. Make sure to check resulting soil depths in the landscape to look for instabilities.  +
S
Extreme deformation of mesh is expected during large deformationin the updated Lagrangian formulation but prevented by regridding. The numerical diffusion due to regridding, however, can sometime make structures (e.g. shear localization) lose desired sharpness. Regridding itself fails from time to time.  +
B
Free surface formulation does not support rapid changes in water elevation like hydraulic jumps in rivers.  +
I
If you will be processing very large landscapes then you may need to configure PETSc with the --with-64-bit-indices option.  +
M
Implicit code. Very stable. Very efficient  +
T
In general, Matlab stores all data in the main memory. Manageable grid size will depend on your available RAM. For conveniently working with grids with ~5000x5000 rows and columns, a 4Gb of RAM will likely be sufficient.  +
C
Limited to medium-size resolutions. Typical runs 256x256  +
R
Linear wave theory  +
C
Matlab may max out memory if drainage basin is too large. Since this is a simplified landscape with wrap-around boundary conditions, one workaround is to do numerous smaller runs and add the output.  +
M
Maximum timestep must be determined by trial.  +
S
Model Assumptions * Mild bottom slope and negligible wave reflection * Spatially homogeneous offshore wave conditions * Steady-state waves, currents, and winds * Linear refraction and shoaling * Depth-uniform current * Bottom friction is neglected  +
T
Model Limitations * TOPMODEL only simulates watershed hydrology, although studies have been conducted to modify it to simulate water quality dynamics. * TOPMODEL can be applied most accurately to watersheds that do not suffer from excessively long dry periods and have shallow homogeneous soils and moderate topography. * Model results are sensitive to grid size, and grid size <=50 m is recommended.  +
I
Model has difficulty with negative (uphill) slopes  +
B
Model is solved implicitly, but can become inaccurate at very large (~1000 year) timesteps. When baselevel forcing is mild and block effects are significant, slope-inversion instabilities can develop. The model catches these and will not continue running.  +
E
Model limitations are related to the use of the goal seek function in excel to find the solution.  +
P
Model should never be numerically unstable but its behavior depends on ratios of various parameters. If the model seems to not be "doing anything", look at the parameter initialization functions in deltaRCM_tools.py  +
M
Model slows down as layers are added making long runs (>2000 years) impractical.  +
H
Model works well for resistant layer dips between 10 and 80 degrees. End members will work, but domain setup must be altered.  +
D
Most Dakota analysis techniques require multiple iterations of a model to explore a requested parameter space, so an experiment created with Dakotathon can take a long time to run and produce a lot of model output.  +
C
Most of the heavy lifting algorithms are implicit, thus numerically stable  +
H
N/A  +
N
S
None identified  +
R
None known; the model requires very little computational expense.  +
G
Numerical instabilities occur if the time step is too large.  +
C
Numerical limitations and issues: # Currently the model runs with a constant timestep, which is limited by the maximum inflow. Future versions may include adaptive time-stepping. # As mentioned above, the model channels tend to be one or two cells wide. Future versions may address this issue with some combination of diffusive regularization or multi-scale modeling.  +
A
Overall, the model is very computationally intensive. It is usually ran on a grid or a cluster.  +
T
Overland flow is currently modeled in a nonstandard way. Diffusive wave and dynamic wave routing routines need more testing. The linkage between the unsaturated zone (infiltration component) and saturated zone (subsurface flow component and water table) is not robust.  +
I
Poor scaling for ice-flow models with direct solvers (improves upon use of iterative solvers, but convergence is not systematic).  +
T
Presently limited to grids up to 4GB  +
C
Probably more than we know but none come to mind.  +
M
Quasi-static tide propagation. Flow neglected when water depth too small.  +
R
ROMS has a predictior-corrector algorithm that is efficient and accuarate. This class of model (terrain-following) exhibits stronger sensitivity to topography which results in pressure gradient errors. ROMS has several pressure gradient algorithms that minimize this problem.  +
C
ROMS has a predictior-corrector algorithm that is efficient and accuarate. This class of model (terrain-following) exhibits stronger sensitivity to topography which results in pressure gradient errors. ROMS has several pressure gradient algorithms that minimize this problem.  +
ROMS has a predictior-corrector algorithm that is efficient and accuarate. This class of model (terrain-following) exhibits stronger sensitivity to topography which results in pressure gradient errors. ROMS has several pressure gradient algorithms that minimize this problem.  +
U
ROMS has a predictior-corrector algorithm that is efficient and accuarate. This class of model (terrain-following) exhibits stronger sensitivity to topography which results in pressure gradient errors. ROMS has several pressure gradient algorithms that minimize this problem.  +
C
Run times can be long (60 +days for large areas over many 100's of years). Flow model is steady state  +
L
Runs slowly - iterates implicit scheme. Some sort of matrix algebra might improve speed.  +
M
Runs with grid sizes greater than about 600x600 may require many days on a PC. Model assumes fluvial streams have gradients determined by steady-state transport. Depositional stratigraphy not modeled.  +
S
SBEACH is an empirically based model that was developed for sandy beaches with uniform representative grain sized in the range of 0.2 to 0.42 mm. SBEACH should be tested or calibrated using data from beach profile surveyed before and after storms on the project coast.  +
P
See 'Description of Input and Examples for PHREEQC Version 3 - A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations'.  +
W
See WRF-Hydro Technical Description https://ral.ucar.edu/projects/wrf_hydro/technical-description-user-guide  +
E
See article: https://doi.org/10.3390/rs10121915  +
C
See documentation. The major limitation is computational time. This can be alleviated with sensible selection of module parameters. See documentation for guidance.  +
S
See manual  +
N
See related publication by J. A. Czuba.  +
R
See related publications by J. A. Czuba.  +
F
See: Version 2.0: Cohen et al. (2019), The Floodwater Depth Estimation Tool (FwDET v2.0) for Improved Remote Sensing Analysis of Coastal Flooding. Natural Hazards and Earth System Sciences (NHESS) Version 1.0: Cohen, S., G. R. Brakenridge, A. Kettner, B. Bates, J. Nelson, R. McDonald, Y. Huang, D. Munasinghe, and J. Zhang (2017), Estimating Floodwater Depths from Flood Inundation Maps and Topography. Journal of the American Water Resources Association (JAWRA):1–12.  +
G
Semi-implicit solution can decrease in accuracy for extremely long (hundreds of millions of years for typical input parameters) time steps  +
P
Solver is efficient and accurate for very stiff systems of equations  +
F
Some of the fuzzy logic methods do not produce unique results as there are a variety of method choices that best 'match' test data. For example, the user can choose a variety of aggregation methods to calculate final carbonate facies and productivity values.  +
A
Some parameter values result in channels that self-intersect. The code outputs both the raw centerline and a simplified centerline with self-intersections removed.  +
I
Some time step limitations due to the *semi* implicit nature of the code  +
C
Subject to CFL stability condition. Sharp depth changes can cause instability even with low Courant numbers. Pre-processing with depth_ssl is recommended (see Cliffs User Manual at http://arxiv.org/abs/1410.0753 )  +
T
TUGS was developed with a fairly low budget, and thus, bugs may still exist. There are, however, no known numerical limitations at this point.  +
S
The Courant number must be less than 1 at all times to maintain stability.  +
O
The calculations assume the waves are wind waves with periods in the range of 0-~30 s. If the waves are much larger or produced by a different mechanism, the calculations are not likely to be accurate.  +
T
The elevation specific equation of mass conservation is integrated with the Euler method. Thus, the user should carefully choose the spatial distance between computational nodes in the vertical and streamwise direction, as well as the temporal increment, to guarantee the numerical stability and mass conservation.  +
C
The fluvial sediment transport equations are quasi-diffusive and typically have orders of magnitude spatial variations in rate coefficient (reflecting differences in water discharge), which makes the system of equations stiff. Small time steps are typically required, which can lead to long compute times for large meshes.  +
S
The model cannot yet fully handle complex coastline geometries, such as those that cannot be represented (after rotation) by a single-valued function.  +
Q
The model does not allow for compaction  +
C
The model handles complex-shaped coastlines, such as cuspate-capes and spits. However, where the shoreline curvature becomes extreme (radius of curvature comparable to the cross-shore shoreface extent), as at the ends of spits, the assumptions of a locally rectilinear coordinate system break down, and sediment is conserved less rigorously locally. See Ashton and Murray (2006a) for details.  +
W
The model is abstract.  +
M
The model might be unstable if the meander bends are too sharp and/or flow parameters are somehow borderline.  +
R
The model was designed for laminar to transitional flows, up to 10 cm/s. Under these conditions, the flow velocity solution is approximate but is realistic and stable.  +
L
The profiles should not be spaced too closely in order to avoid an unstable saw-tooth longitudinal profile of the river.  +
C
This is a large model that takes significant computing resources to spin up run.  +
T
This model is only tested in rectangle domains, and compared the results with idealized experiments; Sediment bed state is affected by the initial condition (mainly due to the frictional stess closure in this model). For 2D or 3D runs, it is suggested that first run 1DV to steady or quasi-steady state, and map the 1DV results to 2DV or 3D, in this way, the initial instability of the sediment bed can be avoided.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +
This model/component needs more rigorous testing.  +