Mondaic%matplotlib inline
import os
import numpy as np
import xarray as xr
import salvus.namespace as sn
SALVUS_FLOW_SITE_NAME = os.environ.get("SITE_NAME", "local")One of the most exciting features of the spectral-element method is the ease with which surface topography or other nonuniform boundaries can be represented in a simulation mesh. In this short tutorial, we will generate a custom topographic model and apply it to a simulation in both 2- and 3-D. The simple trigonometric definition of topography specified below can be expanded to provide much more complexity. Try it out yourself!
d = sn.domain.dim2.BoxDomain(x0=0, x1=20e3, y0=-5e3, y1=0)p = sn.Project.from_domain(path="proj", domain=d, load_if_exists=True)
max_x = d.bounding_box[:, 0].max()
x = np.linspace(0, max_x, 10000)
y = np.sin(x * 2 * np.pi / max_x) * 1e3Wrap in an xarray dataset, specifying the reference elevation
Before application to the mesh, the reference elevation will be subtracted from the topography model in the digital elevation model (DEM).
ds = xr.Dataset(
data_vars={"dem": (["x"], y, {"reference_elevation": 0.0})},
coords={"x": x},
)tm = sn.topography.cartesian.SurfaceTopography(name="surface", data=ds)
tm.ds.dem.plot()[<matplotlib.lines.Line2D at 0x7fdb66718e90>]
p.add_to_project(tm, overwrite=True)vp, vs, rho = 4e3, 2.4e3, 2.6e3
bm = sn.model.background.homogeneous.IsotropicElastic(vp=vp, vs=vs, rho=rho)src = sn.simple_config.source.cartesian.SideSetVectorPoint2D(
point=(5.0e3, 0.0),
direction="y",
offset=-10.0,
side_set_name="y1",
fx=0.0,
fy=1.0,
)recs = [
sn.simple_config.receiver.cartesian.SideSetPoint2D(
point=(x, 0.0),
direction="y",
offset=0.0,
side_set_name="y1",
station_code=f"XX_{i}",
fields=["velocity", "strain"],
)
for i, x in enumerate(np.linspace(12.5e3, 17.5e3, 11))
]p.add_to_project(sn.EventCollection.from_sources(sources=src, receivers=recs))f_max = 10.0
p.add_to_project(
sn.SimulationConfiguration(
tensor_order=2,
name="sim_with_topo",
elements_per_wavelength=2.0,
max_frequency_in_hertz=f_max,
model_configuration=sn.ModelConfiguration(background_model=bm),
topography_configuration=sn.TopographyConfiguration("surface"),
event_configuration=sn.EventConfiguration(
sn.simple_config.stf.Ricker(center_frequency=f_max / 2),
sn.WaveformSimulationConfiguration(end_time_in_seconds=10.0),
),
absorbing_boundaries=sn.AbsorbingBoundaryParameters(
reference_velocity=vp,
number_of_wavelengths=0.0,
reference_frequency=f_max / 2,
),
),
overwrite=True,
)p.viz.nb.simulation_setup("sim_with_topo", events="event_0000")[2024-07-05 09:55:12,678] INFO: Creating mesh. Hang on.
<salvus.flow.simple_config.simulation.waveform.Waveform object at 0x7fdb5f7ffd90>
p.simulations.launch(
"sim_with_topo",
events="event_0000",
site_name=SALVUS_FLOW_SITE_NAME,
ranks_per_job=1,
)
p.simulations.query(block=True)[2024-07-05 09:55:13,177] INFO: Submitting job ... Uploading 1 files... 🚀 Submitted job_2407050955303727_c1f2e83ee4@local
True
p.waveforms.get(
data_name="SYNTHETIC_DATA:sim_with_topo", events=["event_0000"]
)[0].plot(component="Y", receiver_field="velocity")
Tutorial coming soon! In the meantime, check out our tutorial on running simulations through a model of Mt St. Helens.
PAGE CONTENTS