-
Notifications
You must be signed in to change notification settings - Fork 0
/
dctoip.py
169 lines (142 loc) · 4.83 KB
/
dctoip.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
"""
2.5D DC-IP inversion of Dipole Dipole array with Topography
===========================================================
This is an example for 2.5D DC-IP Inversion.
For DC inversion, a resisistivity model (Ohm-m) is generated having conductive
and resistive cylinders; they are respectively located right and left sides
of the subsurface.
For IP inversion, a chargeability model (V/V) is generated having a chargeable
cyinder located at the center.
Default `survey_type` is dipole-dipole, but this can be changed.
to 'pole-dipole', 'dipole-pole', and 'pole-pole'.
By running DC and IP simulations synthetic DC and IP data are generated,
respectively. Following two-stage approach (Oldenburg et al, 1999),
first DC data is inverted to recover a resistivity model. Then by using
the obtained resistivity model, sensitivity function is formed and used for
subsequent IP inversion to recover a chargeability model.
"""
from SimPEG import DC, IP
from SimPEG import Maps, Utils
import matplotlib.pyplot as plt
from matplotlib import colors
import numpy as np
from pylab import hist
try:
from pymatsolver import Pardiso as Solver
except ImportError:
from SimPEG import SolverLU as Solver
survey_type='pole-dipole'
np.random.seed(1)
# Initiate I/O class for DC
IO = DC.IO()
# Obtain ABMN locations
xmin, xmax = 0., 200.
ymin, ymax = 0., 0.
zmin, zmax = 0, 0
endl = np.array([[xmin, ymin, zmin], [xmax, ymax, zmax]])
# Generate DC survey object
survey_dc = DC.Utils.gen_DCIPsurvey(endl, survey_type=survey_type, dim=2,
a=10, b=10, n=10)
survey_dc.getABMN_locations()
survey_dc = IO.from_ambn_locations_to_survey(
survey_dc.a_locations, survey_dc.b_locations,
survey_dc.m_locations, survey_dc.n_locations,
survey_type, data_dc_type='volt', data_ip_type='volt'
)
# Obtain 2D TensorMesh
mesh, actind = IO.set_mesh()
topo, mesh1D = DC.Utils.genTopography(mesh, -10, 0, its=100)
actind = Utils.surface2ind_topo(mesh, np.c_[mesh1D.vectorCCx, topo])
survey_dc.drapeTopo(mesh, actind, option="top")
# Build conductivity and chargeability model
blk_inds_c = Utils.ModelBuilder.getIndicesSphere(
np.r_[60., -25.], 12.5, mesh.gridCC
)
blk_inds_r = Utils.ModelBuilder.getIndicesSphere(
np.r_[140., -25.], 12.5, mesh.gridCC
)
blk_inds_charg = Utils.ModelBuilder.getIndicesSphere(
np.r_[100., -25], 12.5, mesh.gridCC
)
sigma = np.ones(mesh.nC)*1./100.
sigma[blk_inds_c] = 1./10.
sigma[blk_inds_r] = 1./1000.
sigma[~actind] = 1./1e8
rho = 1./sigma
charg = np.zeros(mesh.nC)
charg[blk_inds_charg] = 0.1
# Use Exponential Map: m = log(rho)
actmap = Maps.InjectActiveCells(
mesh, indActive=actind, valInactive=np.log(1e8)
)
mapping = Maps.ExpMap(mesh) * actmap
# Generate mtrue_dc for resistivity
mtrue_dc = np.log(rho[actind])
# Generate 2.5D DC problem
# "N" means potential is defined at nodes
prb = DC.Problem2D_N(
mesh, rhoMap=mapping, storeJ=True,
Solver=Solver
)
# Pair problem with survey
try:
prb.pair(survey_dc)
except:
survey_dc.unpair()
prb.pair(survey_dc)
# Make synthetic DC data with 5% Gaussian noise
dtrue_dc = survey_dc.makeSyntheticData(mtrue_dc, std=0.05, force=True)
IO.data_dc = dtrue_dc
# Generate mtrue_ip for chargability
mtrue_ip = charg[actind]
# Generate 2.5D DC problem
# "N" means potential is defined at nodes
prb_ip = IP.Problem2D_N(
mesh, etaMap=actmap, storeJ=True, rho=rho,
Solver=Solver
)
survey_ip = IP.from_dc_to_ip_survey(survey_dc, dim="2.5D")
prb_ip.pair(survey_ip)
dtrue_ip = survey_ip.makeSyntheticData(mtrue_ip, std=0.05)
IO.data_ip = dtrue_ip
# Set initial model based upon histogram
m0_dc = np.ones(actmap.nP)*np.log(100.)
# Set uncertainty
# floor
eps_dc = 10**(-3.2)
# percentage
std_dc = 0.05
mopt_dc, pred_dc = DC.run_inversion(
m0_dc, survey_dc, actind, mesh, std_dc, eps_dc,
beta0_ratio=1e0,
use_sensitivity_weight=True
)
# Convert obtained inversion model to resistivity
# rho = M(m), where M(.) is a mapping
rho_est = mapping*mopt_dc
rho_est[~actind] = np.nan
rho_true = rho.copy()
rho_true[~actind] = np.nan
# Set initial model based upon histogram
m0_ip = np.ones(actmap.nP)*1e-10
# Set uncertainty
# floor
eps_ip = 10**(-4)
# percentage
std_ip = 0.05
# Clean sensitivity function formed with true resistivity
prb_ip._Jmatrix = None
# Input obtained resistivity to form sensitivity
prb_ip.rho = mapping*mopt_dc
mopt_ip, _ = IP.run_inversion(
m0_ip, survey_ip, actind, mesh, std_ip, eps_ip,
upper=np.Inf, lower=0.,
beta0_ratio=1e0,
use_sensitivity_weight=True
)
# Convert obtained inversion model to chargeability
# charg = M(m), where M(.) is a mapping for cells below topography
charg_est = actmap*mopt_ip
charg_est[~actind] = np.nan
charg_true = charg.copy()
charg_true[~actind] = np.nan