def readgeonet(geonetfile):
"""
Read strong motion data from a GeoNet data file
@param geonetfile: Path to a valid GeoNet data file.
@return: List of ObsPy Trace objects, containing accelerometer data in m/s.
"""
f = open(geonetfile, 'rt')
tracelist = []
headerlist = []
try:
hdrlines = readheaderlines(f)
except:
pass
while len(hdrlines[-1]):
hdrdict = readheader(hdrlines)
numlines = int(np.ceil(hdrdict['npts'] / 10.0))
data = []
for i in range(0, numlines):
line = f.readline()
parts = line.strip().split()
mdata = [float(p) for p in parts]
data = data + mdata
data = np.array(data)
header = hdrdict.copy()
stats = Stats(hdrdict)
trace = Trace(data, header=stats)
#apply the calibration and convert from mm/s^2 to m/s^2
trace.data = trace.data * trace.stats[
'calib'] * 0.001 #convert to m/s^2
tracelist.append(trace.copy())
headerlist.append(header.copy())
hdrlines = readheaderlines(f)
f.close()
return (tracelist, headerlist)
def readgeonet(geonetfile):
"""
Read strong motion data from a GeoNet data file
@param geonetfile: Path to a valid GeoNet data file.
@return: List of ObsPy Trace objects, containing accelerometer data in m/s.
"""
f = open(geonetfile,'rt')
tracelist = []
headerlist = []
try:
hdrlines = readheaderlines(f)
except:
pass
while len(hdrlines[-1]):
hdrdict = readheader(hdrlines)
numlines = int(np.ceil(hdrdict['npts']/10.0))
data = []
for i in range(0,numlines):
line = f.readline()
parts = line.strip().split()
mdata = [float(p) for p in parts]
data = data + mdata
data = np.array(data)
header = hdrdict.copy()
stats = Stats(hdrdict)
trace = Trace(data,header=stats)
#apply the calibration and convert from mm/s^2 to m/s^2
trace.data = trace.data * trace.stats['calib'] * 0.001 #convert to m/s^2
tracelist.append(trace.copy())
headerlist.append(header.copy())
hdrlines = readheaderlines(f)
f.close()
return (tracelist,headerlist)
def readiran(iranfile, doRotation=True):
"""
Read strong motion data from a Iran data file
@param iranfile: Path to a valid Iran data file.
@keyword doRotation: Apply back-azimuth rotation of L & T channels to NS and EW.
@return: List of ObsPy Trace objects, containing accelerometer data in m/s.
"""
f = open(iranfile, 'rt')
tracelist = []
headerlist = []
try:
hdrlines = readheaderlines(f)
except:
pass
while len(hdrlines[-1]):
hdrdict = readheader(hdrlines)
numlines = int(np.ceil(hdrdict['npts'] / 10.0))
data = []
for i in range(0, numlines):
line = f.readline()
parts = line.strip().split()
mdata = [float(p) for p in parts]
data = data + mdata
data = np.array(data)
header = hdrdict.copy()
stats = Stats(hdrdict)
trace = Trace(data, header=stats)
#apply the calibration and convert from mm/s^2 to m/s^2
trace.data = trace.data * trace.stats[
'calib'] * 0.98 #convert to m/s^2 from g/10
tracelist.append(trace.copy())
headerlist.append(header.copy())
endblock = f.readline()
hdrlines = readheaderlines(f)
f.close()
#data from Iran may be rotated so that one channel is aligned in the direction between the earthquake
#epicenter and the station. We want to rotate the data back so that we have what are presumably the
#original NS and EW channels. We presume that the "L" channel will rotate back to become NS, and
#"T" will become EW.
#First, find the channel called L*
if doRotation:
channels = [h['channel'][0:1] for h in headerlist]
lidx = channels.index('L')
tidx = channels.index('T')
ldata = tracelist[lidx].data
tdata = tracelist[tidx].data
backaz = headerlist[0]['rotation']['L']
ndata, edata = rotate.rotate_RT_NE(ldata, tdata, backaz)
tracelist[lidx].data = ndata.copy()
tracelist[lidx].stats[
'channel'] = 'H1' #most probably NS, but we're being cautious
tracelist[tidx].data = edata.copy()
tracelist[tidx].stats[
'channel'] = 'H2' #most probably EW, but we're being cautious
return (tracelist, headerlist)
def readiran(iranfile,doRotation=True):
"""
Read strong motion data from a Iran data file
@param iranfile: Path to a valid Iran data file.
@keyword doRotation: Apply back-azimuth rotation of L & T channels to NS and EW.
@return: List of ObsPy Trace objects, containing accelerometer data in m/s.
"""
f = open(iranfile,'rt')
tracelist = []
headerlist = []
try:
hdrlines = readheaderlines(f)
except:
pass
while len(hdrlines[-1]):
hdrdict = readheader(hdrlines)
numlines = int(np.ceil(hdrdict['npts']/10.0))
data = []
for i in range(0,numlines):
line = f.readline()
parts = line.strip().split()
mdata = [float(p) for p in parts]
data = data + mdata
data = np.array(data)
header = hdrdict.copy()
stats = Stats(hdrdict)
trace = Trace(data,header=stats)
#apply the calibration and convert from mm/s^2 to m/s^2
trace.data = trace.data * trace.stats['calib'] * 0.98 #convert to m/s^2 from g/10
tracelist.append(trace.copy())
headerlist.append(header.copy())
endblock = f.readline()
hdrlines = readheaderlines(f)
f.close()
#data from Iran may be rotated so that one channel is aligned in the direction between the earthquake
#epicenter and the station. We want to rotate the data back so that we have what are presumably the
#original NS and EW channels. We presume that the "L" channel will rotate back to become NS, and
#"T" will become EW.
#First, find the channel called L*
if doRotation:
channels = [h['channel'][0:1] for h in headerlist]
lidx = channels.index('L')
tidx = channels.index('T')
ldata = tracelist[lidx].data
tdata = tracelist[tidx].data
backaz = headerlist[0]['rotation']['L']
ndata,edata = rotate.rotate_RT_NE(ldata,tdata,backaz)
tracelist[lidx].data = ndata.copy()
tracelist[lidx].stats['channel'] = 'H1' #most probably NS, but we're being cautious
tracelist[tidx].data = edata.copy()
tracelist[tidx].stats['channel'] = 'H2' #most probably EW, but we're being cautious
return (tracelist,headerlist)
def getCatData(date, opt):
"""
Download data from IRIS or Earthworm waveserver with padding and filter it. This is
a specialized version getData() for catalog events, pulling a smaller amount of time
around a known event.
date: UTCDateTime of known catalog event
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
if opt.server == "IRIS":
client = Client("IRIS")
else:
client = EWClient(opt.server, opt.port)
st = Stream()
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
date - opt.atrig, date + 3*opt.atrig)
stmp = stmp.filter("bandpass", freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.merge(method=1, fill_value='interpolate')
except (obspy.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
date - opt.atrig, date + 3*opt.atrig)
stmp = stmp.filter("bandpass", freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.merge(method=1, fill_value='interpolate')
except (obspy.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Resample to ensure all traces are same length
if stmp[0].stats.sampling_rate != opt.samprate:
stmp = stmp.resample(opt.samprate)
st.extend(stmp.copy())
st = st.trim(starttime=date-opt.atrig, endtime=date+3*opt.atrig, pad=True,
fill_value=0)
stC = st.copy()
return st, stC
def readgeonet(geonetfile):
"""
Read strong motion data from a GeoNet data file
@param geonetfile: Path to a valid GeoNet data file.
@return: List of ObsPy Trace objects, containing accelerometer data in m/s.
"""
#notes on implementation:
# originally I had written code to read each line of data manually, just
# as I was reading the header lines. However, this became VERY slow for large
# files. I discovered that numpy's genfromtxt function was much faster. However,
# I could not get that function to work when I passed it a file object instead of a file name.
# Consequently, the only way I could keep the genfromtxt() method and file pointer in sync was
# to read and discard all of the lines of data that genfromtxt() parsed. While annoying, the combination
# of these two seems to still be at least an order of magnitude faster than manually reading the file.
f = open(geonetfile,'rt')
tracelist = []
headerlist = []
totlines = 0
hdrlines = readheaderlines(f)
while len(hdrlines[-1]):
totlines += len(hdrlines)
hdrdict = readheader(hdrlines)
numlines = int(np.ceil(hdrdict['npts']/10.0))
data = np.genfromtxt(geonetfile,skip_header=totlines,max_rows=numlines)
totlines += numlines
#now we need to set the file position to where we just ended
for i in range(0,numlines):
f.readline()
data = data.flatten()
header = hdrdict.copy()
stats = Stats(hdrdict)
trace = Trace(data,header=stats)
#apply the calibration and convert from mm/s^2 to m/s^2
trace.data = trace.data * trace.stats['calib'] * 0.001 #convert to m/s^2
tracelist.append(trace.copy())
headerlist.append(header.copy())
hdrlines = readheaderlines(f)
f.close()
return (tracelist,headerlist)
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'file':
# Generate list of files
if opt.server == 'file':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,opt.filepattern)) for root, dirs, files in os.walk(opt.searchdir)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[-1].stats.endtime
if (ststart<=tstart and tstart<=stend) or (ststart<=tend and
tend<=stend) or (tstart<=stend and ststart<=tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend+opt.maxdt)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m] and nets[n] in
netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find "+stas[n]+'.'+chas[n]+'.'+nets[n]+'.'+locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
for m in range(len(stmp)):
stmp[m].data = np.where(stmp[m].data == -2**31, 0, stmp[m].data) # replace -2**31 (Winston NaN token) w 0
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
for m in range(len(stmp)):
stmp[m].data = np.where(stmp[m].data == -2**31, 0, stmp[m].data) # replace -2**31 (Winston NaN token) w 0
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
# Edit 'start' time if using offset option
if opt.maxdt:
dts = np.fromstring(opt.offset, sep=',')
for n, tr in enumerate(st):
tr.stats.starttime = tr.stats.starttime-dts[n]
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
def grab_file_data(filepath, scnl, tstart, tend, fill_value=0):
import obspy
from obspy import Stream, Trace
import glob, os, itertools
stas = []
chas = []
nets = []
locs = []
for sta in scnl:
stas.append(sta.split('.')[0])
chas.append(sta.split('.')[1])
nets.append(sta.split('.')[2])
if len(sta) == 4:
locs.append(sta.split('.')[3])
else:
locs.append('')
st = Stream()
#if opt.server == 'file':
if True:
# Generate list of files
#if opt.server == 'file':
flist = list(
itertools.chain.from_iterable(
glob.iglob(os.path.join(root, "*"))
for root, dirs, files in os.walk(filepath)))
# "*" takes the place of wildcard lists, see REDPy documentation
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart <= tstart and tstart <= stend) or (
ststart <= tend
and tend <= stend) or (tstart <= stend
and ststart <= tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
#stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
# zerophase=True)
#stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
#for m in range(len(stmp)):
# if stmp[m].stats.sampling_rate != opt.samprate:
# stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=fill_value)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find " + stas[n] + '.' + chas[n] + '.' +
nets[n] + '.' + locs[n])
trtmp = Trace()
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
if len(st) > 1:
if fill_value == 0 or fill_value == None:
st.detrend('demean')
st.taper(max_percentage=0.01)
st.merge(fill_value=fill_value)
st.trim(tstart, tend, pad=0)
st.detrend('demean')
print(st)
return st
def get_stream(datasource,
scnl,
tstart,
tend,
fill_value=0,
filepattern='*',
filter=None,
samprate=100,
verbose=False):
"""
Generalized (and more robust) way to retrieve waveform data through ObsPy
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
filepattern='*'
You can specify a pattern for your files to reduce the files within the directory
searched. For example, filepattern=2019.06.*.mseed if your files are miniSEED files
named by date and you only want those from June 2019. Simple wildcarding is supported
(i.e., * and ?, [] for ranges of values or lists) but not full regular expressions.
samprate=100
Resamples all waveforms to the same sample rate.
Returns ObsPy stream objects
Based on code by Alicia Hotovec-Ellis and Aaron Wech.
Example:
>>> get_stream(['vdap.org', 16024], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
>>> get_stream(['file', '/Users/vdapseismo/data/'], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
>>> get_stream(['IRIS'], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
"""
from obspy import UTCDateTime
import obspy
from obspy.clients.fdsn import Client
from obspy.clients.earthworm import Client as EWClient
from obspy.core.trace import Trace
from obspy.core.stream import Stream
from obspy.signal.trigger import coincidence_trigger
import numpy as np
from scipy import stats
from scipy.fftpack import fft
import glob, os, itertools
#print(datasource)
#print(scnl)
#print(tstart)
#print(tend)
tstart = UTCDateTime(tstart)
tend = UTCDateTime(tend)
nets = []
stas = []
locs = []
chas = []
for s in scnl:
#print(s)
nets.append(s.split('.')[2])
stas.append(s.split('.')[0])
locs.append(s.split('.')[3])
chas.append(s.split('.')[1])
st = Stream()
if '/' in datasource:
# Retrieve data from file structure
flist = list(
itertools.chain.from_iterable(
glob.iglob(os.path.join(root, filepattern))
for root, dirs, files in os.walk(datasource)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart <= tstart and tstart <= stend) or (
ststart <= tend
and tend <= stend) or (tstart <= stend
and ststart <= tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# merge
stmp = stmp.taper(max_percentage=0.01)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.merge(method=1, fill_value=fill_value)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print('No data found for {}.{}.{}.{}'.format(
stas[n], chas[n], nets[n], locs[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
# retrieve data from server
if '.' not in datasource:
client = Client(datasource)
else:
datasource = datasource.split(':')
client = EWClient(datasource[0], int(datasource[1]))
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend)
for m in range(len(stmp)):
#stmp[m].data = np.ma.masked_where(stmp[m].data == -2**31, stmp[m].data) # masks out all values of -2**31 (Winston NaN Token)
#stmp[m] = stmp[m].split().merge(method=0, fill_value='interpolate')[0] # splits trace at masked values; then re-merges using linear interpolation
stmp[m].data = np.where(stmp[m].data == -2**31, 0,
stmp[m].data)
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.taper(max_percentage=0.01)
stmp = stmp.merge(method=1, fill_value=fill_value)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n],
chas[n], tstart, tend)
for m in range(len(stmp)):
#stmp[m].data = np.ma.masked_where(stmp[m].data == -2**31, stmp[m].data) # masks out all values of -2**31 (Winston NaN Token)
#stmp[m] = stmp[m].split().merge(method=0, fill_value='interpolate')[0] # splits trace at masked values; then re-merges using linear interpolation
stmp[m].data = np.where(stmp[m].data == -2**31, 0,
stmp[m].data)
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.taper(max_percentage=0.01)
stmp = stmp.merge(method=1, fill_value=fill_value)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n], nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {}.{}.{}.{}'.format(
stas[n], chas[n], nets[n], locs[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
st = st.trim(starttime=tstart,
endtime=tend,
pad=True,
fill_value=fill_value)
return st
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'SAC' or opt.server == 'miniSEED':
# Generate list of files
if opt.server == 'SAC':
flist = glob.glob(opt.sacdir + '*.sac') + glob.glob(opt.sacdir +
'*.SAC')
elif opt.server == 'miniSEED':
flist = glob.glob(opt.mseeddir +
'*.mseed') + glob.glob(opt.mseeddir + '*.MSEED')
# Load data from file
stmp = Stream()
for f in flist:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05, type='hann', max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find " + stas[n] + '.' + chas[n] + '.' +
nets[n] + '.' + locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend)
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05, type='hann', max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n],
chas[n], tstart, tend)
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05,
type='hann',
max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n], nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
def detect_stalta(self):
STA_len = self.params["STA_len"]
LTA_len = self.params["LTA_len"]
STALTA_thresh = self.params["STALTA_thresh"]
det_off_win = self.params["no_det_win"]
try:
devices = self.traces.data["device_id"].unique()
except Exception as exception:
logging.error(exception)
devices = []
for device in devices:
for channel in ["x", "y", "z"]:
trace = self.traces.data[self.traces.data["device_id"] ==
device][channel].to_numpy()
time = self.traces.data[self.traces.data["device_id"] ==
device]["cloud_t"]
sr = self.traces.data[self.traces.data["device_id"] ==
device]["sr"].iloc[0]
if len(trace) > int(np.ceil(sr * (STA_len + LTA_len))):
# set new trace
tr = Trace()
tr.data = trace
tr.stats.delta = 1 / sr
tr.stats.channel = channel
tr.stats.station = device
# tr.filter("highpass", freq=0.2)
tr.detrend(type="constant")
tr_orig = tr.copy()
std = np.std(tr_orig.data[:int(STA_len * sr)])
tr.trigger(
"recstalta",
sta=self.params["STA_len"],
lta=self.params["LTA_len"],
)
(ind, ) = np.where(tr.data > STALTA_thresh)
if len(ind) > 0:
det_time = time.iloc[ind[0]]
past_detections = self.detections.data[
(self.detections.data["device_id"] == device)
& (self.detections.data["cloud_t"] - det_time +
det_off_win) > 0]
if (past_detections.shape[0]
== 0) & (std <= self.params["max_std"]):
# Get event ID
# timestamp
timestamp = datetime.utcfromtimestamp(det_time)
year = str(timestamp.year - 2000).zfill(2)
month = str(timestamp.month).zfill(2)
day = str(timestamp.day).zfill(2)
hour = str(timestamp.hour).zfill(2)
minute = str(timestamp.minute).zfill(2)
detection_id = "D_" + year + month + day + hour + minute
new_detection = pd.DataFrame(
{
"detection_id": detection_id,
"device_id": device,
"cloud_t": det_time,
"mag1": None,
"mag2": None,
"mag3": None,
"mag4": None,
"mag5": None,
"mag6": None,
"mag7": None,
"mag8": None,
"mag9": None,
"event_id": None,
},
index=[0],
)
# plot all detections and save in obj/detections folder
if self.params["plot_detection"]:
self.plot_detection(tr_orig, tr, device,
detection_id, std)
self.detections.update(new_detection)
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'SAC' or opt.server == 'miniSEED':
# Generate list of files
if opt.server == 'SAC':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.sac')) for root, dirs, files in os.walk(opt.sacdir)))+list(
itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.SAC')) for root, dirs, files in os.walk(opt.sacdir)))
elif opt.server == 'miniSEED':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.mseed')) for root, dirs, files in os.walk(opt.mseeddir)))+list(
itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.MSEED')) for root, dirs, files in os.walk(opt.mseeddir)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart<=tstart and tstart<=stend) or (ststart<=tend and
tend<=stend) or (tstart<=stend and ststart<=tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend+opt.maxdt)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m] and nets[n] in
netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find "+stas[n]+'.'+chas[n]+'.'+nets[n]+'.'+locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
# Edit 'start' time if using offset option
if opt.maxdt:
dts = np.fromstring(opt.offset, sep=',')
for n, tr in enumerate(st):
tr.stats.starttime = tr.stats.starttime-dts[n]
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
