class getwaveform:
def __init__(self):
"""
---------------- copied from old getwaveform.py ----------------------
min_dist - minimum station distance (default = 20)
max_dist - maximum station distance (default =300)
before - time window length before the event time (default= 100)
after - time window length after the event time (default = 300)
network - network codes of potential stations (default=*)
channel - component(s) to get, accepts comma separated (default='BH*')
ifresample_TF - Boolean. Request resample or not. Default = False
resample_freq - sampling frequency to resample waveforms (default 20.0)
ifrotate - Boolean, if true will output sac files rotated to baz
unrotated sac files will also be written
ifCapInp - Boolean, make weight files for CAP
ifEvInfo - Boolean, output 'ev_info.dat' containg event info (True)
ifRemoveResponse - Boolean, will remove response (True)
ifDetrend - Boolean, will remove linear trend from data (True)
ifDemean - Boolean, will insult the data (True)
scale_factor - scale all data by one value (10.0**2)
This usually puts the data in the units required by CAP
From m/s to cm/s
pre_filt - list, corner frequencies of filter to apply before deconv
a good idea when deconvolving (ifRemoveResponse=True)
"""
# DEFAULT SETTINGS
self.idb = 1 # =1-IRIS (default); =2-AEC; =3-LLNL; =4-Geoscope
self.client = Client()
# event parameters
self.use_catalog = 1 # =1: use an existing catalog (=1); =0: specify your own event parameters (see iex=9)
self.sec_before_after_event = 10 # time window to search for a target event in a catalog
self.min_dist = 0
self.max_dist = 20000
self.min_az = 0
self.max_az = 360
self.min_lat = None
self.max_lat = None
self.min_lon = None
self.max_lon = None
self.tbefore_sec = 100
self.tafter_sec = 300
# These are used only if self.use_catalog = 0
self.evname = None
self.otime = None
self.elat = None
self.elon = None
self.edep = None
self.emag = None
# Refernce origin (dummy values)
self.rlat = None
self.rlon = None
self.rtime = None
# event objects
self.ev = Event()
self.ref_time_place = Event()
# station parameters
self.network = '*' # all networks
self.station = '*,-PURD,-NV33,-GPO' # all stations except -(these)
self.channel = '*' # all channels
self.overwrite_ddir = 1 # 1 = delete data directory if it already exists
self.icreateNull = 1 # create Null traces so that rotation can work (obsby stream.rotate require 3 traces)
self.phase_window = False # Grab waveforms using phases
self.phases = ["P",
"P"] # Phases to write to sac files or grab data from
self.write_sac_phase = False # put phase information in sac files
self.taupmodel = "ak135"
# Filter parameters
self.ifFilter = False
#------Filter--------------
# for 'bandpass' both f1 and f2 are used
# for 'lowpass' only f2 is used
# for 'highpass' only f1 is used
#
# EXAMPLES
# ifFilter zerophase remove_response ipre_filt
# A. CAP-ready waveforms [DEFAULT]: False NA True 1
# B. plot-ready waveforms, acausal: True True True 2
# C. plot-ready, causal waveforms: True False True 0
# D. possible sensor issues: True False False NA
#
self.filter_type = 'bandpass'
# f1 should consider the requested length of the time series
# f2 should consider the sampling rate for the desired channels
self.f1 = 1 / 40 # fmin - highpass will keep frequencies larger than fmin
self.f2 = 1 / 10 # fmax - lowpass will keep frequencies lower than fmax
self.zerophase = True # = False (causal/one-pass), = True (acausal/two-pass)
# 4 pole filter is more sharper at the edges than 2 pole
self.corners = 4 # Is corner in Obspy same as Pole in SAC?
# Pre-filter parameters
self.ipre_filt = 1 # =0 No pre_filter
# =1 default pre_filter (see getwaveform_iris.py)
# =2 user-defined pre_filter (use this if you are using bandpass filter)
# For tapering down the pre-filter
# Perhaps you want to set ipre_filt = 0 to prevent extra filtering
# pre-filter for deconvolution
# https://ds.iris.edu/files/sac-manual/commands/transfer.html
# Pre-filter will not be applied if remove_response = False
self.f0 = 0.5 * self.f1
self.f3 = 2.0 * self.f2
self.pre_filt = (self.f0, self.f1, self.f2, self.f3
) # applies for ipre_filt = 2 only
# self.pre_filt = (0.005, 0.006, 10.0, 15.0) # BH default
self.water_level = 60
# For CAP
self.resample_TF = True # if False then resample_freq is taken from SAC files
self.resample_freq = 50 # 0 causes errors. Use resample_TF instead
self.scale_factor = 1 # for CAP use 10**2 (to convert m/s to cm/s)
# Pre-processing (mainly for CAP)
self.output_cap_weight_file = True # output cap weight files
self.detrend = True # detrend waveforms
self.demean = True # demean waveforms
self.taper = False # this could also be a fraction between 0 and 1 (fraction to be tapered from both sides)
self.output_event_info = True # output event info file
self.outformat = 'VEL' # Instrument-response-removed waveforms saved as VEL, DISP, or ACC
self.ifsave_sacpaz = False # save sac pole zero (needed as input for MouseTrap module)
self.remove_response = True # remove instrument response
self.iplot_response = False # plot response function
self.ifplot_spectrogram = False # plot spectrograms
self.ifsave_stationxml = True # save station xml file (for adjoint tomo)
# Waveforms to be saved
self.rotateRTZ = True # Rotate and save the RTZ components
self.rotateUVW = False # Rotate and save the UVW components
self.isave_raw = False # save raw waveforms
self.isave_raw_processed = False # save processed waveforms just before rotation to ENZ
#self.rotateENZ = True # rotate extracted waveforms to ENZ
self.isave_ENZ = True # save ENZ
# username and password for embargoed IRIS data
# Register here: http://ds.iris.edu/ds/nodes/dmc/forms/restricted-data-registration/
self.user = None
self.password = None
# To output lots of processing info
self.ifverbose = False
# save RTZ as asdf files
self.ifsave_asdf = False
def run_get_waveform(self):
"""
Get SAC waveforms for an event
basic usage:
run_get_waveform(event)
c - client
event - obspy Event object
ref_time_place - reference time and place (other than origin time and place - for station subsetting)
"""
c = self.client
event = self.ev
ref_time_place = self.ref_time_place
evtime = event.origins[0].time
reftime = ref_time_place.origins[0].time
if self.idb == 1:
print("Preparing request for IRIS ...")
# BK network doesn't return data when using the IRIS client.
# this option switches to NCEDC if BK is
if "BK" in self.network:
client_name = "NCEDC"
print(
"\nWARNING. Request for BK network. Switching to NCEDC client"
)
c = Client("NCEDC")
else:
client_name = "IRIS"
print("Download stations...")
stations = c.get_stations(network=self.network,
station=self.station,
channel=self.channel,
starttime=reftime - self.tbefore_sec,
endtime=reftime + self.tafter_sec,
minlatitude=self.min_lat,
maxlatitude=self.max_lat,
minlongitude=self.min_lon,
maxlongitude=self.max_lon,
level="response")
inventory = stations # so that llnl and iris scripts can be combined
if self.ifverbose:
print("Printing stations")
print(stations)
print("Done Printing stations...")
sta_limit_distance(ref_time_place,
stations,
min_dist=self.min_dist,
max_dist=self.max_dist,
min_az=self.min_az,
max_az=self.max_az,
ifverbose=self.ifverbose)
#print("Printing stations NEW")
#print(stations)
#print("Done Printing stations...")
#stations.plotprojection="local")
# Find P and S arrival times
phases = self.phases
t1s, t2s = get_phase_arrival_times(stations, event, self.phases,
self.phase_window,
self.taupmodel, reftime,
self.tbefore_sec,
self.tafter_sec)
print("Downloading waveforms...")
# this needs to change
bulk_list = make_bulk_list_from_stalist(stations,
t1s,
t2s,
channel=self.channel)
stream_raw = c.get_waveforms_bulk(bulk_list)
# save ev_info object
pickle.dump(
self,
open(self.evname + '/' + self.evname + '_ev_info.obj', 'wb'))
elif self.idb == 3:
client_name = "LLNL"
print("Preparing request for LLNL ...")
# Get event an inventory from the LLNL DB.
event_number = int(event.event_descriptions[0].text)
# event = llnl_db_client.get_obspy_event(event)
inventory = c.get_inventory()
nsta_llnl = len(inventory.get_contents()["stations"])
print("--> Total stations in LLNL DB: %i" % nsta_llnl)
sta_limit_distance(event,
inventory,
min_dist=self.min_dist,
max_dist=self.max_dist,
min_az=self.min_az,
max_az=self.max_az)
print("--> Stations after filtering for distance: %i" %
(len(inventory.get_contents()["stations"])))
stations = set([sta.code for net in inventory for sta in net])
_st = c.get_waveforms_for_event(event_number)
stream_raw = obspy.Stream()
for tr in _st:
if tr.stats.station in stations:
stream_raw.append(tr)
# set reftime
stream = obspy.Stream()
stream = set_reftime(stream_raw, evtime)
print("--> Adding SAC metadata...")
if self.ifverbose: print(inventory)
st2 = add_sac_metadata(stream,
idb=self.idb,
ev=event,
stalist=inventory,
taup_model=self.taupmodel,
phases=phases,
phase_write=self.write_sac_phase)
# Do some waveform QA
do_waveform_QA(st2, client_name, event, evtime, self.tbefore_sec,
self.tafter_sec)
if self.demean:
st2.detrend('demean')
if self.detrend:
st2.detrend('linear')
if self.ifFilter:
prefilter(st2, self.f1, self.f2, self.zerophase, self.corners,
self.filter_type)
if self.remove_response:
resp_plot_remove(st2, self.ipre_filt, self.pre_filt,
self.iplot_response, self.water_level,
self.scale_factor, stations, self.outformat,
self.ifverbose)
else:
# output RAW waveforms
decon = False
print("WARNING -- NOT correcting for instrument response")
if self.scale_factor > 0:
amp_rescale(st2, self.scale_factor)
if self.idb == 3:
amp_rescale_llnl(st2, self.scale_factor)
# Set the sac header KEVNM with event name
# This applies to the events from the LLNL database
# NOTE this command is needed at the time of writing files, so it has to
# be set early
st2, evname_key = rename_if_LLNL_event(st2, evtime)
self.evname = evname_key
# save station plot
# Note: Plotted are stations in the inventory and NOT the ones with the traces
# It could be possible that there might not be waveforms for some of these stations.
try:
fig = inventory.plot(projection="local",
resolution="i",
label=False,
show=False)
Catalog([self.ev]).plot(fig=fig,
outfile=self.evname + '/station_map.pdf')
except:
print("There is a problem with creating the station map!")
# Get list of unique stations + locaiton (example: 'KDAK.00')
stalist = []
for tr in stream.traces:
if self.ifverbose: print(tr)
station_key = "%s.%s.%s.%s" % (tr.stats.network, tr.stats.station,
tr.stats.location,
tr.stats.channel[:-1])
stalist.append(station_key)
# Crazy way of getting a unique list of stations
stalist = list(set(stalist))
# Resample
if self.resample_TF == True:
# NOTE !!! tell the user if BOTH commands are disabled NOTE !!!
if (client_name == "IRIS"):
resample(st2, freq=self.resample_freq)
elif (client_name == "LLNL"):
resample_cut(st2, self.resample_freq, evtime, self.tbefore_sec,
self.tafter_sec)
else:
print(
"WARNING. Will not resample. Using original rate from the data"
)
# match start and end points for all traces
st2 = trim_maxstart_minend(stalist, st2, client_name, event, evtime,
self.resample_TF, self.resample_freq,
self.tbefore_sec, self.tafter_sec,
self.ifverbose)
if len(st2) == 0:
raise ValueError("no waveforms left to process!")
# save raw waveforms in SAC format
if self.isave_raw:
path_to_waveforms = evname_key + "/RAW"
write_stream_sac_raw(stream_raw,
path_to_waveforms,
evname_key,
self.idb,
event,
stations=inventory)
# Taper waveforms (optional; Generally used when data is noisy- example: HutchisonGhosh2016)
# https://docs.obspy.org/master/packages/autogen/obspy.core.trace.Trace.taper.html
# To get the same results as the default taper in SAC, use max_percentage=0.05 and leave type as hann.
# Note: Tapering also happens while resampling (see util_write_cap.py)
if self.taper:
st2.taper(max_percentage=self.taper,
type='hann',
max_length=None,
side='both')
# save processed waveforms in SAC format
# evname_key/RAW_processed = traces after waveform_QA + demean + detrend +
# resample + remove response + filtering +
# resampling + scaling + tapering
# NOTE: The orientation is same as that of extracted waveforms
# Waveforms are rotated to ENZ, in case they are not already orientated,
# in the next step (self.rotateRTZ)
if self.isave_raw_processed:
path_to_waveforms = os.path.join(evname_key, 'RAW_processed')
write_stream_sac(st2, path_to_waveforms, evname_key)
# Rotate to ENZ (save: optional)
#if self.rotateENZ:
st2 = rotate2ENZ(st2, evname_key, self.isave_ENZ, self.icreateNull,
self.ifverbose)
# rotate to UVW and save
if self.rotateUVW:
rotate2UVW(st2, evname_key)
# Rotate to RTZ and save
if self.rotateRTZ:
rotate2RTZ(st2, evname_key, self.ifverbose)
# save CAP weight files
if self.output_cap_weight_file:
write_cap_weights(st2, evname_key, client_name, event,
self.ifverbose)
# save event info
if self.output_event_info:
write_ev_info(event, evname_key)
# Plot spectrograms
if self.ifplot_spectrogram:
plot_spectrogram(st2, evname_key)
# save pole zero file (Needed for MouseTrap)
if self.ifsave_sacpaz:
write_resp(inventory, evname_key)
# save station inventory as XML file
if self.ifsave_stationxml:
xmlfilename = evname_key + "/stations.xml"
try:
inventory.write(xmlfilename,
format="stationxml",
validate=True)
except:
print('Could not create stationxml file')
# Path to the asdf_converter script
if self.ifsave_asdf:
# save RTZ
asdf_filename = evname_key + "/" + evname_key + ".h5"
os.system("../asdf_converters/asdf_converters/sac2asdf.py " +
evname_key + " " + asdf_filename + " observed")
# save NEZ
nez_dir = evname_key + "/ENZ/"
nez_asdf_filename = nez_dir + evname_key + ".h5"
os.system("../asdf_converters/asdf_converters/sac2asdf.py " +
nez_dir + " " + nez_asdf_filename + " observed")
def copy(self):
'''
create of copy of itself
'''
return deepcopy(self)
def reference_time_place(self):
'''
returns an event object with different origin time and location
(i.e. not centered around the earthquake). Stations will be subsetted
based on reference origin time and location
'''
self.ref_time_place = self.ev.copy()
self.ref_time_place.origins[0].latitude = self.rlat
self.ref_time_place.origins[0].longitude = self.rlon
self.ref_time_place.origins[0].time = self.rtime
def get_event_object(self):
'''
update events otime,lat,lon and mag with IRIS (or any other clients) catalog
'''
# get parameters from the cataog
if self.use_catalog == 1:
print("WARNING using event data from the IRIS catalog")
cat = self.client.get_events(
starttime=self.otime - self.sec_before_after_event,
endtime=self.otime + self.sec_before_after_event)
self.ev = cat[0]
# use catalog parameters
self.otime = self.ev.origins[0].time
self.elat = self.ev.origins[0].latitude
self.elon = self.ev.origins[0].longitude
self.edep = self.ev.origins[0].depth
self.emag = self.ev.magnitudes[0].mag
# use parameters from the input file
else:
print("WARNING using event data from user-defined catalog")
#self.ev = Event()
org = Origin()
org.latitude = self.elat
org.longitude = self.elon
org.depth = self.edep
org.time = self.otime
mag = Magnitude()
mag.mag = self.emag
mag.magnitude_type = "Mw"
self.ev.origins.append(org)
self.ev.magnitudes.append(mag)
def get_events_client(self):
'''
get absolute and reference event object
'''
# IRIS
if self.idb == 1:
# import functions to access waveforms
if not self.user and not self.password:
self.client = Client("IRIS")
else:
self.client = Client("IRIS",
user=self.user,
password=self.password)
# will only work for events in the 'IRIS' catalog
# (future: for Alaska events, read the AEC catalog)
# get event object
self.get_event_object()
# use a different reference time and place for station subsetting
if self.rlat is not None:
self.reference_time_place()
# or use reference same as the origin
else:
self.ref_time_place = self.ev
# LLNL
if self.idb == 3:
import llnl_db_client
self.client = llnl_db_client.LLNLDBClient(
"/store/raw/LLNL/UCRL-MI-222502/westernus.wfdisc")
# get event time and event ID
cat = self.client.get_catalog()
mintime_str = "time > %s" % (self.otime -
self.sec_before_after_event)
maxtime_str = "time < %s" % (self.otime +
self.sec_before_after_event)
print(mintime_str + "\n" + maxtime_str)
self.ev = cat.filter(mintime_str, maxtime_str)
if len(self.ev) > 0:
self.ev = self.ev[0]
# Nothing happens here. We can change later
self.ref_time_place = self.ev
print(len(self.ev))
else:
print(
"WARNING. No events in the catalog for the given time period"
)
#sys.exit()
# print client and event info
print(self.ev)
def save_extraction_info(self):
# track git commit
os.system('git log | head -12 > ./' + self.evname + '/' + self.evname +
'_last_2git_commits.txt')
# save filenames in a file for checking
fname = self.evname + '/' + self.evname + '_all_filenames'
fcheck = open(fname, 'w')
os.system('ls -1 ' + self.evname + '/* > ' + fname)
class getwaveform:
def __init__(self):
"""
---------------- copied from old getwaveform.py ----------------------
min_dist - minimum station distance (default = 20)
max_dist - maximum station distance (default =300)
before - time window length before the event time (default= 100)
after - time window length after the event time (default = 300)
network - network codes of potential stations (default=*)
channel - component(s) to get, accepts comma separated (default='BH*')
ifresample_TF - Boolean. Request resample or not. Default = False
resample_freq - sampling frequency to resample waveforms (default 20.0)
ifrotate - Boolean, if true will output sac files rotated to baz
unrotated sac files will also be written
ifCapInp - Boolean, make weight files for CAP
ifEvInfo - Boolean, output 'ev_info.dat' containg event info (True)
ifRemoveResponse - Boolean, will remove response (True)
ifDetrend - Boolean, will remove linear trend from data (True)
ifDemean - Boolean, will insult the data (True)
scale_factor - scale all data by one value (10.0**2)
This usually puts the data in the units required by CAP
From m/s to cm/s
pre_filt - list, corner frequencies of filter to apply before deconv
a good idea when deconvolving (ifRemoveResponse=True)
"""
# DEFAULT SETTINGS (see getwaveform.py)
self.idb = 1 # default: =1-IRIS; =2-AEC; =3-LLNL; =4-Geoscope
self.client = Client()
# event parameters
self.use_catalog = 1 # use an existing catalog (=1) or specify your own event parameters (see iex=9)
self.sec_before_after_event = 10 # time window to search for a target event in a catalog
self.min_dist = 0
self.max_dist = 2000
self.min_az = 0
self.max_az = 360
self.min_lat = None
self.max_lat = None
self.min_lon = None
self.max_lon = None
self.tbefore_sec = 100
self.tafter_sec = 300
# These are used only if self.use_catalog = 0
self.evname = None
self.otime = None
self.elat = None
self.elon = None
self.edep = None
self.emag = None
# Refernce origin (dummy values)
self.rlat = None
self.rlon = None
self.rtime = None
# event objects
self.ev = Event()
self.ref_time_place = Event()
# station parameters
self.network = '*' # all networks
self.station = '*,-PURD,-NV33,-GPO' # all stations except -(these)
self.channel = '*' # all channels
self.overwrite_ddir = 1 # 1 = delete data directory if it already exists
self.icreateNull = 1 # create Null traces so that rotation can work (obsby stream.rotate require 3 traces)
self.phase_window = False # Grab waveforms using phases
self.phases = ["P",
"P"] # Phases to write to sac files or grab data from
self.write_sac_phase = False # put phase information in sac files
self.taupmodel = "ak135"
# Filter parameters
self.ifFilter = False
#------Filter--------------
# for 'bandpass' both f1 and f2 are used
# for 'lowpass' only f2 is used
# for 'highpass' only f1 is used
#
# EXAMPLES
# ifFilter zerophase remove_response ipre_filt
# A. CAP-ready waveforms [DEFAULT]: False NA True 1
# B. plot-ready waveforms, acausal: True True True 2
# C. plot-ready, causal waveforms: True False True 0
# D. possible sensor issues: True False False NA
#
self.filter_type = 'bandpass'
# f1 should consider the requested length of the time series
# f2 should consider the sampling rate for the desired channels
self.f1 = 1 / 1000 # NOTE DOES NOT CHANGE WHEN ipre_filt=2
self.f2 = 20 # NOTE DOES NOT CHANGE WHEN ipre_filt=2
self.zerophase = True # = False (causal), = True (acausal)
# 4 pole filter is more sharper at the edges than 2 pole
self.corners = 4 # Is corner in Obspy same as Pole in SAC?
# Pre-filter parameters
self.ipre_filt = 1 # =0 No pre_filter
# =1 default pre_filter (see getwaveform_iris.py)
# =2 user-defined pre_filter (use this if you are using bandpass filter)
# For tapering down the pre-filter
# Perhaps you want to set ipre_filt = 0 to prevent extra filtering
# pre-filter for deconvolution
# https://ds.iris.edu/files/sac-manual/commands/transfer.html
# Pre-filter will not be applied if remove_response = False
self.f0 = 0.5 * self.f1
self.f3 = 2.0 * self.f2
self.pre_filt = (self.f0, self.f1, self.f2, self.f3
) # applies for ipre_filt = 2 only
# self.pre_filt = (0.005, 0.006, 10.0, 15.0) # BH default
# For CAP
self.resample_TF = True # if False then resample_freq is taken from SAC files
self.resample_freq = 50 # 0 causes errors. Use resample_TF instead
self.scale_factor = 1 # for CAP use 10**2 (to convert m/s to cm/s)
# Pre-processing (manily for CAP)
self.output_cap_weight_file = True # output cap weight files
self.detrend = True # detrend waveforms
self.demean = True # demean waveforms
self.taper = False # this could also be a fraction between 0 and 1 (fraction to be tapered from both sides)
self.output_event_info = True # output event info file
self.outformat = 'VEL' # Intrument response removed waveforms could be saved as 'VEL' 'DISP' 'ACC'
self.ifsave_sacpaz = False # save sac pole zero (needed as input for MouseTrap module)
self.remove_response = True # remove instrument response
self.iplot_response = False # plot response function
self.ifplot_spectrogram = False # plot spectrograms
self.ifsave_stationxml = True # save station xml file (for adjoint tomo)
# Waveforms to be saved
self.rotateRTZ = True # Rotate and save the RTZ components
self.rotateUVW = False # Rotate and save the UVW components
self.isave_raw = False # save raw waveforms
self.isave_raw_processed = False # save processed waveforms just before rotation to ENZ
#self.rotateENZ = True # rotate extracted waveforms to ENZ
self.isave_ENZ = True # save ENZ
# username and password for embargoed IRIS data
# Register here: http://ds.iris.edu/ds/nodes/dmc/forms/restricted-data-registration/
self.user = None
self.password = None
# To output lots of processing info
self.ifverbose = False
def run_get_waveform(self):
"""
Get SAC waveforms for an event
basic usage:
run_get_waveform(event)
c - client
event - obspy Event object
ref_time_place - reference time and place (other than origin time and place - for station subsetting)
"""
c = self.client
event = self.ev
ref_time_place = self.ref_time_place
evtime = event.origins[0].time
reftime = ref_time_place.origins[0].time
if (self.idb == 1) or (self.idb == 5):
# 2019-03-12 I added idb #5 here and not in a separate block since
# the station retrieval is common to #1 and #5 and I rather not repeat the same block.
# note that CH stations may not provide all sensor metadata (required when using idb #1) and this crashes with later requests in the code
# another option is to add sta retrieval after #3 below but unsure if this creates conflicts.
# TODO use 'is' instead of 'in' to avoid wrong calls (eg FZ or GFZ)
if ("BK" in self.network) or ("BG" in self.network) or (
"BP" in self.network):
# NOTE BK network doesn't return data when using the IRIS client.
# this option switches to NCEDC if BK is in networkname.
client_name = "NCEDC"
elif "CH" in self.network:
client_name = "ETH"
elif "GFZ" in self.network:
client_name = "GFZ"
elif "Z3" in self.network: # 2019-07-31 for AlpArray
client_name = "ORFEUS"
#client_name = "GFZ"
#client_name = "ETH"
elif "IV" in self.network:
client_name = "INGV"
elif "DE" in self.network:
client_name = "EMSC" # or BGR GFZ LMU ...
else:
client_name = "IRIS" # default
c = Client(client_name)
#-----------------------------------------------------------
# Check if requesting AlpArray data
# the following are ongoing TESTS (alparray data)
#c = Client(client_name, password='******')
#c = Client(client_name, password='******')
#c = Client(client_name, password='******')
#c = RoutingClient('iris-federator') # 2019-06-18 This downloads a list of station data, but no waveforms. Also it requires removing radius-related infor (below) and doesn't have a field for eidatoken.
#c = RoutingClient('iris-federator', credentials={'EIDA_TOKEN': './eidatoken'}, debug=True)
#c = RoutingClient('iris-federator', credentials={'EIDA_TOKEN': './eidatoken'})
#c = RoutingClient('eida-routing', credentials={'EIDA_TOKEN': '/home/calvizur/.eidatoken'}, debug=True) # 2019-07-31 TEST?
## try again AlpArray, arclink
# NB doesnt work. 'response' level is not accepted which means I would have to prepare data manually
# option: use script similar to Veronica's
#c = Client(user='******') # test, AlpArray Client
#c.max_status_requests = 2000
if "Z3" in self.network:
# 2019-07-31 WORKS! (bondo event)
# replace client if requesting AlpArray data
c = RoutingClient(
'iris-federator',
credentials={'EIDA_TOKEN': '/home/calvizur/.eidatoken'},
debug=True)
#-----------------------------------------------------------
print(c)
print("\n** WARNING ** Preparing request for client", client_name)
print("Download stations...")
if "Z3" in self.network:
# *** alparray only ***
stations = c.get_stations(network=self.network,
station=self.station,
channel=self.channel,
starttime=reftime - self.tbefore_sec,
endtime=reftime + self.tafter_sec,
level="response")
else:
# default
stations = c.get_stations(
network=self.network,
station=self.station,
channel=self.channel,
starttime=reftime - self.tbefore_sec,
endtime=reftime + self.tafter_sec,
#----------------------------------------------------------
# NOTE RoutingClient (AlpArray) does not have the following
minlatitude=self.min_lat,
maxlatitude=self.max_lat,
minlongitude=self.min_lon,
maxlongitude=self.max_lon,
longitude=self.elon, # required for maxradius
latitude=self.elat, # required for maxradius
maxradius=18, # note 18 deg ~ 2000 km
#----------------------------------------------------------
level="response")
inventory = stations # so that llnl and iris scripts can be combined
if self.ifverbose:
print("Printing stations")
print(stations)
print("Done Printing stations...")
sta_limit_distance(ref_time_place,
stations,
min_dist=self.min_dist,
max_dist=self.max_dist,
min_az=self.min_az,
max_az=self.max_az,
ifverbose=self.ifverbose)
#print("Printing stations NEW")
#print(stations)
#print("Done Printing stations...")
#stations.plotprojection="local")
# Find P and S arrival times
t1s = []
t2s = []
phases = self.phases
if self.phase_window:
#model = TauPyModel(model=taupmodel)
model = TauPyModel(model=self.taupmodel)
for net in stations:
for sta in net:
dist, az, baz = obspy.geodetics.gps2dist_azimuth(
event.origins[0].latitude,
event.origins[0].longitude, sta.latitude,
sta.longitude)
dist_deg = kilometer2degrees(dist / 1000, radius=6371)
Phase1arrivals = model.get_travel_times(
source_depth_in_km=event.origins[0].depth / 1000,
distance_in_degree=dist_deg,
phase_list=[phases[0]])
if len(Phase1arrivals) == 0:
if phases[0] == "P":
phases[0] = "p"
elif phases[0] == "p":
phases[0] = "P"
elif phases[0] == "S":
phases[0] = "s"
elif phases[0] == "s":
phases[0] = "S"
Phase1arrivals = model.get_travel_times(
source_depth_in_km=event.origins[0].depth /
1000,
distance_in_degree=dist_deg,
phase_list=[phases[0]])
Phase2arrivals = model.get_travel_times(
source_depth_in_km=event.origins[0].depth / 1000,
distance_in_degree=dist_deg,
phase_list=[phases[1]])
if len(Phase2arrivals) == 0:
if phases[1] == "P":
phases[1] = "p"
elif phases[1] == "p":
phases[1] = "P"
elif phases[1] == "S":
phases[1] = "s"
elif phases[1] == "s":
phases[1] = "S"
Phase2arrivals = model.get_travel_times(
source_depth_in_km=event.origins[0].depth /
1000,
distance_in_degree=dist_deg,
phase_list=[phases[1]])
#somearr = model.get_travel_times(source_depth_in_km=event.origins[0].depth/1000,distance_in_degree=dist_deg)
#print("Print arrivals")
#print(somearr)
try:
if Phase2arrivals[0].time < Phase1arrivals[0].time:
# You are assuming that the first index is the first arrival. Check this later.
t1s.append(event.origins[0].time +
Phase2arrivals[0].time -
self.tbefore_sec)
t2s.append(event.origins[0].time +
Phase1arrivals[0].time +
self.tafter_sec)
else:
t1s.append(event.origins[0].time +
Phase1arrivals[0].time -
self.tbefore_sec)
t2s.append(event.origins[0].time +
Phase2arrivals[0].time +
self.tafter_sec)
except:
t1s.append(reftime - self.tbefore_sec)
t2s.append(reftime + self.tafter_sec)
else:
t1s = [reftime - self.tbefore_sec]
t2s = [reftime + self.tafter_sec]
print("Downloading waveforms...")
# this needs to change
bulk_list = make_bulk_list_from_stalist(stations,
t1s,
t2s,
channel=self.channel)
stream_raw = c.get_waveforms_bulk(bulk_list)
# save ev_info object
pickle.dump(
self,
open(self.evname + '/' + self.evname + '_ev_info.obj', 'wb'))
elif self.idb == 3:
client_name = "LLNL"
print("Preparing request for LLNL ...")
# Get event an inventory from the LLNL DB.
event_number = int(event.event_descriptions[0].text)
# event = llnl_db_client.get_obspy_event(event)
inventory = c.get_inventory()
nsta_llnl = len(inventory.get_contents()["stations"])
print("--> Total stations in LLNL DB: %i" % nsta_llnl)
sta_limit_distance(event,
inventory,
min_dist=self.min_dist,
max_dist=self.max_dist,
min_az=self.min_az,
max_az=self.max_az)
print("--> Stations after filtering for distance: %i" %
(len(inventory.get_contents()["stations"])))
stations = set([sta.code for net in inventory for sta in net])
_st = c.get_waveforms_for_event(event_number)
stream_raw = obspy.Stream()
for tr in _st:
if tr.stats.station in stations:
stream_raw.append(tr)
#-----------------------------------------------------------
elif self.idb == 99:
# IRIS federator to figure out who has waveform data for that
# particular query and subsequently call the individual data
# centers to actually get the data.
# https://docs.obspy.org/packages/obspy.clients.fdsn.html
print("Preparing request through iris-federator")
c = RoutingClient('iris-federator', debug=True)
stations = c.get_stations(
#network = self.network,
#station = self.station,
channel=self.channel,
starttime=reftime - self.tbefore_sec,
endtime=reftime + self.tafter_sec,
#minlatitude = self.min_lat,
#maxlatitude = self.max_lat,
#minlongitude = self.min_lon,
#maxlongitude = self.max_lon,
longitude=self.elon, # required for maxradius
latitude=self.elat, # required for maxradius
#maxradius = 18, # note 18 deg ~ 2000 km
maxradius=18, # note 18 deg ~ 2000 km
level="response")
print("Downloading waveforms...")
# this needs to change
bulk_list = make_bulk_list_from_stalist(stations,
t1s,
t2s,
channel=self.channel)
stream_raw = c.get_waveforms_bulk(bulk_list)
# save ev_info object
pickle.dump(
self,
open(self.evname + '/' + self.evname + '_ev_info.obj', 'wb'))
#-----------------------------------------------------------
# set reftime
stream = obspy.Stream()
stream = set_reftime(stream_raw, evtime)
print("--> Adding SAC metadata...")
if self.ifverbose: print(inventory)
st2 = add_sac_metadata(stream,
idb=self.idb,
ev=event,
stalist=inventory,
taup_model=self.taupmodel,
phases=phases,
phase_write=self.write_sac_phase)
# Do some waveform QA
do_waveform_QA(st2, client_name, event, evtime, self.tbefore_sec,
self.tafter_sec)
if self.demean:
st2.detrend('demean')
if self.detrend:
st2.detrend('linear')
if self.remove_response:
resp_plot_remove(st2, self.ipre_filt, self.pre_filt,
self.iplot_response, self.scale_factor, stations,
self.outformat, self.ifverbose)
else:
# output RAW waveforms
decon = False
print("WARNING -- NOT correcting for instrument response")
if self.ifFilter:
prefilter(st2, self.f1, self.f2, self.zerophase, self.corners,
self.filter_type)
if self.scale_factor > 0:
amp_rescale(st2, self.scale_factor)
if self.idb == 3:
amp_rescale_llnl(st2, self.scale_factor)
# Set the sac header KEVNM with event name
# This applies to the events from the LLNL database
# NOTE this command is needed at the time of writing files, so it has to
# be set early
st2, evname_key = rename_if_LLNL_event(st2, evtime)
self.evname = evname_key
# save station plot
# Note: Plotted are stations in the inventory and NOT the ones with the traces
# It could be possible that there might not be waveforms for some of these stations.
try:
fig = inventory.plot(projection="local",
resolution="i",
label=False,
show=False)
Catalog([self.ev]).plot(fig=fig,
outfile=self.evname + '/station_map.pdf')
except:
print("There is a problem with creating the station map!")
# Get list of unique stations + locaiton (example: 'KDAK.00')
stalist = []
for tr in stream.traces:
if self.ifverbose: print(tr)
station_key = "%s.%s.%s.%s" % (tr.stats.network, tr.stats.station,
tr.stats.location,
tr.stats.channel[:-1])
stalist.append(station_key)
# Crazy way of getting a unique list of stations
stalist = list(set(stalist))
# Resample
if self.resample_TF == True:
# NOTE !!! tell the user if BOTH commands are disabled NOTE !!!
if (client_name == "IRIS"):
resample(st2, freq=self.resample_freq)
elif (client_name == "LLNL"):
resample_cut(st2, self.resample_freq, evtime, self.tbefore_sec,
self.tafter_sec)
else:
print(
"WARNING. Will not resample. Using original rate from the data"
)
# match start and end points for all traces
st2 = trim_maxstart_minend(stalist, st2, client_name, event, evtime,
self.resample_TF, self.resample_freq,
self.tbefore_sec, self.tafter_sec,
self.ifverbose)
if len(st2) == 0:
raise ValueError("no waveforms left to process!")
# save raw waveforms in SAC format
if self.isave_raw:
path_to_waveforms = evname_key + "/RAW"
write_stream_sac_raw(stream_raw,
path_to_waveforms,
evname_key,
self.idb,
event,
stations=inventory)
# Taper waveforms (optional; Generally used when data is noisy- example: HutchisonGhosh2016)
# https://docs.obspy.org/master/packages/autogen/obspy.core.trace.Trace.taper.html
# To get the same results as the default taper in SAC, use max_percentage=0.05 and leave type as hann.
if self.taper:
st2.taper(max_percentage=self.taper,
type='hann',
max_length=None,
side='both')
# save processed waveforms in SAC format
# evname_key/RAW_processed = traces after waveform_QA + demean + detrend +
# resample + remove response + filtering +
# resampling + scaling + tapering
# NOTE: The orientation is same as that of extracted waveforms
# Waveforms are rotated to ENZ, in case they are not already orientated,
# in the next step (self.rotateRTZ)
if self.isave_raw_processed:
path_to_waveforms = os.path.join(evname_key, 'RAW_processed')
write_stream_sac(st2, path_to_waveforms, evname_key)
# Rotate to ENZ (save: optional)
#if self.rotateENZ:
st2 = rotate2ENZ(st2, evname_key, self.isave_ENZ, self.icreateNull,
self.ifverbose)
# rotate to UVW and save
if self.rotateUVW:
rotate2UVW(st2, evname_key)
# Rotate to RTZ and save
if self.rotateRTZ:
rotate2RTZ(st2, evname_key, self.ifverbose)
# save CAP weight files
if self.output_cap_weight_file:
write_cap_weights(st2, evname_key, client_name, event,
self.ifverbose)
# save event info
if self.output_event_info:
write_ev_info(event, evname_key)
# Plot spectrograms
if self.ifplot_spectrogram:
plot_spectrogram(st2, evname_key)
# save pole zero file (Needed for MouseTrap)
if self.ifsave_sacpaz:
write_resp(inventory, evname_key)
# save station inventory as XML file
if self.ifsave_stationxml:
xmlfilename = evname_key + "/stations.xml"
try:
inventory.write(xmlfilename,
format="stationxml",
validate=True)
except:
print('Could not create stationxml file')
print("### DEBUG ###")
def copy(self):
'''
create of copy of itself
'''
return deepcopy(self)
def reference_time_place(self):
'''
returns an event object with different origin time and location
(i.e. not centered around the earthquake). Stations will be subsetted
based on reference origin time and location
'''
self.ref_time_place = self.ev.copy()
self.ref_time_place.origins[0].latitude = self.rlat
self.ref_time_place.origins[0].longitude = self.rlon
self.ref_time_place.origins[0].time = self.rtime
def get_event_object(self):
'''
update events otime,lat,lon and mag with IRIS (or any other clients) catalog
'''
# get parameters from the cataog
if self.use_catalog == 1:
print("WARNING using event data from the IRIS catalog")
cat = self.client.get_events(
starttime=self.otime - self.sec_before_after_event,
endtime=self.otime + self.sec_before_after_event)
self.ev = cat[0]
# use catalog parameters
self.otime = self.ev.origins[0].time
self.elat = self.ev.origins[0].latitude
self.elon = self.ev.origins[0].longitude
self.edep = self.ev.origins[0].depth
self.emag = self.ev.magnitudes[0].mag
# use parameters from the input file
else:
print("WARNING using event data from user-defined catalog")
#self.ev = Event()
org = Origin()
org.latitude = self.elat
org.longitude = self.elon
org.depth = self.edep
org.time = self.otime
mag = Magnitude()
mag.mag = self.emag
mag.magnitude_type = "Mw"
self.ev.origins.append(org)
self.ev.magnitudes.append(mag)
def get_events_client(self):
'''
get absolute and reference event object
'''
# IRIS
if (self.idb == 1) or (self.idb == 5):
# import functions to access waveforms
if not self.user and not self.password:
self.client = Client("IRIS")
else:
self.client = Client("IRIS",
user=self.user,
password=self.password)
# will only work for events in the 'IRIS' catalog
# (future: for Alaska events, read the AEC catalog)
# get event object
self.get_event_object()
# use a different reference time and place for station subsetting
if self.rlat is not None:
self.reference_time_place()
# or use reference same as the origin
else:
self.ref_time_place = self.ev
# LLNL
if self.idb == 3:
import llnl_db_client
self.client = llnl_db_client.LLNLDBClient(
"/store/raw/LLNL/UCRL-MI-222502/westernus.wfdisc")
# get event time and event ID
cat = self.client.get_catalog()
mintime_str = "time > %s" % (self.otime -
self.sec_before_after_event)
maxtime_str = "time < %s" % (self.otime +
self.sec_before_after_event)
print(mintime_str + "\n" + maxtime_str)
self.ev = cat.filter(mintime_str, maxtime_str)
if len(self.ev) > 0:
self.ev = self.ev[0]
# Nothing happens here. We can change later
self.ref_time_place = self.ev
print(len(self.ev))
else:
print(
"WARNING. No events in the catalog for the given time period"
)
#sys.exit()
# print client and event info
print(self.ev)
def save_extraction_info(self):
# track git commit
os.system('git log | head -12 > ./' + self.evname + '/' + self.evname +
'_last_2git_commits.txt')
# save filenames in a file for checking
fname = self.evname + '/' + self.evname + '_all_filenames'
fcheck = open(fname, 'w')
os.system('ls -1 ' + self.evname + '/* > ' + fname)
