ObsPy Tutorial
Handling Event Metadata
In [1]:
%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use('ggplot')
plt.rcParams['figure.figsize'] = 12, 8
- for event metadata, the de-facto standard is QuakeML (an xml document structure)
- QuakeML files can be read using
read_events()
In [2]:
from obspy import read_events
catalog = read_events("./data/event_tohoku_with_big_aftershocks.xml")
print(catalog)
read_events()function returns aCatalogobject, which is a collection ofEventobjects.
In [3]:
print(type(catalog))
print(type(catalog[0]))
In [4]:
event = catalog[0]
print(event)
- Event objects are again collections of other resources.
- the nested ObsPy Event class structure (Catalog/Event/Origin/Magnitude/FocalMechanism/...) is closely modelled after QuakeML
In [5]:
print(type(event.origins))
print(type(event.origins[0]))
print(event.origins[0])
In [6]:
print(type(event.magnitudes))
print(type(event.magnitudes[0]))
print(event.magnitudes[0])
In [7]:
# try event.<Tab> to get an idea what "children" elements event has
- The Catalog object contains some convenience methods to make working with events easier.
- for example, the included events can be filtered with various keys.
In [8]:
largest_magnitude_events = catalog.filter("magnitude >= 7.8")
print(largest_magnitude_events)
- There is a basic preview plot using the matplotlib basemap module.
In [9]:
catalog.plot(projection="local");
- a (modified) Catalog can be output to file (currently there is write support for QuakeML only)
In [10]:
largest_magnitude_events.write("/tmp/large_events.xml", format="QUAKEML")
!ls -l /tmp/large_events.xml
- the event type classes can be used to build up Events/Catalogs/Picks/.. from scratch in custom processing work flows and to share them with other researchers in the de facto standard format QuakeML
In [11]:
from obspy import UTCDateTime
from obspy.core.event import Catalog, Event, Origin, Magnitude
from obspy.geodetics import FlinnEngdahl
cat = Catalog()
cat.description = "Just a fictitious toy example catalog built from scratch"
e = Event()
e.event_type = "not existing"
o = Origin()
o.time = UTCDateTime(2014, 2, 23, 18, 0, 0)
o.latitude = 47.6
o.longitude = 12.0
o.depth = 10000
o.depth_type = "operator assigned"
o.evaluation_mode = "manual"
o.evaluation_status = "preliminary"
o.region = FlinnEngdahl().get_region(o.longitude, o.latitude)
m = Magnitude()
m.mag = 7.2
m.magnitude_type = "Mw"
m2 = Magnitude()
m2.mag = 7.4
m2.magnitude_type = "Ms"
# also included could be: custom picks, amplitude measurements, station magnitudes,
# focal mechanisms, moment tensors, ...
# make associations, put everything together
cat.append(e)
e.origins = [o]
e.magnitudes = [m, m2]
m.origin_id = o.resource_id
m2.origin_id = o.resource_id
print(cat)
cat.write("/tmp/my_custom_events.xml", format="QUAKEML")
!cat /tmp/my_custom_events.xml