search

PeterMakus / SeisMIC

SeisMIC is a python software suite to monitor velocity changes using ambient seismic noise.
https://petermakus.github.io/SeisMIC/
European Union Public License 1.2
53 stars 12 forks source link

Possible issue with computing velocity changes with dev versiom #72

Closed imme-w closed 3 months ago

imme-w commented 3 months ago

Hi Peter! I have succesfully installed the dev branch of seisMIC and used it to compute cross-correlations between different locations codes and it works! (see below) Thanks a lot for adding this feature :). image I run into problems with computing the velocity changes with monitor.py. The issue is as follows, if I have my network parameter in params.yaml like this: 

net:
    # list of stations used in the project
    # type: list of strings or string, wildcards allowed
    network : ['MT', 'MT', 'MT']
    station : ['GAL', 'RUI', 'THE']
    component : 'Z'

monitor.py does not find the correlation files, as the output says this:

2024-07-09 13:46:36,845 - INFO - Found correlation data for the following station and network combinations []

0it [00:00, ?it/s]
0it [00:00, ?it/s]

I have discoved that changing the network parameters like this solves the issue:

net:
    # list of stations used in the project
    # type: list of strings or string, wildcards allowed
    network : 'MT'
    station : ['GAL', 'RUI', 'THE']
    component : 'Z'

because then the output of monitor.py (see below) gives the right combinations and velocity changes appear in the right directory.

2024-07-09 13:49:47,411 - INFO - Found correlation data for the following station and network combinations ['MT-MT.GAL-RUI', 'MT-MT.GAL-RUI', 'MT-MT.GAL-RUI', 'MT-MT.GAL-RUI', 'MT-MT.GAL-RUI', 'MT-MT.GAL-RUI', 'MT-MT.GAL-THE', 'MT-MT.GAL-THE', 'MT-MT.GAL-THE', 'MT-MT.GAL-THE', 'MT-MT.GAL-THE', 'MT-MT.GAL-THE', 'MT-MT.RUI-THE', 'MT-MT.RUI-THE', 'MT-MT.RUI-THE', 'MT-MT.RUI-THE']

But at some point I will be dealing with different network codes so the network parameter will have to be a list. I also discoverd if I have the net parameter as below, so a list with two different networks: 

#### parameters that are network specific
net:
    # list of stations used in the project
    # type: list of strings or string, wildcards allowed
    network : ['MT', 'MT', 'MT', 'FR']
    station : ['GAL', 'RUI', 'THE', 'OSG1']
    component : 'Z'

then the output of monitor.py looks like shown below, it appears to only compute cross-correlations between station combinations with a different network code.

2024-07-09 13:33:54,990 - INFO - Found correlation data for the following station and network combinations ['FR-MT.OGS1-GAL', 'FR-MT.OGS1-GAL', 'FR-MT.OGS1-GAL', 'FR-MT.OGS1-RUI', 'FR-MT.OGS1-RUI', 'FR-MT.OGS1-THE', 'FR-MT.OGS1-THE']

So my question is, am I doing something wrong? Thank you for thinking with me!

PeterMakus commented 3 months ago

Hi Imme, thanks for reporting the issue. I am looking into it!

PeterMakus commented 3 months ago

Hi Imme, Could you update your SeisMIC version to the one I released to the dev branch a minute ago?

Pretty sure that I fixed the issue.

A small side note: If you want to correlate from all available network codes, you could use a wildcard ('*') in the network parameters.

Any ways, let me know if this fixed your problem!

imme-w commented 3 months ago

Hi Peter! Thank you very much, I updated my SeisMIC version and indeed it works now! Also thanks for the tip about the wildcard, I had not thought about that yet.