Category: Uncategorized

New calibrated cluster posted to GCCEL:

The Dezhgan cluster is named for the village of Dezhgan in Hormozgan
Province in southern Iran. The cluster includes seismicity at Qeshm
Island in the Strait of Hormuz. The cluster contains many earthquakes
with magnitudes in the range 5.0-5.9, and three larger events, 6.0 mb on
September 10, 2008 (at Qeshm Island) and 6.0 and 6.3 mb on July 1, 2022.
In order to obtain adequate azimuthal coverage for direct calibration it
was necessary to extend the distance limit used for estimating the
hypocentroid to 1.2°, to pick up readings from seismograph stations to
the south, across the Strait of Hormuz, in the UAE. In the distance
range 1.0-1.2° Pn and Sn arrivals are often found following close to the
arrival times of direct crustal phases Pg and Sg. This required careful
inspection of residuals to avoid biasing depth estimates. The data set
includes arrival times from temporary network stations that were
operated on Qeshm Island by Yamini Fard and colleagues from the IIEES,
as well as observations of S-P reported from strong motion stations very
near epicenters on Qeshm Island by Reza Ghods from IASBS. These data
help constrain focal depths and also helped guard against location bias
from unmodeled lateral heterogeneity, which is likely rather strong in
this area.

Latest upload for GCCEL:

The Kurikoma cluster is named for Mt. Kurikoma, a stratovolcano in
northern Honshu, Japan. Much of the seismicity in the region is thought
to be related to magmatic activity. The cluster includes 5 earthquakes
of magnitude 6.0-7.0, and most of the smaller events appear to have
occurred as aftershocks or swarms related to these larger events.
Station coverage for location calibration and depth control is
excellent. All events have depth control from near-source or
local-distance observations. All events are recorded at teleseismic
distances.

Version 12 of mloc is now being used for new calibrated clusters. Version 11 was meant to be the release of this greatly updated code but it ended up being mainly used for development and testing. The changes in v12 vastly improve memory management and execution time, making it possible to analyze much larger clusters than before. Clusters of 500 events or more are practical on a Mac Mini with M2 chip and only 8 GB of RAM. This version of mloc will be posted soon, when I’ve finished updating the User’s Manual, but if you’re anxious to get started with it, get in touch with me directly.

Since the last update on new clusters in August I’ve posted eight new ones, in China, Iran, Tajikistan, Russia, and Uzbekistan.

Since I last posted about new calibrated clusters for GCCEL in April, I have uploaded 15 new ones, bringing the total number of calibrated clusters to 359. The regions I’ve been working in include the western U.S., Japan and China. You can see them at the “MLOC data for GCCEL” page .

mloc version 11 is a major update to the software that has been in progress since March of this year. I have done a considerable amount of testing and I’ve been using it for recent calibrated clusters uploaded to the GCCEL database. I’m no longer finding (m)any bugs with the way I typically use it, but several other people are using mloc for rather different studies however and they may well uncover a few more bugs that need attention. I expect to declare this version ready for release in the next month or two.

The main advantage of version 11 is the management of memory, which is much more efficient now. As a result the limitation that previously existed on the number of events that can be in a cluster has been lifted. I have made runs with clusters of over 500 events and one of the current users is running clusters approaching 1000 events. Of course, these runs still take a lot of time, and careful analysis of such large clusters is very time-consuming. The other main advantage of version 11 is not evident to the user, but very useful for the developer: placing all subroutines in Fortran modules and getting rid of large named common blocks in include files to handle global variables.

One known issue that I will try to resolve before release of v11 is the annoying discovery that the scripts it produces for plots are in some cases incompatible with GMT v6.5. For the time being, GMT 6.4 or below must be used.

As I’ve added more features to mloc over the years its memory requirements have grown to the point where we are now encountering problems on some systems when trying to run larger clusters. I won’t try to describe the details here but I’m happy to discuss it with you if you get in touch. In the old days it was referred to as “stack overflow” but modern chips and OSes have much more complex memory management systems and I don’t really know what is specifically failing. The original software structure, written in the 1980s is based mainly on Fortran 77 and relies heavily on include files and named common blocks that hit the stack very hard. The good news is that Fortran 90 (which is the standard used for mloc) has some features that should make it possible to substantially reduce the memory requirements for a given problem size, and thus make it feasible to run clusters with more events and/or arrival time data. I am now working on the next generation of the code that takes advantage of some of those features, in particular, modules and allocatable arrays. The reorganization should also make it easier to maintain and develop the code further. So for some period of time I will not be working on new calibrated clusters (much). I will post here when I have a version ready for testing.

The Salem cluster in northwest Oregon has been posted in the “mloc data for GCCEL” section.

A calibrated cluster in the Seattle, Washington area has been posted to GCCEL.

The distribution package for mloc v10.6.6, release date April 18, 2024, is now available on the Distribution page.