[webservices] fdsnws-event service - New Release

Philip Crotwell crotwell at seis.sc.edu
Fri Oct 4 12:51:06 PDT 2013


Hi

Did you all come up with a use case for requesting "all magnitudes, but
only of a single type"? I just don't see how someone wanting all 'mb'
magnitudes would be mad if they also got some "Mw" and "ML" magnitudes? Not
like the extra magnitudes contribute that much overhead, and as you say
they can be filtered out or ignored on the client side.

I do see a use case for filtering based on a magType, but still wanting to
see ALL magnitudes. Searching for "mb>X" is common, but with mb saturating
for large earthquakes, many seismologists would prefer to use the mw if it
exists for a given event, but falling back on the mb if not. To bo honest,
the main reason for me may be that the parameter name
"includeallmagnitudes" is confusing if it doesn't actually "include all
magnitudes".

I'll cast my vote for includeallmagnitudes to really "include all
magnitudes," regardless of whether or not magType is specified.

My $0.02
Philip




On Tue, Oct 1, 2013 at 11:04 AM, Chad Trabant <chad at iris.washington.edu>wrote:

>
> Hi Philip,
>
> That is correct.  We struggled with this one quite a bit and there is
> really no right answer.  The issue is that magtype can mean either what is
> searched on (as specified by the spec) but also specify what is returned.
>  If we change the logic to what you propose there will be no way to request
> only all magnitudes of a given type for events.  There are simply not
> enough parameters to separate these use cases unfortunately.  Understand
> that it remains quite easy to get all magnitudes without specifying any
> magtype for the server and doing more advanced selection downstream if
> needed.
>
> It would be useful to hear from other users regarding what would be
> useful.  The 'magtype' parameter, per the specification, specifies which
> magnitudes should be used for matching the criteria (min and max).  Should
> the 'magtype' parameter also be interpreted as a selection of what to
> return?
>
> thanks
> Chad
>
> On Sep 17, 2013, at 2:43 PM, Philip Crotwell <crotwell at seis.sc.edu> wrote:
>
>
> Hi
>
> So, with these changes and if I am reading the docs correctly, if I
> specify magType, there is no way to get all magnitudes because
> includeallmagnitudes only includes all the magnitudes that also match the
> magtype? Meaning if I query "magtype=MB&includeallmagnitudes=true" then I
> might get 5 magnitudes, but they will all be of type MB?
>
> Am I misunderstanding the docs, because this doesn't seem to match what
> you proposed a few weeks ago. I think there is definite value in being able
> to get ALL magnitudes when searching on a particular type. Being able to
> get all magnitudes of a given type actually seems less useful to me. In
> most cases I am quit happy to get the "best" magnitude of a given type, but
> there are cases where I want to see the best Mb and the Mw and ML for the
> events I am interested in.
>
> Can you clarify?
> Philip
>
>
>
> On Tue, Sep 17, 2013 at 4:26 PM, Yazan Suleiman <yazan at iris.washington.edu
> > wrote:
>
>> Hello WS users,
>>
>> The DMC has updated the fdsnws-event service to version 1.0.6.  Changes
>> include:
>>
>> * Fixed bug where updatedafter operated on the wrong timezone
>> * Updated behavior of magnitudes for combinations of includeallmagnitudes
>> and magtype. (Details<http://service.iris.edu/fdsnws/event/docs/1/help/#magnitudebehavior>
>> )
>>
>> This release should be backwards compatible with all request tools.
>>
>> regards,
>> IRIS DMC web services team
>>
>> _______________________________________________
>> webservices mailing list
>> webservices at iris.washington.edu
>> http://www.iris.washington.edu/mailman/listinfo/webservices
>>
>>
> _______________________________________________
> webservices mailing list
> webservices at iris.washington.edu
> http://www.iris.washington.edu/mailman/listinfo/webservices
>
>
>
> _______________________________________________
> webservices mailing list
> webservices at iris.washington.edu
> http://www.iris.washington.edu/mailman/listinfo/webservices
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.iris.washington.edu/pipermail/webservices/attachments/20131004/6b126172/attachment-0001.htm>


More information about the webservices mailing list