Re: proposal new blockettes

[Rick Benson]

Hi, Reinoud-
I think what would help is a recommendation for what the blockette
number is that they propose (are they talking data record blockette
so we'd have a blockette 300 or 3000 type of addition, or is this
in the station headers. A specific proposal for what to number these.
You mention there was an attached file, but unless you are referring to
the inline text, we did not receive an "attachment" per se.

Thanks
Rick

On Apr 21, 2010, at 12:22 PM, Sleeman, Reinoud (KNMI) wrote:

> Dear WG-II members,
>
> I received the following request from JAMSTEC for 2 new blockettes within SEED,
> including a detailed description (see attached file):
>
> ______________________________________________________________________
>
> One is ``Delay time Information Blockette" which describes breakdown of total
> delay time. The other is ``Quartz Pressure Blockette" which includes 14 coefficients
> to convert count value of quartz-type pressure gauge data to physical value.
>
>
> 1. ``Delay time Information Blockette"
> We are going to install some seismometers and quartz-type pressure gauges in seafloor this year.
> The clock time of seafloor data stamped at land base station is different from the time in seafloor
> observation because of the delay of time from sensor to seafloor station, node point of optic submarine
> cable, and the land base station.
> This time difference is corrected in the land base station, and the time correction is written in field
> 16 (100 micro-sec unit) in Fixed Section of Data Header.
> We here want to describe the breakdown of total delay time in SEED and to utilize the system check of
> data transfer and the acquisition of more precise (more than 100 micro-sec precision) time at sensor in
> seafloor.
> We want to define the new blockette on delay time for seafloor observation.
>
> 2. ``Quartz Pressure Blockette"
> We observe water pressure using quartz-type pressure gauge (Paroscientific, Inc.).
> This gauge is monitoring not only water pressure but also temperature in frequency domain for high-
> accuracy measurement.
> Data user needs to use conversion equations and 14 conversion coefficients if they analyze the data with
> physical value such as Pascal unit.
> We want to define the new blockette on the coefficient for quartz-type pressure gauge.
>
> _____________________________________________________________________________
>
>
> I'm not an expert in this type of measurements but the proposed blockettes seems OK for me
> within the SEED format. Please give me your opinion, if any. If there are no objections within 3
> weeks I will approve this proposal.
>
> Best regards
> Reinoud
>
>
>
> _______________________________________________
> fdsn-wg2-data mailing list
> fdsn-w...@iris.washington.edu
> http://www.iris.washington.edu/mailman/listinfo/fdsn-wg2-data

================
Rick Benson
Director of Operations
IRIS DMC
(206)547-0393 ext. 119(office)
ri...@iris.washington.edu

[Reinoud Sleeman]

[Reinoud Sleeman]

Thanks, Rick - please find the attachment here.

R

[Chad Trabant]

Dear WG-II,

Regarding the two proposed data blockettes:

1) The "Delay time Information Blockette" seems reasonable and we would suggest that it be accompanied by a paragraph describing the intended use, such as is done for blockette 405. We are assuming that is this is a bookkeeping blockette that documents components of the total time correction in field 16 of the fixed section data header, please let us know if we misunderstood the intention. The descriptive paragraph should clearly state that data parsers do not need to perform any further operation when encountering this blockette to determine sample times. A less important characteristic is the usage of "10001" as a blockette type, we would recommended "2001".

2) We feel the proposal for a "Quartz Pressure Blockette" is too specific being limited to a particular instrument type, a general solution would be preferred.

Also it is not clear if the coefficients are changing continuously and represent the actual data stream or are they fixed for a given set of input values. We ask the authors to clarify.

In the case where the coefficients represent the actual data an alternative solution might be the use of blockette 2000 which can contain nearly any data.

In the case where the coefficients are fixed for a given set of input samples the question arises for why this would be a data blockette instead of written as a station control header blockette. Blockette 62 could potentially be extended by defining a new response type (equation), similar to the entry for the MacLaurirn pressure sensors, and create a new field 7.

regards,
Chad Trabant
Tim Ahern
Rick Benson

On Apr 21, 2010, at 12:22 PM, Sleeman, Reinoud (KNMI) wrote:

[Seiji Tsuboi]

Dear members of WG-II

I am sorry for late replies to the comments. We try to clarify our proposal
in the following explanation. Please let us know if we need further
explanation.

Thank you very much.
Seiji Tsuboi

1) This blockette documents components of the total time correction.
As you pointed out, data users do not need to perform any further operation
when encountering this blockette to determine sample times. This blockette
also enables us to derive more precise time information, which depends on
the frequency corresponding to the time unit (field 8,9,10), at each
data-transfer stage before applying time correction. We modified the
explanation of this blockette and temporarily named blockette 2001. Please
see the attached file.

2) To clalify the instrument what we are suing, we would explain some
details.
The quartz pressure gauge we install outputs periods of temperature and
pressure.
Data users need to use conversion equations and 14 conversion coefficients

if they analyze the data with physical value such as Pascal unit.

The conversion equations are as follows:

Temperature=Y1*U^1+Y2*U^2+Y3*U^3 (physical value)

U:Measured period of temperature (sensor output)
Y1-Y3:Coefficinets

Pressure=C(1-T0^2/T^2)(1-D(1-T0^2/T^2)) (physical value)

C=C1+C2*U^1+C3*U^2
D=D1+D2*U^1
T0=T1+T2*U^1+T3*U^2+T4*U^3+T5*U^4

T:Measured period of pressure (sensor output)
C1-C3, D1-D2, T1-T5:Coefficients

The coefficients are fixed for a given set of input samples.
For pressure, the equation is complex and is not a polynomial equation,
which may be different with an assumption of the Blockette 62 (Response
[Polynomial] Blockette).
Could you tell me whether we can extend or create a new field 7 of the
Blockette 62 for the above pressure equation?
We want to avoid the use of Maclaurin expansion because of the problem of
the precision of data conversion.
As another way, is it better to use the Blockette 62 to document the
coefficient and also use the Blockette 31 to explain the above complex
equation?

[Seiji Tsuboi]

Dear WG-II members,

We have proposed new blocketts for a while ago. I would be grateful if you
would let us know any problems concerning with this proposal.

Thank you very much.
Seiji Tsuboi

> -----Original Message-----
> From: fdsn-wg2-d...@iris.washington.edu
> [mailto:fdsn-wg2-d...@iris.washington.edu] On Behalf Of Sleeman,
> Reinoud (KNMI)
> Sent: Thursday, April 22, 2010 4:23 AM
> To: fdsn-w...@iris.washington.edu
> Subject: [fdsn-wg2-data] proposal new blockettes
>
> Dear WG-II members,
>
> I received the following request from JAMSTEC for 2 new blockettes within
> SEED, including a detailed description (see attached file):
>
> ______________________________________________________________________
>
> One is ``Delay time Information Blockette" which describes breakdown of

> totaldelay time. The other is ``Quartz Pressure Blockette" which includes
> 14coefficients to convert count value of quartz-type pressure gauge data

[Seiji Tsuboi]

Dear WG-II members,

Sorry for sending this email repeatedly but I would be grateful if you would
consider this issue and send us any possible problems.