Chapter 2.

HYPOELLIPSE Users Guide

Chapter 2. HYPOELLIPSE Users Guide.....................................................................................1
 2.1 Introduction............................................................................................................................3
 2.2 Specifications for the data-input records..............................................................................4
 2.2.1 Jump record - Format: JUMP character..................................................................................................5
 2.2.2 Travel-time calculations..........................................................................................................................5
 2.2.2.1) Velocity model specifications (Models 1-25).................................................................................6
 2.2.2.2) Vp/Vs ratios....................................................................................................................................7
 2.2.2.3) Travel-time-table specification.......................................................................................................8
 2.2.2.4) Elevation corrections......................................................................................................................9
 2.2.3 Option records.......................................................................................................................................11
 2.2.3.1) Printer option record - Format: PRINTER OPTION Integer.......................................................11
 2.2.3.2) Summary option record - Format: SUMMARY OPTION Integer...............................................12
 2.2.3.3) Magnitude option record - Format: MAGNITUDE OPTION Integer.........................................12
 2.2.3.4) Tabulation option record - Format: TABULATION OPTION Integer.........................................13
 2.2.3.5) Variable-layer option record.........................................................................................................13
 2.2.3.6) Delay model, velocity model, and starting depth option record..................................................15
 2.2.3.7) Missing stations option record - Format: MISSING STATIONS Integer-code...........................15
 2.2.3.8) Sort option record - Format: SORT OPTION Integer-code.........................................................16
 2.2.3.9) Compress option record - Format: COMPRESS OPTION Integer-code.....................................16
 2.2.3.10) Debug option record - Format: DEBUG OPTION Integer-code...............................................16
 2.2.3.11) Find the global minimum in depth - Format: GLOBAL OPTION Integer................................17
 2.2.3.12) Residual option record - Format: RESIDUAL OPTION Integer...............................................19
 2.2.3.13) Weight assignment option..........................................................................................................20
 2.2.3.14) Ignore summary records - Format: IGNORE SUMMARY REC Integer-code..........................22
 2.2.3.15) Header content record - Format: HEADER OPTION Any alphanumeric heading...................22
 2.2.3.16) Comment records - Format: C* Any Comment.........................................................................22
 2.2.3.17) Relocate after revising the delays - Format: RELOCATE Integer-code....................................22
 2.2.3.18) Uofacal option - Format: UOFACAL OPTION Filename.........................................................22
 2.2.3.19) Reset all control parameters to the initial default values............................................................23
 2.2.3.20) Constants noprint option - Format: CONSTANTS NOPRINT Integer......................................23
 2.2.3.21) Blank source option - Format: BLANK SOURCE One-Character Code..................................23
 2.2.4 Reset test records - Format: RESET TEST Integer Real.......................................................................23
 2.2.5 Station list..............................................................................................................................................29
 2.2.5.1) Begin station list - Format: BEGIN STATION LIST Integer-code Integer-date..........................29
 2.2.5.2) Primary station parameters - Formatted as indicated below.........................................................29
 2.2.5.3) Time-dependent station parameters..............................................................................................32
 2.2.5.4) End station list - Format: END STATION LIST..........................................................................36
 2.2.5.5) Additional delays..........................................................................................................................36
 2.2.6 Arrival times..........................................................................................................................................36
 2.2.6.1) Arrival-times-next record.............................................................................................................36
 2.2.6.2) Arrival-time-record format...........................................................................................................37
 2.2.6.3) Instruction record.........................................................................................................................40
 2.2.6.4) Comment records.........................................................................................................................41
 2.2.7 Station parameter changes during run (see 2.2.5.2 and 2.2.5.3)............................................................41
 2.2.8 Change input items 2.2.1 through 2.2.5 during run...............................................................................42

Chapter 2 - 1
  2.2.8.1) Reset record..................................................................................................................................42
 2.2.8.2) Reset S record...............................................................................................................................42
 2.2.8.3) Standard test record......................................................................................................................42
 2.2.9 Use of S-P intervals...............................................................................................................................42
 2.2.10 How to add your own calibration curve - Format: CALIBRATION Integer......................................43
 2.2.11 How to run the same data more than once...........................................................................................43
 2.2.12 Summary of first trial-location specifications.....................................................................................43
 2.2.13 Run data with random errors added.....................................................................................................45
 2.2.14 Define a MASTER event for a random location study.......................................................................46
 2.2.15 How to generate an ARCHIVE-PHASE FILE....................................................................................46
 2.2.16 How to close current ARCHIVE-PHASE FILE and open a new one.................................................50
 2.2.17 Use of magnitudes not determined by HYPOELLIPSE......................................................................51
 2.3 Printed output.......................................................................................................................52
 2.3.1 List of stations available (each station uses two lines in the printout)..................................................52
 2.3.2 Program specifications..........................................................................................................................53
 2.3.3 Vp/Vs ratio............................................................................................................................................53
 2.3.4 Iteration output......................................................................................................................................53
 2.3.5 Quality...................................................................................................................................................54
 2.3.5.1 Based on error estimates................................................................................................................54
 2.3.5.2 Quality based on many parameters................................................................................................55
 2.3.6 Final summary output lines...................................................................................................................56
 2.3.7 Detailed station output: TRAVEL TIMES AND DELAYS:..................................................................58
 2.3.8 Auxiliary RMS sphere output................................................................................................................61
 2.3.9 Focal mechanism plot............................................................................................................................61
 2.3.10 Final Tabulation...................................................................................................................................62
 2.4 Summary record output.......................................................................................................63
 2.4.l Summary record.....................................................................................................................................63
 2.4.2 Phase records in input format with corrected arrival times...................................................................67

Chapter 2 - 2
  2.1 Introduction
HYPOELLIPSE has been developed to meet some of the research needs of the USGS and is in a
constant state of modification and revision to meet new needs and implement new ideas.
There are many subtle uses of the various options, and a complete description of these would
expand the current work to book length. The program and manual are not error-free, and the
author would greatly appreciate feedback on any errors or problems encountered.
The number of "options" available is large, and hence a new user may have trouble deciding
where to begin and which options to use. The easiest course is to start out by specifying only
the minimum amount of information necessary to run the program, including station
locations, velocity model, and a few sets of earthquake arrival times. Then review section 2.2
for modifications to the default parameters or additional calculations required by your data
set. After the run, refer to section 2.3 to interpret the printed output.
The choice of which velocity model and which variable layer thickness options to use will
depend upon how much information one has about the region of study. Some of the possible
choices are described below.

Very little is known about the velocity structure.

In this case a single, simple model consistent with available information could be used. The
linear increase with depth over a half-space model might be chosen, in that few assumptions
need to be made.
Detailed information is available about the thickness of the upper sediment-layer.
In this case, a single, simple model could be used with the thickness of the upper layer varying
from station to station. VMOD of section 2.2.3.5 would be set equal to 0.0. On the STATION
LIST records (section 2.2.5) the layer thickness for variable layer model 1 would be filled out
for each station, and the preferred layer-thickness model for each station would be set equal
to 1.
The region covered by the network includes two or more distinct velocity structures, which
are well known.
In this case the multiple velocity structure option could be used. Each station would be assigned
to one of the velocity models, and that model will always be used to calculate traveltimes to
that station. Note that ray tracing is not done so that a shallow earthquake whose waves pass
through a number of different velocity structures in the earth will be poorly modeled.
However, events deeper than the velocity variations will be modeled relatively well.
Alaska data

Chapter 2 - 3
In the case of Alaska data, three different velocity models are used, depending upon the location
of the earthquake. The station delays may also be a function of earthquake location (see
Chapter 7). Selection of the correct velocity model and station delay is done by the
subroutine USEDLY (see 2.2.3.6), which has been set up specifically for the Alaska region.
This subroutine would need to be modified for use with data from another region. These
options allow all of the events from a very large area to be run in chronological order without
pre-sorting by the source area.
Fault zone time delays
Work in California indicates that there is a low-velocity zone along the San Andreas Fault. To
model this situation, two delays are assigned to each station, one termed delay-model 1 and
the other termed delay-model 2. In addition, the stations on the east side of the fault are
assigned the delay-model preference number 1, while those on the west side are assigned the
number 2. The delay-model used (1 or 2) in locating a particular earthquake is determined by
the delay-model preference number of the closest station to the event. For example, an
earthquake near station XYZ on the west side of the fault would use delay-model 2. Delay-
model 2 has fault zone delays added to the delays of stations on the east side of the fault. The
reverse would be true for the earthquakes on the east side of the fault.
Poisson ratio variation between different velocity models and within one velocity model
If desired, the ratio of the P-wave velocity to the S-wave velocity (Vp/Vs) may be specified
independently for each velocity model and for each model layer. A simple use of this option
would be to specify a different Vp/Vs ratio for each velocity model, but constant within each
model. A more complex use would be to vary the Vp/Vs ratio within each model.

 2.2 Specifications for the data-input records

The input records for this program provide three types of information:
Parameters specific to each user and required for program operation such as the four-character
code and location of each station and the velocity model(s) to be used in travel-time
calculations;
Parameters that control the iterative location procedure or that specify which of the available
output options are to be used;
Arrival-time data to be used in the location of each earthquake.
Except where otherwise noted, items 1 and 2 above have the following format: Columns 1
through 18 contain a keyword which is scanned to determine the number and type of free
format variables on the remainder of the record. The directions for each input item indicate
how many variables are required and whether they are real, integer, or character. For
example,

Chapter 2 - 4
 Format: VELOCITY real, real, real
would indicate that the keyword is VELOCITY and that three real variables must be specified.
All columns beginning with an ! mark are ignored, so comments may be placed on any input
parameter record. Records starting with C* are processed as comments, and the contents are
written to the output file. Note that the number of variables on each record must agree with
the instructions, so leaving columns 19 and above blank is not equivalent to specifying a
value of 0.0, but instead will generate an error message. If two records with the same
keyword in columns 1 through 18 are found, the second one encountered will update the
value(s) specified by the first.
Each of the input items is described below. In many cases, the parameters have default values,
which are enclosed in brackets, []. If the default is desired, then the record does not need to
be included in the input file. The order of sections 2.2.1 through 2.2.5 makes no difference
except that the RESET TEST(1) record, which specifies the Vp/Vs ratio, must precede the
VELOCITY model records (see 2.2.5).
WARNING: Do not include any tabs on the data-input records. Fortran will not read the record
as expected, but the problem can be very difficult to understand, because the record will look
correct when printed or viewed with a text editor.

 2.2.1 Jump record - Format: JUMP character

If a record of the form JUMP FILENAME is encountered, where JUMP begins in column 1 and
the filename is in columns 6-55, then input is switched to the file 'FILENAME'. Input
resumes from the original input file after an end-of-file is reached. The jump file may not
include any additional JUMP records. A JUMP record may not be imbedded within the
STATION LIST records or within an earthquake's PHASE records. A JUMP record after the
ARRIVAL TIMES NEXT record or between the earthquakes may optionally contain in
columns 56-65 the contents of the start of the next record to be processed. For example, if the
columns 56-65 contain the character string '7901' followed by 6 blanks, then all of the
records will be skipped until one beginning with '7901' is found. All of the records
encountered prior to that one will be skipped.

 2.2.2 Travel-time calculations

Both the velocity models and the travel-time tables may be used by HYPOELLIPSE in
computing travel times.

 2.2.2.1) Velocity model specifications (Models 1-25)

Format: VELOCITY real, real, real
The maximum number of velocity model records that can be used is given in section 1.3. There
are three types of model that may be specified and up to 25 models may be defined by up to
Chapter 2 - 5
 96 records (10 models and 36 records on the PC). Models are placed in order starting with
model 1. For these models a reference elevation, Eo (see Test(8) in section 2.2.4), is specified
for the highest elevation in the region in kilometers above sea level. The "top" of the model is
set to Eo. Earthquake depths are still computed with respect to sea level, so negative depths,
up to -Eo km are allowed. Station elevations on the station records must be specified in
meters above sea level. Specification of Eo allows for the correct location of earthquakes
within a region of great topographic relief, such as within a volcano; travel times and take-off
angles are computed correctly, even to the stations that are at a lower elevation than the
hypocenter.
Constant velocity in each layer.
The three real variables to be specified are the P-phase velocity (km/s), the depth to top of layer
(km), and the Vp/Vs ratio. The first record of each model must have a depth of 0.0 km
specified. The model may consist of from 1 to 19 layers over a half space for the SUN
version and 1 to 11 layers over a half space for the PC version. Embedded low velocity
zones are allowed. For example, a 5-km thick layer with velocity of 5.2 km/s and a Vp/Vs
ratio of 1.85 over a 7.0 km/s half space with a Vp/Vs ratio of 1.78 would be specified by the
following two records:

VELOCITY 5.2 0.0 1.85

VELOCITY 7.0 5.0 1.78

Linear increase in velocity over a half space.

For this model the velocity increases linearly from Vo at the surface by K km/s per km until the
half space is reached at a depth of D km. The velocity within the half space is Vh km/s. To
use this type of model, the VELOCITY records are defined as follows:

VELOCITY Vo 0.0 Vp/Vs

VELOCITY K 1.0 Vp/Vs

VELOCITY D 2.0 Vp/Vs

VELOCITY Vh 3.0 Vp/Vs

VELOCITY 200. 4.0 Vp/Vs

Vp/Vs must be specified on each record and must remain constant.

Linear increase in velocity.
For this model the velocity begins at Vo km/s and increases at a rate of K km/s per km. To use
this type of model, the VELOCITY records are defined as follows:

Chapter 2 - 6
 VELOCITY Vo 0.0 Vp/Vs

VELOCITY K 1.0 Vp/Vs

VELOCITY 300. 3.0 Vp/Vs

Vp/Vs must be specified on each record and must remain constant.

Example: To specify two models, the first model with 20 km of 6.0 km/s over a half space of
7.5 km/s and the second model with a linear increase with depth starting at 4.0 km/s at the
surface, increasing 0.11 km/s per km of depth down to 30 km, overlying a half space with a
velocity of 8.1 km/s, the following records would be used:

VELOCITY 6.0 0.0 0.0

VELOCITY 7.5 20.0 0.0

VELOCITY 4.0 0.0 0.0

VELOCITY 0.11 1.0 0.0

VELOCITY 30.0 2.0 0.0

VELOCITY 8.1 3.0 0.0

VELOCITY 200.0 4.0 0.0

A blank record between the models is optional. See the next section (2.2.2.2) for more discussion
of the Vp/Vs ratio.

 2.2.2.2) Vp/Vs ratios

The Vp/Vs ratio must be specified for each velocity layer. If specified as 0.0 then the current
value on the TEST(1) will be used. Use of this feature will be described by considering
various cases:
For all models to use the same Vp/Vs ratio
In this case set the TEST(1) (see 2.2.4) record to the desired value and place it ahead of the
VELOCITY-model records in the program input. Also set the Vp/Vs ratio to zero on all of
the VELOCITY records.
For different models to use different Vp/Vs ratios
Specify Vp/Vs ratio on each VELOCITY model record. Do not vary Vp/Vs within a given
model.

Chapter 2 - 7
Variation in Vp/Vs ratio within a given model
This feature is allowed only for models with constant velocity layers (not for the linear-increase
models). If the Vp/Vs ratio changes within a given model, then a separate S-velocity model is
defined and used for the S-phase travel times. The S-phase model is assigned a number one
higher than the corresponding P-phase model. For example, if two models are specified in the
input stream, and the first model has a variable Vp/Vs ratio, then three models will be
defined and used as follows:

P-Phase Travel S-Phase Travel Vp/Vs

Times Times

Stations Using Model 1 Model 1 Model 2 Variable

Stations Using Model 3 Model 3 Model 3 Fixed

In this example, no PRIMARY STATION parameter record should specify velocity model 2.
(See 2.2.5)
One limitation of this feature is that the S-P interval times (see 2.2.9) do not use the S model, but
instead assume the constant Vp/Vs ratio defined by the TEST(1) record. Another limitation is
that each variable Vp/Vs-ratio model uses up two models, so that a maximum of 12 models
with variable Vp/Vs may be specified.

 2.2.2.3) Travel-time-table specification

These are Models 26-28 on the SUN and Models 11-13 on the PC
In addition to the 25 models previously described, up to three velocity models may be specified
by travel-time tables. The first, which will be model number 26, is read from file number 21.
The second and third are model numbers 27 and 28 and are read from file numbers 22 and
23, respectively. The program prompts for the names of the travel-time tables. See Chapter 6
for the operation of the program TTGEN that may be used to generate a travel-time table.
This option uses portions of the code written by Klein (1985) for HYPOINVERSE. Note that
the elevation of the top of the models (TEST(8), see 2.2.4) must be zero when the travel-time
tables are in use.
The Vp/Vs-velocity ratio is specified for each travel-time model (see 6.5), and is used for
computing S-phase travel times. However, if the specified Vp/Vs ratio is negative, then the
next model will be used for computing the S-phase travel times. For example, if model 26
specifies a negative Vp/Vs ratio, then model 27 will be used for the S-phase travel times for
the stations assigned to model 26, and in this case no station may be assigned to use model
27.

Chapter 2 - 8
  2.2.2.4) Elevation corrections
For stations using travel-time tables, the elevation delay is computed from the station elevation
(E) divided by the apparent vertical velocity within a surface layer with velocity VS. VS is
specified by TEST(2) (see 2.2.4). By using apparent velocity, the elevation correction will
vary from a maximum of E/VS for vertical incidence to a minimum of zero for a horizontal
direct path. This formulation is reasonable for refracted ray paths but will underestimate the
elevation delay for direct ray paths with non-vertical incidence at the surface.
For stations using travel-time calculations, the method above is only used to compute the
elevation correction if the elevation of the top surface of the layer models is set to zero by the
TEST(8) record (see 2.2.4).
If the elevation (in km above sea level) of the surface of computed travel-time models set by the
TEST(8) record is not zero, then the stations with lower elevations are "embedded" within
the model, while stations with higher elevations are fixed at the surface. In this case elevation
corrections are not needed and the value of TEST(2) is ignored. Note that earthquakes may
occur above the "embedded" stations, and that in this case travel times and angles of
incidence are correctly computed. See Figures below for an example of the use of
"embedded" stations.
A consequence of allowing the surface of the models to be above sea level is that earthquakes
may also occur above sea level, and will in this case be given a negative depth. This depth
value given on the printed output and on the summary record in columns 113-117 could then
be negative. The TEST(9) record (see 2.2.4) is used to control how the depth will be entered
on columns 34-36 of the summary record. If TEST(9) is 0.0 then this field of the summary
record has the same value as columns 113-117. If, however, the TEST(9) value is not equal
to 0.0, then negative depths will be reported as -00 on the summary record in columns 34-36.
This was done to accommodate some older software that reads the summary records but does
not expect any negative depths.
The following diagrams illustrate the input parameters for the "old" and "new" ways of dealing
with elevation corrections. In the "old" system, ocean bottom stations can not be
accommodated because negative depths are not allowed. In the illustration for the "new"
system, note that stations below sea level are given a negative elevation.

Chapter 2 - 9
"Old" System of Accommodating Station Elevations. The velocity model is defined only below
sea level and negative elevations are not allowed.

Chapter 2 - 10
"New" System of Accommodating Station Elevations. The "top" of the velocity model is 2 km
above sea level and negative elevations are allowed.

 2.2.3 Option records

This set of records is optional. Include only those required.

 2.2.3.1) Printer option record - Format: PRINTER OPTION Integer

Code Printed output

Chapter 2 - 11
 -2 Only warning messages.

-1 Date and time of each earthquake and warning messages.

0 Final solution for each event showing the residuals at each station.

[1] Above plus one line per iteration. [ ] denotes default value.

2 Above plus residuals at each station for each iteration

3, 4, or 5 Above plus details from many subroutines. Used for debug purposes only.

 2.2.3.2) Summary option record - Format: SUMMARY OPTION Integer

See 2.4.1 for the SUMMARY record format and 2.2.15 for the archive format.

Code SUMMARY record output

0 No SUMMARY records

[1] SUMMARY records on FILE4

2 SUMMARY records on FILE4 and ARCHIVE-PHASE FILE on FILE11

3 ARCHIVE-PHASE FILE on FILE11

4 PHASE records in input format with "fake" arrival times on FILE11

(see 2.2.13 for example of use)

 2.2.3.3) Magnitude option record - Format: MAGNITUDE OPTION Integer

Local magnitude (XMAG) and coda magnitude (FMAG) are computed from formulas given in
Chapter 4.

Code Preferred magnitude used on the SUMMARY record and in the final output
line (See 2.2.17, 1.3.5, and 2.4.1)

[0] XMAG (Amplitude magnitude)

1 FMAG (coda length magnitude)

2 (XMAG + FMAG)/2

3 Prefer FMAG but use XMAG if FMAG is not calculated

4 Prefer XMAG but use FMAG if XMAG is not calculated

Chapter 2 - 12
(Add ten to the code for median rather than the average value to be used as the preferred
magnitude.)
The MAGTYP in column 80 of the SUMMARY record is set to X, F, or A, (corresponding to
XMAG, FMAG, and average) to denote which type of magnitude was used. If no location
can be obtained, then the magnitude is left blank and MAGTYP is set to K.
If the code is negative, the calculation will be based on the F minus S (F - S) rather than the F
minus P (F - P) time. F - P is still entered on the ARRIVAL TIME records as the coda length,
but the S-P interval is subtracted. If S has not been read, the S residual is greater than (F -
P)/10., or the computed S weight is zero, then the calculated rather than the observed S-P
interval is subtracted. See 2.2.17 for the use of magnitudes computed outside of
HYPOELLIPSE.

 2.2.3.4) Tabulation option record - Format: TABULATION OPTION Integer

The tabulation at the very end of each run gives various statistics such as the average residual for
each station.

Code Events included in the final tabulation

0 No tabulation

+1 Tabulation for A quality only

[+2] Tabulation for A and B quality

-2 Tabulation for A and B quality

+3 Tabulation for A, B, and C quality

+4 Tabulation for A, B, C, and D quality

Positive for quality based on error ellipsoid.

Negative for quality defined in HYPO71 (Lee and Lahr, 1972).
See 2.3.5 for definition of A, B, C, and D.

 2.2.3.5) Variable-layer option record

Format:

VARIABLE LAYER Integer Integer Integer

NLAY VMOD LOWV

This record is required for the variable-layer-thickness option. NLAY is the number of the layer
to be varied, VMOD determines how the layer-thickness model is chosen, and LOWV is set
Chapter 2 - 13
 to 1 if an equal and opposite change in the thickness of the layer below the variable layer is
to be made.
For each station two thicknesses are specified for the variable layer, a model 1 thickness and a
model 2 thickness. In the calculation of each travel-time two stations are considered, the
closest station to the epicenter and the receiving station. VMOD is used to specify which of
three options is desired:

VMOD Usage

+1 The thickness specified for the receiving station's preferred model (1 or

2) is used. For example, the station STA has the layer thickness for
variable-layer model 1 equal to 3 km, and model 1 is its preferred
layer-thickness model. Then all of the travel times to station STA will
use 3 km as the variable-layer thickness. With this option only one
thickness needs be specified for each station.

0 The depth to the lower boundary of the variable layer is calculated for the
receiving and for the closest station. If the hypocentral depth is below
the average of the two lower-boundary depths, then the receiving
station's variable-layer thickness is used. For shallower depths, the
lower-boundary depth is set to the average of the receiving station's
and closest station's lower- boundary depths.

-1 The thickness model (1 or 2) preferred by the closest station to the

epicenter is used to determine the variable-layer thickness used at
each station

Example of variable-layer-model velocity in the case where the first-layer thickness is variable
and LOWV equals 1:

VELOCITY STRUCTURE AS SPECIFIED ON CRUSTAL STRUCTURE RECORDS

Depth 0 – Top of 5 km/s layer
Depth 10 – Top of 6 km/s layer
Depth 15 – Top of 7 km/s layer
VELOCITY STRUCTURE WHEN VARIABLE LAYER (1) THICKNESS EQUALS 5 KM
Depth 0 – Top of 5 km/s layer
Depth 5 – Top of 6 km/s layer
Depth 15 – Top of 7 km/s layer

Chapter 2 - 14
Note that in this example no station should be given a variable-layer thickness greater than 15
km.

 2.2.3.6) Delay model, velocity model, and starting depth option record
Format: SELECT DELAY Integer-code
If the code is positive, which is the default, the delay-model used (1-9) will be the one preferred
by the closest station. The velocity model used will be the one preferred by each station, as
indicated on the primary station record (2.2.5.2)
If the code is less than or equal to zero, then the subroutine USEDLY is used to control the
delays, velocity model and starting depth of each event. Subroutine USEDLY, as distributed,
has been tailored for use in processing data from Alaska, and would need to be modified for
another region. The current, Alaska, algorithm in USEDLY, which is used when the code is
negative, does the following:
Reads the first record following the SELECT DELAY record for the name of a file defining the
cylindrical-delay regions.

Selects a velocity model to match earthquake location. Northern model if north of 62.5N,
southern model if south of 62.5N unless within a cylindrical region or within the Gulf of
Alaska. The velocity model is updated before each iteration, but not after the iteration
defined by TEST(37).
The delay model is set to one unless the earthquake location is within a cylindrical-delay region
or within the Gulf of Alaska. The Gulf uses delay model 5 and delay models 2, 3, and 4 are
assigned to cylinders. Up to 10 delay-models may be specified. The delay-model selection
is updated each iteration, but not after TEST(37). Chapter 7 describes the use of cylindrical
regions in more detail.
Sets an upper limit on the maximum starting depth depending on each event's starting location.
Section 2.2.12 summarizes how the first trial depth is determined.
For events in the Gulf of Alaska, fixes depth at 10 km.

 2.2.3.7) Missing stations option record - Format: MISSING STATIONS Integer-code

Code Effect

0 The station list will be searched after each event is located for stations that
would possibly improve the earthquake solution quality. Stations are listed
which are closer to the epicenter than the third-closest station used to
compute the solution or that would reduce the GAP (see 2.3.6) by 30o or
more.

Chapter 2 - 15
 [1] A search for "missing" stations is not conducted.

Searching for missing stations requires that the entire station list be initialized, so this option can
not be used with a negative code on the BEGIN STATION LIST record. (See 2.2.5)

 2.2.3.8) Sort option record - Format: SORT OPTION Integer-code

Code Effect

[0] Stations are listed in the output in order of increasing epicentral distance.

1 Stations are listed in the same order as the input ARRIVAL TIME records.

 2.2.3.9) Compress option record - Format: COMPRESS OPTION Integer-code

Code Effect

0 Printout is compressed by not skipping to the top of a new page for each
solution.

[1] Printout for each earthquake starts on a new page.

 2.2.3.10) Debug option record - Format: DEBUG OPTION Integer-code

Code Effect

[0] This option is not called into play and no additional record is needed.

1 This option is used and the record below must be included:

Debug limits record - the following format must be used:

( 10X, F5.2, 5X, F5.2, 5X, F5.2, 5X, i5, 5X, i5, 5X, i5)
For example:

1-10 11-15 16-20 21-15 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65

Max RMS 0.65 PRES 1.00 SRES 1.00 NWOUT 2 NMAX 8 SEMX 25

The words are typed only for your convenience and the order of the variables is fixed.
The value of any variable that is not specified will be read as zero. This setup would give

Chapter 2 - 16
 a detailed printout of travel times, residuals, etc., for each station only for "Debug events"
defined by:
RMS > MAX RMS, 0.65 s, or
Largest P-Res with computed weight greater than 0.2 > PRES = 1 s, or
Largest S-Res with computed weight greater than 0.2 > SRES = 1 s, or
Combined number of P and S readings weighted out by the program > NWOUT = 2, or
(this excludes readings assigned weight-codes 4-8)
Total number of iterations > NMAX = 8, or
The maximum error estimate > SEMX = 25 km.
The DEBUG OPTION can be used with the COMPRESS OPTION so that each event
will not start on a new page.

TABLE OF DEBUG OPTIONS

CODE EVENT PRINTOUT SUMMARY RECORD

[0] Controlled by PRINT OPTION. Controlled by

SUMMARY
OPTION

+1 Summary line for good events. Detailed for For both good and debug
debug events. events

-1 Same as for +1 Only for good events

+2 No print for good events. Detailed for debug For both good and debug
events. events

+3 Detailed for all events To rerun debug events For both good and debug
with only critical stations set TEST(44) = events
1.0

 2.2.3.11) Find the global minimum in depth - Format: GLOBAL OPTION Integer
This option is now set up to find the best solution in regions where the depths vary from the
surface to 100 km or more. This option can not be used if the hypocenter is fixed on a plane,
so TEST(47) (see 2.2.4) must be set to 0.0. Also, a global solution will not be attempted if the
instruction record fixed location indicator is set for fixed depth (1) or fixed hypocenter (7 or
9).

TABLE OF GLOBAL OPTIONS

Code Effect

Chapter 2 - 17
[1] GLOBAL OPTION is turned off.

0 Global search option is turned on. The "global" search begins by solving two
fixed-depth solutions: S(1) with the event at the Earth's surface (z =
-TEST(8)) and S(2) with z first fixed at DEEPZ = TEST(42) km below sea
level and then allowed to go free.
If the depth of S(2) is within 0.1 km of the surface, the surface solution is
reported.
If the depth of S(2) is less than CUTZ = TEST(27) km below sea level and
 the surface RMS is significantly lower than RMS of S(2) then a free
depth solution starting at S(1) is reported.
 the RMS of S(2) is significantly less than the surface RMS, then S(2) is
reported.
 neither solution has significantly lower RMS, so the one with lower RMS
is reported and the printer output file will give the RMS and depth of both
solutions.
If the depth of S(2) is greater than or equal CUTZ = TEST(27) below sea level
then a solution S(3) with z fixed at SHALZ = (CUTZ - TEST(8))/2.0 km
below sea level is computed.
 If the RMS of the surface solution S(1) is less than the RMS of S(3) then
find a free-depth solution called S(4) starting at S(1).
 If the difference in depth between S(3) and S(4) is less than DEEPZ/10.0
then use S(4) as the reported solution.
 If the solutions S(3) and S(4) are more than DEEPZ/10.0 km apart, then
if S(4) has significantly lower RMS value it is reported as the final
solution.
if S(4) does not have significantly lower RMS than S(3), both are
reported to the printed output, and S(4) is taken as the final solution.
 If the RMS of the surface solution S(1) is greater than the RMS of S(3)
then find a free-depth solution called S(4) starting at S(3).
 If the difference in depth between S(1) and S(4) is less than DEEPZ/10.0
then use S(4) as the reported solution.
 If the solutions S(1) and S(4) are more than DEEPZ/10.0 km apart, then

Chapter 2 - 18
 if S(4) has significantly lower RMS value it is reported as the final
solution.
if S(4) does not have significantly lower RMS than S(1), both are
reported to the printed output, and S(4) is taken as the final solution.

The error limits reported to the output and to the SUMMARY record are the greatest deviations
of depth with RMS less than RMSLIM (See Chapter 3.2 for the definition of RMSLIM),
even if there are intervening peaks in the RMS. This limit is approximately equivalent to one
standard deviation in depth.
For different networks and regions, the values of DEEPZ and CUTZ may need to be adjusted for
best results.

 2.2.3.12) Residual option record - Format: RESIDUAL OPTION Integer

After the initial location of an earthquake, the printed output can be checked for large residuals.
When the original seismograms are reviewed to correct errors, inexperienced processors can
be influenced in their revisions by the printed computer residual. Printing the calculated
value of the residual may be hazardous and not recommended because, in many cases, the
true error, if any, is not reflected by the individual residuals.
To prevent this ill-advised feedback, the "preliminary option" may be used.

TABLE OF RESIDUAL OPTIONS

Code Effect

0 Prints station residual in "residual" format, as described below.

[1] Prints station residual in normal format.

The "residual" format consists of the following:

P and S residuals less than 2.25 s are shown only in absolute value and are rounded to the nearest
0.5 s. The printed residual (Rp) is related to the absolute value of the calculated residual (R)
as follows:

0 < R < .25 Rp = 0

0.25 < R < .75 Rp = 0.5

0.75 < R < 1.25 Rp = 1.0

1.25 < R < 1.75 Rp = 1.5

1.75 < R < 2.25 Rp = 2.0

Chapter 2 - 19
Large residuals are flagged. For residuals (R) that have not been weighted out automatically by
the program, an * is placed after the residual if it meets one of the following criteria:
P Residuals
 R > 0.6 for one of the closest five stations
R > 0.9 for distance less than 150 km
R > 1.5 for epicentral distance less than 350 km
S Residuals
 R > 0.9 for one of the closest five stations
R > 1.5 for epicentral distance less than 150 km
R > 2.0 for epicentral distance less than 350 km

 2.2.3.13) Weight assignment option

Format: WEIGHT OPTION Real Real Real

The relative weight assigned to each reading is dependent upon an integer weight-code, which
can range from 0 for the most impulsive to 4 for readings that are too uncertain to be used in
the hypocentral solution. It is recommended that TEST(29) (see 2.2.4) be set to minus the
standard error of the best readings, and that these readings be assigned zero weight-code.
Less certain readings are then assigned larger weight-codes. If, for example, weight-codes of
1, 2, and 3 are to be assigned to readings that have standard errors that are 5, 10, and 20 times
less certain than the best, respectively, then the three WEIGHT OPTION parameters should
be set to 5, 10, and 20. For processing Alaska seismic data, 0, 1, 2, and 3 weight-codes are
assigned to readings with standard errors ranging up to 0.1, 0.5, 1.0, and 2.0 s, respectively.

Chapter 2 - 20
The graph below is used to assign weight-codes according to these limits.

The default weight and corresponding relative standard error corresponding to each weight-code
is given in the table below:

WEIGHT- STANDARD STANDARD ERROR RELATIVE TO COMPUTED

CODE ERROR (S) READINGS WITH WEIGHT-CODE WEIGHT
ZERO

0 0.1 1.0 1.0

1 0.5 5.0 1/25

2 1.0 10.0 1/100

3 2.0 20.0 1/400

4 INFINITE INFINITE 0.0

To change the default weight assignments, include a record with WEIGHT OPTION starting in

Chapter 2 - 21
 Column 1 followed by the relative standard errors for weight-codes of 1, 2, and 3. For
example, the default weights that used to be set in older versions of HYPOELLIPSE could be
reset with:
WEIGHT OPTION 1.33 2. 4.

 2.2.3.14) Ignore summary records - Format: IGNORE SUMMARY REC Integer-code

Code Effect

[1] The starting location parameters (latitude, longitude, depth, and origin time)
may be taken from the SUMMARY record. (see 2.2.12)

0 The SUMMARY record will be ignored in determining the starting location

parameters.

 2.2.3.15) Header content record - Format: HEADER OPTION Any alphanumeric heading
Used to write a heading of up to 50 characters above each earthquake in the output.

 2.2.3.16) Comment records - Format: C* Any Comment

Any record with C* in columns 1-2 will be printed out during program execution but is
otherwise ignored. Comments may not be imbedded within the station list.

 2.2.3.17) Relocate after revising the delays - Format: RELOCATE Integer-code

This will cause the input file to be rewound and rerun N = Code times. Prior to each rerun, the
station delays for delay-model 1 will be revised by adding the event-weighted, average
station residual. At the end of the job a file is generated on Unit 13 with one record for each
station used. These records are in the format of the primary station records. (see 2.2.5) The P-
and S-delay field for delay-model 1 will contain the revised station delays. The field for
delay-models 2 through 5 will be blank. The default is Code = 0.

 2.2.3.18) Uofacal option - Format: UOFACAL OPTION Filename

Used to specify the name of the file that contains calibration data in the format used by the
UAGI. (see 4.2.3). Type UOFACAL beginning in column 1 and the filename in columns 19-
68.

 2.2.3.19) Reset all control parameters to the initial default values

Format: STANDARD TEST

Chapter 2 - 22
  2.2.3.20) Constants noprint option - Format: CONSTANTS NOPRINT Integer
Controls whether the station list, the user-specified constants, and the control options will be
written to the output file.

Code Effect

0 Do not write out these items.

[1] Write out these items.

 2.2.3.21) Blank source option - Format: BLANK SOURCE One-Character Code

The station list may contain calibration, polarity, and telemetry-delay information for various
sources of data. The arrival-time records have columns to specify the source of the P- and S-
arrival times, the first motion, and the amplitude. However, some sets of arrival-time data
may not have any source code entries. The BLANK SOURCE code specified on this record
will be used whenever the arrival-time source code is blank in computing delays, magnitudes
and corrected polarities.

 2.2.4 Reset test records - Format: RESET TEST Integer Real

These records have RESET TEST typed starting in column 1 and beginning in column 19 the test
number and the new value for the test variable. All of these records are optional and need be
included only if a non-default value is required.

Test No. Default Description

Test Value

1 1.78 Ratio of P-wave to S-wave velocity

2 5.0 P-phase velocity for elevation corrections (km/s). If the value is

negative, make no elevation corrections. If zero, use first-
layer velocity for elevation corrections. If greater than zero,
use this for elevation corrections. In the latter case, value
must be less than first layer P-phase velocity. Used with
computed models if TEST(8) = 0 and with travel-time
tables.

Trial Location

3 0.0 First trial latitude (degrees). North positive. If TEST(3) or (4) =

0, then ignore. See 2.2.12 for use.

Chapter 2 - 23
4 0.0 First trial longitude (degrees). West positive. If TEST(3) or (4) =
0, then ignore. See 2.2.12 for use.

5 -99.0 Used for first trial depth (km with respect to sea level) unless
equal -99 or unless Global Option is in effect. See 2.2.12
for use.

6 0.0 RMS may optionally be computed at additional points on a

sphere surrounding the final hypocenter. This is the radius of
the sphere (km). If zero, no auxiliary RMS values are
calculated. If negative, and if one or more points have lower
RMS than the final solution, continue iteration once starting
at point with lowest RMS value.

7 10.0 Focal Mechanism Plot. Minimum number of first motions for a

first-motion plot to be made. If negative, make a second plot
showing station codes.

8 0.0 Elevation of top of computed models with respect to sea level

(km).

9 0.0 If not zero, reset negative depths in summary record cols. 32-36
to -00. True depth below (positive) or above (negative) sea
level always given in cols. 113-117 of the summary record..

Distance Weighting

10 0.0 Apply distance weighting on this iteration. See also TEST(11)

and (12).

11 50.0 XNEAR = Greatest distance (km) with assigned weight

multiplied by 1.0

12 100.0 XFAR = Least distance (km) with assigned weight of multiplied

by 0.0. See also TEST(46).

13 50.0 Azimuthal Weighting. Apply azimuthal weighting on this

iteration. Warning: this option has not been tested.

Truncation Weighting

14 50.0 If there are 6 or more phases, then weight out large residuals on
this iteration. See also TEST(15).

15 10.0 Give zero weight to residuals greater than this (s).

Chapter 2 - 24
Boxcar Weighting

16 50.0 If there are 6 or more phases, then apply boxcar weighting on

this iteration. See also TEST(17), (34) and (35).

17 2.0 Give zero weight to residuals greater than this times the
standard deviation (s).

Jeffrey's Weighting

18 50.0 Begin Jeffrey's weighting on this iteration. (See also TEST(34)

and (35)).

19 0.05 Use Jeffrey's weighting only if RMS is greater than this (s).

20 0.05 Mu of Jeffrey's weighting function

Test No. Default Description

Test Value

21 9.0 Maximum number of iterations allowed

22 35.0 Limit change in focal depth to approximately this amount (km).

23 0.7 If move would take earthquake above surface, move this

proportion of the way to the surface

24 35.0 Limit change in epicenter to approximately this amount (km).

25 40.0 Fix depth if epicentral change greater than this (km).

26 0.0025 Stop iterations if adjustment squared is less than this (km).

27 20.0 Global solution option: if deep solution converges below this

depth with respect to sea level, continue at a depth half way
between this depth and the surface of the velocity models.
See also TEST(42).

28 0.0 To fix the hypocenter on a plane, set absolute value of this equal
to azimuth of plunge line of plane (0o to 360o measured
clockwise from North). If negative, then a free solution will
be determined starting at the best location on the plane. See
also TEST(30) and TEST(47).

29 -0.1 If TEST(29) is positive, the standard error of readings assigned

Chapter 2 - 25
 zero weight-code is set equal to the RMS residual, unless
there are zero degrees-of-freedom or the estimated reading
standard error falls below TEST(29). In that case TEST(29)
is used for the standard error of the readings. If TEST(29) is
negative, the standard error of the zero weight-code readings
is always set equal to minus TEST(29). See also 2.2.3.13.

30 0.0 Used if TEST(28) is positive, causing solution to be fixed on a

plane. If positive, this is dip of plunge vector of the plane.
See also TEST(28) and TEST(47). If negative, then fix
epicenter and solve only for depth and origin time, ignoring
TEST(47).

Duration Magnitude Parameters (See also TEST(40) and (43))

31 -1.15 C1, constant

32 2.0 C2, *log((F-P) * FMGC)

33 0.0 C3, *DELTA

34 0.0 If not equal 0, scale the normal equations.

35 0.001 Minimum damping of normal equations.

36 100.0 Maximum first trial depth (km), if computed from P-arrival

times.

37 3.0 If termination occurs before this iteration, set iteration number

to this and continue. Prevents iteration from stopping before
all forms of weighting have been applied. After this iteration,
velocity and delay models will not be changed by the
SELECT DELAY (2.2.3.6) option.

38 0.0 If 0, use of S arrivals depends upon S-data indicator on

INSTRUCTION record.
If 1, locate all with and without S arrivals.
If 2 locate all with S arrivals.
If 3, locate all without S arrivals.
If 4, fix all solutions at starting hypocenter, and use S arrival.
If negative, use S arrivals only to fix origin time.

39 1.0 Multiply the S and S-P weight-code weights by this factor.

Chapter 2 - 26
Test No. Default Description
Test Value

40 0.007 Duration magnitude parameter C4; multiplies the DEPTH (see

also TEST(31)-(33) and TEST(43))

41 0.0 If this equals 1, PRINT OPTION is greater than or equal 1, and

SUMMARY OPTION equals plus or minus 1, then write a
new SUMMARY record after each iteration.

42 75.0 Global solution option: deep starting depth (km with respect to
sea level). See also TEST(27).

43 0.0 Duration magnitude parameter C5; multiplies (log ((F-P)

*FMGC)**2)
(see also TEST(31)-(33) and TEST(40))

44 0.0 If 1, rerun "debug events" again (See 2.2.3.10) with critical

stations; if 2, make a second run for all events with critical
stations See note below for definition of critical stations.

45 0.1379 X-scale factor for focal mechanism plots. Adjust for printer in
use. (See 2.3.9)

46 0.0 If TEST(46) not equal 0.0, distance-weighting constant XFAR

(see TEST(12)) will be set to a minimum of 10 km beyond
the distance of the TEST(46)th station. If TEST(46) is
negative, then any station beyond XFAR that would reduce
a gap greater than 60 o by 30o or more is given a distance
weight of 0.5.

47 0.0 Constraint equation weight for hypocenter fixed on plane. A

large value, such as 1000, will prevent out-of-plane
movement. If equal to 0, this option is not used. See also
TEST(28) and (30). This option may not be used with the
GLOBAL OPTION (see 2.2.3.11).

48 6.5 Half-space velocity used for first trial location (km/s).

49 0.0 If absolute value equals 1, compute Vp/Vs and origin time; if

equals 2, also make printer plot of S-P vs P. If negative, use
this origin time for earthquake location.

50 0.0 Compute this number of fixed depth solutions, starting with

Chapter 2 - 27
 Z(1) = -TEST(8) and continuing with Z(i+1) = 1.2*Z(i) +
1.0. The maximum this value can be is 22., which produces
a maximum depth of 225 km.

51 1000.0 Beyond this epicentral distance use first travel-time table

model.

52 2800.0 Wood-Anderson magnification used in XMAG calculations.

53 1.0 If equal to 1, then assume stations with 4-letter codes ending

with e or n are horizontal east-west and north-south stations,
respectively.

54 200.0 If 1st computed trial epicenter is greater than this from closest
station, start location at closest station.

55 19.0 Default century if not specified on the summary record.

Note on TEST(44) - critical stations

In an effort to speed up the identification of reading errors during preliminary runs of data, an
option to automatically rerun each event using only the most important arrivals was
developed. In some cases comparing the solution using only critical stations with the normal
solution can identify reading errors. In the printed output for critical-station reruns, readings
that are not used are marked with an 'X' between the residual and the weight. Critical stations
are defined to be:
The closest four stations with P-phase readings that have weight codes less than 4;
b) Additional stations with P- or S-phase readings are considered one at a time and are added
only if they reduce a gap of greater than 72o by 5o or more;
c) S arrivals are used when available at "critical" stations. If no S arrival is available from a
critical station, then S is used from the closest non-critical station with a weight code less
than 4.

 2.2.5 Station list

The station list is set up so that a complex history of station changes can be maintained, such as
the opening and closing dates and changes in gain and polarity (see discussion in 2.2.7). For
the southern Alaska seismic network, a complex history beginning in 1971 has been
developed. However, in situations where this information is not needed, the station list may
consist of just two entries for each station, with many of the fields left blank. Comment
records that begin with C* in columns 1 and 2 may be included within the station list.

Chapter 2 - 28
  2.2.5.1) Begin station list - Format: BEGIN STATION LIST Integer-code Integer-date
The first record has BEGIN STATION LIST typed in columns 1-18, followed by code and date
of the first event to be run. The date includes year, month, and day (for example: 19921028).
If the station list contains many stations that expired before the time of the first event,
specifying the correct starting date will eliminate the expired stations from the initial printed
station list. The STATION records follow this record. (Note that in this Y2K version the date
must include the century.)

CODE Meaning

0 or 1 Print station list updated to date specified and print new station
parameters during run as changes occur.

-1 Do not print station list or print new station parameters when a

station is updated during run.

 2.2.5.2) Primary station parameters - Formatted as indicated below.

For each station there is one entry with PRIMARY STATION PARAMETERS, such as latitude
and longitude, and one or more entries with TIME-DEPENDENT STATION
PARAMETERS, including calibration parameters and polarity indicator. To speed up the
search for station parameters, the current version of HYPOELLIPSE requires the station list
to be in alphabetical order according to the extended 5-character station code. Right
justifying the station code and concatenating the component (z, n, or e) forms this extended
name. The station list must be arranged so that those stations with 1-character codes precede
those with 2-character codes, which perched those with 3-letter codes, which preceded those
with 4-character codes.
Alphabetical order is not required if the alternate version of subroutine PHAIDX, which is
included in with the source code, is used. In either case, the first station should be near the
center of the network, as it is used as a reference location for calculating the azimuth of
approach of a plane wave. A fake station with the code AAA can be used as the first station.
See 1.3 for the maximum of number of stations allowed.
Format for PRIMARY STATION PARAMETER records:

Item Column Nos. Format

Station Code 1 4 A4

Latitude (Degrees) 5 6 i2

N or blank for North, S for South 7 A1

Chapter 2 - 29
 Latitude (Minutes) 8 12 F5.3

Longitude (Degrees) 14 16 i3

W or blank for West, E for East 17 A1

Longitude (Minutes) 18 22 F5.3

THE FOLLOWING ITEMS ARE OPTIONAL, AND IF LEFT BLANK THE DEFAULT
VALUES WILL BE USED.

Item Default Column Nos. Format

Value

Elevation (Meters) 0 23 27 i5

Preferred velocity model 1 28 29 i2

(SUN: models 1-25 computed, 26-28
from tt table)
(PC: models1-10 computed, 11-13 from
tt table)

Preferred layer thickness model (1-2) 1 30 i1

Layer thickness for model 1 0.0 31 34 F4.2

Layer thickness for model 2 0.0 35 38 F4.2

Preferred delay-model (1-9) 1 39 i1

P-delay for model 1 0.0 40 43 F4.2

S-delay for model 1 0.0 44 47 F4.2

P-delay for model 2 0.0 48 51 F4.2

S-delay for model 2 0.0 52 55 F4.2

P-delay for model 3 0.0 56 59 F4.2

S-delay for model 3 0.0 60 63 F4.2

P-delay for model 4 0.0 64 67 F4.2

S-delay for model 4 0.0 68 71 F4.2

P-delay for model 5 0.0 72 75 F4.2

Chapter 2 - 30
 S-delay for model 5 0.0 76 79 F4.2

*Component, Blank or Z, N, or E for Z 80 A1

Vertical, North-South and East-West
component stations.

*Extended, 5-character, station codes are formed from the station code given in columns 1-4
(shifted to the right) plus the component from column 80. If the component is blank, z is
assumed. This 5-character name must agree with the name on the arrival-time record (see
2.2.6.2).
NOTE ON DELAYS: The total delay used for the S-phase is just the S-delay given on the
STATION record. [Some earlier versions of HYPOELLIPSE added a term to the S-delay
equal to (Vp/Vs ratio)*(P-delay).]

Chapter 2 - 31
  2.2.5.3) Time-dependent station parameters
Format for TIME-DEPENDENT STATION PARAMETER records:

Item Default Column Nos. Format

Value

Station code 1 4 A4

* 5 A1

If any of the following items are left blank, default values will be used.

Station weight (Multiplies the weight 1.0 6 9 F4.2

derived from weight-code. (see 2.2.6.)

Primary system response code: 1 10 11 i2

0 = for Wood-Anderson
1= USGS Central California Network Standard
2 = EV-17 and Develco
3 = EV-17 and Teledyne
4 = HS-10 and Develco
5 = L-4C and Develco
6 = L-4C and Teledyne
7 = L-4C replacing HS-10 and Develco
8 = Ten-day Recorders
9-17 = User specified calibration curve (See 2.2.10 and Chapter 4)
18 = Use UAGI magnitude calculation (See 4.2.3)

A1VCO 5-Hz Calibration (mm peak-to- 12 15 F4.0

peak) (not used by HYPOELLIPSE)

XMAG calibration constant-C10 (See 0.0 16 20 F5.2

4.2.2.2)

XMAG correction (added to amplitude 0.0 21 24 F4.2

magnitude) NOTE: If the number typed
is the actual magnitude correction plus
10, the magnitude for this station will
be computed and listed but not used in
computing the average event
magnitude. For example, if XMAG
correction = 10.2, a correction of +0.2
is applied to all XMAG's for this

Chapter 2 - 32
 station, but none are used in the
average magnitude computed for each
event.

XMAGWT magnitude weight (If zero, 1 25 i1

exclude XMAG's from this station from
event mean and median XMAG
magnitude calculations. Print "e" next
to XMAG value in output.

FMAGWT magnitude weight (If zero, 1 27 i1

exclude FMAG's from this station from
event mean and median FMAG
magnitude calculations. Print "e" next
to FMAG value in output.

FMAG correction (Multiplies the observed 1.0 28 31 F4.2

coda).

P-weight-code replacement. For this none 32 i1

station, an ARRIVAL TIME record
with P-weight-code of 0, 1, 2, or 3 will
be replaced by this code. If blank, use
assigned weight.

S-weight-code replacement. For this none 33 i1

station an ARRIVAL TIME record with
an S-weight-code of 0, 1, 2, or 3 will be
replaced by this code. If blank, use
assigned weight.

Field gain setting (Not used by 0 35 36 i2

HYPOELLIPSE).

Palmer attenuator setting of 0, 1, or 2. (Not 0 37 i1

used by HYPOELLIPSE.)

Year, month, and day (e.g. 19891231). 99999999 38 45 i8

Hour of expiration of information in these 0 46 47 i4

entries. If another entry with revised
time dependent parameters does not
follow, then this is time of station
expiration.

Chapter 2 - 33
Two telemetry delays are specified. The primary delay is used unless (1) the source of data is one
of those specified for the alternate delay or (2) the source code is "T" or "N", both of which
have already been corrected. Certain source codes are equivalent for the purposes of the
telemetry delays. For example, the USGS film viewer (source code "V") is equivalent to the
USGS one-film digitizer (source code "1"). A complete list of source codes is given in
section 4.2.2.2. The following table shows delay-equivalent codes. Only one code from a set
of equivalents need be included as an alternate delay code.

Source of data Equivalent codes

USGS Film V, *, 1, 4

USGS Tape S, E, 2

UAGI Film %, A, F

UAGI Masscomp or D, J, X,
SUN Computers

USGS./UAGI PC's P, O, U, I, G, K

Primary telemetry time correction 0.0 48 51 F4.2

Source codes of arrival times that will use 52 55 4A1

the alternate telemetry time correction

Alternate telemetry time correction 0.0 56 59 F4.2

Polarity indicators may be any of the following:

Indicator Meaning Focal Mechanism Symbol:

N Normal Same as phase record

R Reversed Reverse of phase record

+ Probably normal Same as phase record

- Probably reversed Same as phase record

? Unknown Question Mark

Five date-source indicators may be specified, one for each of the following sources of
Alaska data:

Chapter 2 - 34
 Item Default Column Nos. Format
Value

Source of data

USGS Film Source

codes
effected

USGS Tape or USGS/UAGI PC's V, *, 1, 4 60 A1

ATWC Film S, E, 2, P, 61 A1
O, U, I,
G, K

UAGI Film W 62 A1

UAGI Computers %, A, F 63 A1

Other sources D, J, X 64 A1

any other 65 A1
code,
including
blank

The following items allow amplitude measurements for a given station to be made on up to four
additional recording systems, each with a different frequency response:

System response code 72 73 i2

A1VCO 5-Hz, Calibration (mm peak-to-peak) (not used 74 77 F4.0

by HYPOELLIPSE)

XMAG Calibration constant-C10 78 82 F5.2

System response code 84 85 i2

A1VCO 5-Hz, Calibration (mm peak-to-peak) (not used by 86 89 F4.0

HYPOELLIPSE)

XMAG Calibration constant-C10 90 94 F5.2

System response code 96 97 i2

A1VCO 5-Hz, Calibration (mm peak-to-peak) (not used by 98 101 F4.0

HYPOELLIPSE)

Chapter 2 - 35
 XMAG Calibration constant-C10 102 106 F5.2

System response code 108 109 i2

A1VCO 5-Hz, Calibration (mm peak-to-peak) (not used by 110 113 F4.0
HYPOELLIPSE)

Calibration constant-C10 114 118 F5.2

If station has been moved a small distance, these are the new coordinates:

New minutes of latitude 119 123 F5.3

New minutes of longitude 124 128 F5.3

New elevation 129 132 i4

 2.2.5.4) End station list - Format: END STATION LIST

A record with END typed starting in column 1 follows the last STATION record. The record
immediately following the BEGIN STATION LIST record may be of the form JUMP
FILENAME, where the STATION records are contained in the file 'FILENAME'. In this
case, no END record is used in either the original input file or the jump file.

 2.2.5.5) Additional delays

Delays may be specified for 5 additional models (numbers 6-10). Each set of additional delays
begins with a record with DELAY typed in columns 1-5 and an integer delay-model number
(6-10) beginning after column 18. This record is followed by a set of records with station
code in columns 1-4, followed by P Delay and S Delay (s) in free format. After the last
station, a record with END starting in column 1 ends the set of delays. The stations need not
be in alphabetical order, and stations not included will be given delays of zero.

 2.2.6 Arrival times

 2.2.6.1) Arrival-times-next record

A record with ARRIVAL TIMES NEXT starting in column 1 signals the start of the ARRIVAL
TIME records. Each earthquake consists of four types of records: SUMMARY ('/' or '\' in
column 83), ARRIVAL TIME, COMMENT ('C*' in columns 1-2) and INSTRUCTION. If
any SUMMARY records are present, the first record of the event must be the primary
SUMMARY record ('/' in column 83) and this record provides the starting location for the
event unless the IGNORE SUMMARY records is in effect (see 2.2.3.14 and 2.2.12). Each
event must end with a series of one or more INSTRUCTION records. A maximum of NPA
(see 1.3) records may be associated with each event.
Chapter 2 - 36
  2.2.6.2) Arrival-time-record format
For each seismograph station recording the earthquake, an ARRIVAL TIME record is typed as
follows. A maximum of 256 phases, counting P and S phases, may be used for each
earthquake.

Item Column Numbers Format

Station code 1 4 A4

Alphanumeric symbol describing P-phase arrival 5 A1

(for example, I or E)

P-Phase descriptor 6 A1

P = P arrival read on vertical component

N = P arrival read on North-South component
E = P arrival read on East-West component
If N or E is used, then neither the coda duration
nor the maximum amplitude will be used and the
first motion direction will not be plotted.
The extended 5-character station code is formed
from the station code in columns 1-4 (shifted to
the right) plus the phase descriptor. If the phase
descriptor is not n or e, then z is assumed. This
extended name must agree with the extended
name in the station list (see 2.2.5.2).

First-motion direction of P arrival 7 A1

c,
C,
u, Compression
or
U

d,
Dilatation
D

+ Questionable compression

-
Poor dilatation

Chapter 2 - 37
 z, Z Nodal, and not clearly up or down

n, N Noisy

. or Blank Not readable

P-weight-code 8 F1.0

0 or blank Full weight

l Partial weight

2 Partial weight

3 Partial weight

4, 5, 6, 7, 8 No weight

9 Use S-P interval (see 2.2.9)

If the P phase is a secondary arrival refracted along the

bottom of the ith layer, type the value of I here. If
event is in the (i + 1)th layer, direct wave calculation
is made. If the event is deeper than the (i + 1)th layer
9 i1
or the distance is too short for this refraction to be
possible, then the computed weight is reset to zero.
This option only operates with computed layer
models with constant velocity (see 2.2.2.1.a).

Year, month, day, hour, minute (e.g. 8912312359) 10 19 i10

Seconds of P arrival. If left blank, the assigned weight- 20 24 F5.2

code will be set to 8 both during this run and on the
ARCHIVE ARRIVAL TIME record.

Seconds of S arrival 32 36 F5.2

S remark (analogous to columns 7-9, e.g. iSN) 37 39 A3

(Not used by HYPOELLIPSE.)

S-weight-code 40 F1.0

Chapter 2 - 38
Maximum peak-to-peak amplitude. Values from .001 to 44 47 F4.0
9,999 are entered as positive. Negative entries are
multiplied by -10,000 to allow for values of 10,000 to
9,990,000. Units depend on calibration data (see
Chapter 4).

Period of maximum amplitude in hundredths of s. If left 48 50 F3.2

blank, 0.1 will be used.

Siemens gain state: 0 = high; 1 = low (gain times 1/4) 61 i1

A1VCO gain-range state. 0 = high; 1 = gain times 1/10; 2 62 i1

= gain times 1/500.

Any remark 63 64 A2

Time correction (s) 66 70 F5.2

F-P time interval (s), for FMAG calculation. In USGS 71 75 F5.0

practice, one measures the time between the first P
arrival and the point where the peak-to-peak
amplitude of the signal drops below 1 cm on a
Teledyne Geotech Develocorder film viewer (X 20
magnification). If the F-P time is less than 1.25 times
the S-P time, then the FMAG is not calculated.

Polarity source code 105 A1

P-arrival source code 106 A1

S-arrival source code 107 A1

Amplitude source code 108 A1

Coda-duration source code 109 A1

 See 4.2.2.2 for a listing of source codes used in Alaska.

Chapter 2 - 39
  2.2.6.3) Instruction record
After each set of ARRIVAL TIME records for a particular earthquake, at least one
INSTRUCTION record follows.

Item Column Nos. Format

MORE Indicator for another INSTRUCTION record 1 4 A4

following this one. Leave blank if no additional
INSTRUCTION records follow. Type MORE if
another one follows. The earthquake will be processed
once for each INSTRUCTION record.

Processing Status 9 A1

Event Type. See 2.4.1 for definition of Processing Status 10 A1

and Event type. If a SUMMARY record does not
precede an event, then the Processing Status and Event
Type from the INSTRUCTION record are placed in
columns 74 and 92 of the SUMMARY record
generated when HYPOELLIPSE is run. If the event is
in Archive Format, and is therefore already preceded
by a SUMMARY record, then the event type and
processing status on the INSTRUCTION record are
ignored.

S-data indicator. 0 if S data is not to be used. 1 for use of 18 i1

S data in solution. TEST(38) must be set to 0 for S-
data indicator to be used (see 2.2.4).

Fixed location indicator 19 i1

0 implies nothing fixed.

1 implies depth fixed at trial depth.
7 implies hypocenter fixed but origin time free.
8 implies origin time fixed at trial origin time.
9 implies location fixed at trial latitude, longitude, and depth.
If origin time is entered on this record (cols. 74-80), or on a SUMMARY record,
then origin time will also be fixed (see 2.2.12).
If event type is T, R, or N on SUMMARY RECORD, fixed location indicator is
ignored (see 2.4.1).

Chapter 2 - 40
 Trial depth 20 24 F5.2

Trial latitude (Degrees) 41 42 F2.0

N or blank for North, S for South 43 A1

Trial latitude (Minutes) 44 48 F5.2

Trial longitude (Degrees) 54 56 F3.0

W or blank for West and E for East 57 A1

Trial longitude (Minutes) 58 62 F5.2

USGS "CUSP ID" (not used by HYPOELLIPSE) 63 73 A11

Trial origin time (Minutes) 74 75 F2.0

Trial origin time (Seconds) 76 80 F5.2

Sequence Number - will be transferred to columns 94-98 92 96 A5

of SUMMARY record.

 2.2.6.4) Comment records

Any phase record with C* in columns 1-2 will be printed out during program execution but
otherwise will be ignored. There is no limit to the number of COMMENT records per event,
except that they count along with SUMMARY records, ARRIVAL TIME records, and the
INSTRUCTION record toward the maximum number of records allowed per event -
currently set at 256. In this way a comment can be made, for example:
C* Station XYZ may have cross-feed
or
C* these readings may be from two earthquakes.
2.2.6.5) Jump records
If a record with 'JUMP FILENAME' beginning in column 1 is encountered then input is
transferred to file 'FILENAME'. When an end-of-file is reached on the subsidiary input file,
input resumes from the original file.

 2.2.7 Station parameter changes during run (see 2.2.5.2 and 2.2.5.3)
The station list record file may be set up so that station parameter changes will automatically be
made as during a single batch run of a set of earthquakes. Each STATION record has an

Chapter 2 - 41
 expiration date and time. If left blank the year is set to 1999. But if, for example, the station
calibration changed on 760120 at 1432 from 5.1 to 8.3, then two STATION records would be
included in the file. The first would have 5.1 for calibration and an expiration of
7601201432. The second, which must directly follow the first, would be identical except 8.3
would replace 5.1 and the expiration date and time would be updated. As many STATION
records as required can be grouped together like this. The expiration date of each station is
checked against the current event time before each event is processed.
CAUTION: In order to use this system of automatic updating of station parameters, the
earthquake data set must be run in chronological order. All updates for a particular station
must also be in chronological order. Note that if the events are rerun, as described in section
2.2.11, then they will not be run in chronological order, so station parameter updating will not
operate correctly and errors may result.

 2.2.8 Change input items 2.2.1 through 2.2.5 during run

Any or all of these items may be changed as follows:

 2.2.8.1) Reset record

Type RESET starting in column 1. This record is placed following the last INSTRUCTION
record of an event or set of events, and it switches the program to the input mode in which
items 2.2.1 through 2.2.5 may be entered. All location parameters will remain in effect except
those for which new 2.2.1-2.2.5 records are included. To switch back to reading PHASE
records, terminate the 2.2.1-2.2.5 items with an ARRIVAL TIMES NEXT record.

 2.2.8.2) Reset S record

Type RESET S starting in column 1. This is the same as RESET described above except that (1)
a tabulation of average residuals will be printed for the set of events preceding the RESET S
record, and (2) a tabulation will begin anew on the events that follow.

 2.2.8.3) Standard test record

A record with STANDARD TEST starting in column 1 will reset input items 2.2.3 and 2.2.4 to
default values. This record is placed after a RESET or RESET S record.

 2.2.9 Use of S-P intervals

If the same time base is not available for some stations, it is still possible to use the recorded S-P
intervals in the hypocentral solution. To do this, set the weight-code assigned to the P arrival
(column 8) to 9, and the weight-code assigned to the S arrival (column 40) to that desired for
the S-P interval. This will only work with velocity models that have the same ratio of P-wave
to S-wave velocity in each layer.

Chapter 2 - 42
  2.2.10 How to add your own calibration curve - Format: CALIBRATION Integer
In addition to the 8 system response tables stored with HYPOELLIPSE, the user may supply
tables for up to nine additional seismic systems, corresponding to system numbers 9 through
17. To do this, place a record with CALIBRATION starting in column 1 and the number of
additional system calibration tables to be added beginning in column 19 in free format. Each
table consists of two records with the values of RSPA for n = 1,20 on the first and n = 21,40
on the second. The format is (20F4.2). The first two records correspond to system number 9,
the second two to number 10, etc., up to the total number of tables to be added. See Chapter 4
for the definition of RSPA(n).

 2.2.11 How to run the same data more than once

A set of ARRIVAL TIME records may be run with a variety of velocity models, station lists, trial
depths, or any other of the variable parameters defined in sections 2.2.1 through 2.2.5. First,
place the ARRIVAL TIME records in a separate file (named PHDATA in the example below).
Then set up the input file as in this example:
Items 2.2.1 through 2.2.5 as desired for 1st run
ARRIVAL TIMES NEXT
JUMP PHDATA
RESET
New items 2.2.1 through 2.2.5 as desired for second run
ARRIVAL TIMES NEXT
JUMP PHDATA
NOTE: HYPOELLIPSE always assumes that the earthquakes being processed are in
chronological order when keeping the station parameters up-to-date. Therefore, if the events
being rerun span any changes in the station parameters, the station file will have to be
included again prior to rerunning the data.

 2.2.12 Summary of first trial-location specifications

For each parameter, the sources are given in order of decreasing priority.
LATITUDE AND LONGITUDE
INSTRUCTION record, if specified (columns 41-44 contain some non-blank characters),
else
SUMMARY* record, if specified (columns 17-20 contain some non-blank characters) -
(See 2.2.3.14 to ignore SUMMARY records), else

Chapter 2 - 43
 TEST(3) and (4) values, if both not equal zero, else
Inversion* of up to the first 10 P-arrival times.
DEPTH
If Global Option is in effect, then use multiple starting depths as described in
2.2.3.11.
Use INSTRUCTION record, if depth is specified (columns 21-24 contain some non-
blank characters), else
SUMMARY* record, if specified (columns 33-36 contain some non-blank
characters) - (See 2.2.3.14 to ignore SUMMARY records), else
Trial depth specified in SUBROUTINE USEDLY if the SELECT DELAY
code is not zero and USEDLY sets a depth not equal to 99999. (See 2.2.3.6),
else
TEST(5), if not equal -99.0, else
Inversion* of up to the first 10 P-arrival times, rounded to nearest 5 km and
not less than 15 km or greater than TEST(36).
ORIGIN TIME
Computed from distribution of S vs P if TEST(49) equals -1 or -2 and the number of
pairs of S and P times is greater than 2.
INSTRUCTION record, if specified (columns 74-77 contain some non-blank
characters), else
SUMMARY* record, if specified (columns 11-14 contain some non-blank
characters) - (See 2.2.3.14 to ignore SUMMARY records), else
Computed from distribution of S vs P if TEST(49) equals 1 or 2 and the
number of pairs of S and P times is greater than 2.
Define TO(i) = TP(i) - (TS(i) - TP(i))/(TEST(1) - 1.0) where
TP(i) is the P-arrival time at the ith station,
TS(i) is the S-arrival time at the ith station,
TEST(1) is the Vp/Vs ratio
Use the average value of TO(i) if at least one station has both P- and S- arrivals, else
Inversion of up to the first 10 P-arrival times based on a halfspace with velocity =
TEST(48).

Chapter 2 - 44
 NOTE: If TEST(38) is negative, then the origin time will be fixed at the average
value of TO(i), unless TEST(49) is negative, in which case it will be fixed at the value
determined by the Wadati plot (extrapolation of the (TS - TP) vs TP curve to the point
where (TS - TP) equals zero).
*NOTE: Starting parameters will not be taken from the SUMMARY record if the
IGNORE SUMMARY RECORD is in effect (See 2.2.3.14).

 2.2.13 Run data with random errors added

In order to study the effectiveness of this inversion program for varying station distributions or
earthquake distributions, the following method may be used:
a) Generate a "fake" set of perfect phase data for the desired earthquake distribution. This can be
done by setting up a group of ARRIVAL TIME records with the stations of interest and
including arbitrary P- and S-phase data for each station. Then use a series of INSTRUCTION
records, each with a fixed location indicator and one of the desired test earthquake locations
(See 2.2.6.3). Run this data with SUMMARY OPTION 4 in order to generate the "fake" set
of arrival time data. If these data were fed back into the program, the result should be the
desired test earthquake locations, each with a zero RMS residual.
b) In order to have random errors added to each arrival time each time the "fake" data are
relocated, type a control record with SCATTER starting in column 1, P-standard error
starting in column 20 (F5.2), and S-standard error starting in column 32 (F5.2). Note that this
format is fixed. These should be the standard errors for readings with a weight-code of zero.
For readings with weight-codes of 1, 2, and 3, larger deviations will be added, based on the
WEIGHT parameters (see 2.2.3.13). Place this record directly before the first ARRIVAL
TIME record of the fake data set. If an archive phase file is being generated, the arrival times
will reflect the added random errors. Each time this job is run a new series of random
numbers will be generated because the random number seed is initialized by the time, month,
and year of the run.
If the velocity model is also changed, then one can also simulate the systematic errors introduced
by not knowing the true earth structure.
Note that an archive file written when SCATTER is in effect will have all of the original arrival
times perturbed by random errors.
c) To simulate a non-Gaussian distribution, add a record with NONGAUS starting in column 1,
fraction of readings with larger error starting in column 20 (f5.2), and factor by which to
increase P & S standard errors starting in column 32 (f5.2). for example, to randomly
increase P & S standard errors by a factor of 5 10% of the time, use:
NONGAUS 0.1 5.0

Chapter 2 - 45
  2.2.14 Define a MASTER event for a random location study
This option may be used to test the hypothesis that all of the events in a set of locations are
located at the same hypocenter, and are only apparently shifted due to random reading errors
and variations in which phases were read for each event. Set up a "perfect" event (an event
with zero RMS residual) located at the hypothesized hypocenter. Place this as the first of a
set of earthquakes and precede it with a record that has MASTER starting in column 1.
Specify the reading errors with a SCATTER record next (see 2.2.13). All events following
the first will have their minutes and seconds of P and S replaced with the "master" values.
Due to the SCATTER option, the computed locations will vary randomly. The spatial
distribution of the resulting random distribution and the summary statistics at the end of the
run can be compared with the actual data set to test the hypothesis.

 2.2.15 How to generate an ARCHIVE-PHASE FILE

HYPOELLIPSE can be used in a mode that utilizes and generates data in a data base that
combines the raw data measurements of the PHASE records, the SUMMARY record, and
certain derived parameters for each station such as the distance, the azimuth, and the angle of
incidence (see Data Base Organization table below). To set up HYPOELLIPSE to generate
this file, named the ARCHIVE-PHASE FILE, set the SUMMARY OPTION code equal to 2
or 3 (see 2.2.3.2). The file is generated by the following steps:
Write out SUMMARY record(s).
Case 1. The input event included either zero or one SUMMARY records. (Note that the first
(primary) summary record will have a "/" in column 83. Any additional summary records
will have a "\" in column 83.)
Case 1a. A new earthquake solution was generated. Then: Write a
SUMMARY record for the new solution and discard the previous SUMMARY
record, if there was one.
Case 1b. A new solution was not generated. This could happen, for example, if
all of the arrivals are weighted out. Then: Write out a FAKE SUMMARY
record followed by the SUMMARY record that was previously associated
with the event, if there was one.
Case 2. The input event included two or more SUMMARY records.
Case 2a. A new earthquake solution was generated. Then: Write out the new
SUMMARY record followed by all but the first of the SUMMARY records
previously associated with the event.
Case 2b. A new solution was not generated. Then: Write out a FAKE
SUMMARY record followed by all of the SUMMARY records previously
associated with the event. A "FAKE" summary record is just a placeholder,
Chapter 2 - 46
 with the earthquake location and depth fields left blank.
For each station, write out an augmented PHASE record with original phase data and
computed data.
Write out the original INSTRUCTION record.
The generation of archive-phase files is not compatible with option of running
events with more than one INSTRUCTION record (see 2.2.6.3, 'MORE' ).
Data Base Organization

Organization of the old and new data base structures. Raw and derived data that were
previously stored in three files are now combined into a single ARCHIVE-PHASE
FILE

OLD DATA BASE NEW DATA BASE

PHASE records with raw data measurements One archive-phase file with raw and
in one file. derived information for each station as
well as the derived earthquake
SUMMARY records with: derived earthquake solution parameters.
solution parameters, such as location
depth, origin time and magnitude in
another file.

Printed listings with: derived station

information, such as distance, azimuth,
angle of incidence, and magnitude.

The ARCHIVE-PHASE FILE may be used as a HYPOELLIPSE input phase file. In that case,
the starting location, depth, and origin time will be taken from the first SUMMARY record
associated with the event unless overridden by a location, depth, or origin time on the
INSTRUCTION record (section 2.2.6.3). The format specification for reading an ARCHIVE
ARRIVAL TIME record is as follows:

Item Col. Nos. Format

Station code 1 4 A4

Any two alphanumeric symbols to describe P phases. See 5 6 A2

2.2.6.2)

First motion direction of P arrival 7 A1

Chapter 2 - 47
 c, C, u, or U Compression

d, D Dilatation

+ Questionable compression

- Questionable dilatation

n, N Noisy

.or Blank Not readable

z, Z Nodal, and not clearly up or down

P-weight-code 8 F1.0

0 or blank Full weight

1 Partial weight

2 Partial weight

3 Partial weight

4, 5, 6, 7, 8 No weight

9 Use S-P interval (see 2.2.9)

If the P phase is a secondary arrival refracted along the bottom 9 i1

of the Ith layer, type the value of I here. If event is in the (I
+ 1)th layer, a direct wave calculation will be made. Below
that, the weight is reset to zero.

Year, month, day, hour, minute (e.g. 9812312358) 10 19 i10

Seconds of P arrival 20 24 F5.2

Distance (km)* 25 28 F4.1

AZM - Azimuth from epicenter to station (degrees)* 29 31 F3.0

Seconds of S arrival 32 36 F5.2

S remark 37 39 A3

S-weight-code 40 F1.0

AIN - Angle of ray leaving hypocenter (degrees)* 41 43 F3.0

Chapter 2 - 48
Maximum peak-to-peak amplitude. Values from .001 to 9,999 44 47 F4.0
are entered as positive. Negative entries are multiplied by
-10,000 to allow for values of 10,000 to 9,990,000.

Period of maximum amplitude (s). If left blank, the standard 48 50 F3.2

period as specified in the station list will be used

P travel-time computed* 51 54 F4.2

P standard error* 55 57 F3.2

D, B, M, J, X, R, G, or * weight-code (See 2.3.7 for 58 A1

definition)*

Instrument Period (S=short, L=long, B=broad) 59 A1

Instrument Gain (H=high, L=low) 60 A1

Siemens gain state: 0 = high; 1 = low (gain times 1/4) 61 i1

A1VCO gain range state. 62 i1

0 = high; 1 = gain times 1/10; 2 = gain times 1/500

Remark (Recorder location in column 63, e.g. F=Fairbanks) 63 64 A2

Corrected first-motion symbol 65 A1

Time correction (s) 66 70 F5.2

F-P time interval (s) for FMAG 71 75 F5.0

P:RES - Residual of P arrival (s)* 76 80 F5.2

S standard error* 81 83 F3.2

D, B, M, J, X, R, G, or "*" weight-code* 84 A1

S:RES - Residual of S arrival (s)* 85 89 F5.2

P delay* 90 92 F3.1

S delay* 93 95 F3.1

P elevation delay* 96 98 F3.1

System response code* 99 100 i2

XMAG* 101 102 F2.1

Chapter 2 - 49
 FMAG* 103 104 F2.1

Polarity source code** 105 A1

P-arrival source code** 106 A1

S-arrival source code** 107 A1

Amplitude source code** 108 A1

Coda-duration source code** 109 A1

Number of satellite hops (NHOP) in telemetry path, each 110 i1

producing a delay of 0.27 s. (If the P-arrival and S arrival
sources are not the same, then NHOP is set according to
the P-arrival source.) *

* These items are added to the original ARRIVAL TIME record, but are not used in
subsequent runs of HYPOELLIPSE
** See 4.2.2.2 for a listing of source codes.
Use of Polarity Source Code by HYPOELLIPSE:
STATION records have two fields for keeping track of station polarity (see 2.2.5), the Polarity-
Reversal-Indicator in column 34 and the Tape-Polarity-Reversal-Indicator in column 48, and
these two indicators do not always agree. The Polarity-Reversal-Indicator is used in
correcting the observed first motion unless the polarity source code is S, in which case the
Tape-Polarity-Reversal-Indicator is used.

 2.2.16 How to close current ARCHIVE-PHASE FILE and open a new one
In some situations it is desirable to close the archive-phase file specified when the program was
started and to open a new file with a different name. To do this, use a record with ARC in
columns 1-3 and the new file name in columns 19-68. The ARC record must either be
included with the initial input data or follow a RESET record. The following example
illustrates a run in which P1.ARC corresponds to P1.PHA and P2.ARC corresponds to
P2.PHA.
 {Items 2.2.1 through 2.2.5}
 ARC P1.ARC
ARRIVAL TIMES NEXT
JUMP P1.PHA
RESET
ARC P2.ARC

Chapter 2 - 50
 ARRIVAL TIMES NEXT
JUMP P2.PHA

 2.2.17 Use of magnitudes not determined by HYPOELLIPSE

For some earthquakes it is desirable to use a magnitude calculated by another organization, and
to enter this magnitude in columns 37-38 of the SUMMARY record as the preferred
magnitude. In this situation MAGTYP in column 80 is also set to some code other than F, X,
or A. For example, in Alaska we use the following codes:

Mag. Type Source

B PDE mb

C Canadian ML

G UAGI ML

H Helicorder (approximate ML)

P Palmer ML

L Lamont-Doherty Earth Observatory

O Other

S PDE MS

W Moment magnitude (Mw)

When earthquakes are being rerun, if the SUMMARY record precedes the phase data (See
2.2.15) and has MAGTYP not equal to F, X, or A, then the preferred magnitude and
MAGTYP on the newly generated SUMMARY record will not be changed. Thus, the
preferred magnitude is preserved through repeated runs of HYPOELLIPSE.

 2.3 Printed output

The line-printer outputs of HYPOELLIPSE are generally self-explanatory. The following
explanations may be helpful for first-time users.

 2.3.1 List of stations available (each station uses two lines in the printout)
Heading Explanation

P THK Preferred variable layer thickness. Either 1 or 2.

VAR LAYER THICKNESS Two thicknesses may be specified for the variable layer.
Chapter 2 - 51
 1 and 2

P MOD Number of velocity model to be used with this station.

P DLY Preferred delay model 1 through 5.

PDLY1, SDLY1 Model 1 time delays for P and S arrivals. Followed by

delays for models 2 through 5.

SYS System response code. See 2.2.5 on station list for code
number assignments.

CALR Standard calibration for XMAG

XMGC Amplitude magnitude correction

FMWT Weight for F-P magnitudes

FMGC F-P magnitude correction. Multiplies observed F-P

interval.

WT P Replace P-weight-code of 0, 1, 2, or 3 with this. Ignored

if equal to 10.

WT S Replace S-weight-code of 0, 1, 2, or 3 with this. Ignored

if equal to 10.

POL Two codes are printed, the first for Develocorder polarity
and the second for tape polarity. If 1, then reverse
observed polarity before plotting on focal-sphere. If >
1, plot as a question mark.

STAWT The reading weight is multiplied by STAWT.

TLDLY Correction to be added to the observed time.

Used for satellite delays for USGS Alaska data.

YRMODY Year, month, and day of expiration of time-dependent

station parameters.

HRMN Hour and minute of expiration of these parameters

 2.3.2 Program specifications

The TEST variables and abbreviated definitions are printed out so that each run is well
documented. The station list and the velocity models are also printed out.

Chapter 2 - 52
  2.3.3 Vp/Vs ratio
If the Vp/Vs-value is in effect (see TEST(49), section 2.2.4) the computed Vp/Vs ratios are
printed out. See Chapter 5 for an explanation of the calculations.

 2.3.4 Iteration output

It is recommended that PRINTER 1 be used. One line will then be printed per iteration as
follows:

Heading Explanation

I Iteration step. If a particular step is repeated, I is not

incremented.

LAT Minutes of latitude

LON Minutes of longitude

DEPTH In kilometers

* If depth is constrained.

RMS Root-mean-square travel-time residual (s). See equation

below

NO Number of P, S and S-P readings used

PRMS RMS predicted for after the next step. See 4.4 for
discussion.

DAMP Value of damping constant in use. See 4.

EIGENVALUES The three eigenvalues of the spatial normal equations.

ADJUSTMENTS The adjustments in the principal directions are converted

COMPUTED into changes in latitude, longitude, and depth.
DLAT, DLON, DZ

ADJUSTMENTS TAKEN This adjustment will be taken to reach the next iterative
location. The limits imposed by the TEST variables
DLAT, DLON, DZ have been applied.

Equation for root-mean-square travel-time residual (RMS). For i phases, i = 1, N, Ri is the

observed minus computed time of the ith phase. Wi is the computed weight of the ith phase.

Chapter 2 - 53
 1
 nWi Ri2 
2

RMS   1 N 
 1 Wi 

 2.3.5 Quality

 2.3.5.1 Based on error estimates

This quality is based on the values of SEH (the horizontal 68% confidence limit in the least well-
constrained direction) and SEZ (the 68% confidence limit for depth). See Chapter 3 for
further explanation of SEH and SEZ. Note that the following limits are modified from those
used previous to April 1984 to reflect revised definitions of ERH and SEZ.

Quality Larger of SEH and

SEZ

A < 1.34

B < 2.67

C < 5.35

D > 5.35

 2.3.5.2 Quality based on many parameters

SQD-HYPO71 Quality
S is the solution quality as defined in HYPO 71:

S RMS SEH* SEZ**

A < 0.15 < 1.0 < 2.0

B < 0.30 < 2.5 < 5.0

C < 0.50 < 5.0

D Others

 SEH is the horizontal 68% confidence limit in the least-well-constrained direction.

** SEZ is 68% confidence limit for depth.
The letter "Q" is used just as a spacer in the string "SQD".

Chapter 2 - 54
D is the station distribution quality as defined in HYPO71:

D No. GAP DMIN

A >6 < 90 < DEPTH or 5 km

B >6 < 135 < 2*DEPTH or 10

km

C >6 < 180 < 50 km

D Others

DMIN is the distance to the nearest station.

Chapter 2 - 55
 2.3.6 Final summary output lines
Heading Explanation

SE OF ORIG Standard deviation of origin time.

# OF ITERATIONS Total number of iterations

DMAX Distance weighting maximum distance.

SEQUENCE NUMBER Sequence number from columns 94-98 of SUMMARY record

preceding this event.

EVENT TYPE Column 92 of SUMMARY record (see 2.4.1).

PROCESSING STATUS Column 74 of SUMMARY record (see 2.4.1).

DMAX Final value of XFAR, based on TEST(12) and TEST(46).

DATE If solution based only on S-P data, an * will follow date.

ORIGIN Hour Minute Second

LAT Degrees and minutes

LON Degrees and minutes

DEPTH Kilometers

MAG Preferred magnitude. Also entered on SUMMARY record in

columns 37-38. See 2.2.3.3, 2.2.17, and 2.4.1.

NO Number of P, S, and S-P readings used in the solution.

D1 Distance to the closest station used in the solution (km).

GAP Largest azimuthal separation in degrees between stations as

seen from the epicenter.

D Number of delay-model used (1 to 5).

RMS If the residuals are Ri and the weights are Wi, i = 1, N, then
equation for RMS is as given below.

AVWT Weights are normalized so that their sum equals NO by

dividing each weight by the average weight, AVWT.

SEH Horizontal 68% confidence-limit for the least-well-

Chapter 2 - 56
 constrained direction.

SEZ 68% confidence-limit for depth.

Q - HYPO71 Average of S and D qualities defined in 2.3.5.2. Rounded to

lower quality when necessary.

SQD S and D qualities defined in 2.3.5.2.

ADJ Length (km) of final adjustment of hypocenter.

I S-data indicator. 0 - S not used 1 - S is used

N Fixed location indicator. 0 - nothing fixed, 1 - depth fixed at

trial depth, 8 - origin time fixed at trial origin time, 9 -
location fixed at trial hypocenter

NR Total number of P, S, and S-P readings

AVR Average-weighted residual.

AAR Average of the absolute value of the weighted residuals.

NM Number of stations at which amplitude magnitude (XMAG)

was calculated.

AVXM Average XMAG.

MDXM Median XMAG.

SDXM Standard deviation of XMAG's calculated, with respect to

AVXM.

NF Number of stations at which F-P magnitude (FMAG) was

calculated.

AVFM Average FMAG.

MDFM Median FMAG

SDFM Standard deviation of FMAG's calculated, with respect to

AVFM.

VPVS Computer slope of Tp vs Ts.

Chapter 2 - 57
Equation for root-mean-square residual:

1
 nWi Ri2 
2

RMS   1 N 
 1 Wi 

 2.3.7 Detailed station output: TRAVEL TIMES AND DELAYS:

Heading Explanation.

STN Station code.

C Component.
Z - vertical.
N - north-south horizontal.
E - east-west horizontal.

PHA Phase.
Blank for P-phase.
S for S-phase.
SMP for S minus P interval.

REMK Phase remark (columns 5-8 of PHASE record).

P First-motion polarity, corrected as per station history.

PSEC Seconds of P-arrival as typed on ARRIVAL TIME record.

SSEC Seconds of S-arrival as typed on ARRIVAL TIME record.

RESID Residual (s).

If a character follows the residual the meaning is:

D = Weight reduced to zero by distance weighting.
B = Weight reduced to zero by boxcar weighting.
M = Weight reduced to zero by truncation weighting.
J = Residual is greater than 3 standard deviations from the mean. Used with Jeffrey's
weighting.
X = Weight reduced to zero during critical station run (see TEST(44)).
R = Computed weight less than 0.0005, so set to zero.
G = Beyond distance weighting cutoff but included in order to reduce gap (See
TEST(46)).

Chapter 2 - 58
 * Large residual flagged by RESIDUAL OPTION (See 2.2.3.12).

STD-ER Standard error(s) used for this arrival in hypocentral solution.

The weight assigned to each phase is proportional to the
inverse square of the standard error of the phase.

DIST Epicentral distance of station (km).

AZM Azimuth of station from epicenter (degree).

AIN Angle of ray leaving hypocenter measured with respect to

downward vertical (degree).

TC Station clock correction (s) from ARRIVAL TIME record.

Added to observed arrival time.

C Velocity model used for this travel time.

VTHK Thickness of the variable layer in km for the velocity model

used.

TTOB Travel time observed (s).

TTCAL Travel time calculated (s).

DELAY Station delay (s) for model preferred by closest station.

EDLY Elevation delay (s).

RMK Remark from columns 63-64 of PHASE record.

STN Station code.

SOURCES Phase data sources followed by number of satellite hops,

columns 105-110 of original PHASE record.

MAGNITUDE DATA:

SOURCE Amplitude source code from column 108 of ARRIVAL TIME

Chapter 2 - 59
 record.

SYS System response used in computing magnitude.

C10 XMAG calibration constant for USGS magnitude subroutine;

For UAGI magnitude subroutine, system magnification in
counts per 10-6 mm at the period (PER) of wave. Set to -1.0
if calibration data is not available.

AMX Maximum amplitude from input data, peak-to-peak (mm).

GR A1VC0 gain range state, 0 = normal; 1 = gain times 1/10;

2 = gain times 1/500.

INK Siemens playback gain, 0 = high; 1 = low (gain times1/4).

AMF Maximum amplitude (mm), corrected for gain state and

Siemens playback record.

PER Period (s) of wave where maximum amplitude was read. If PER
is not given on ARRIVAL TIME record, then standard
period from station archive is used. The default is 0.1s.

UNIT/MM Station gain at period PER.

GND MOT m Ground motion (microns), peak-to-peak.

XMGC XMAG station correction.

XMAG Amplitude magnitude. An 'e' (for Excluded) follows if

XMAGWT = 0; if not, an '*' follows XMAG if XMAG -
AVXM > 0.5.

FMP F-P time interval (s).

FMAG Coda magnitude. An S follows if the coda was too short with
respect to the S-P time for a coda magnitude to be
computed; if not, an 'e' (for Excluded) follows the FMAG if
FMAGWT = 0; if not, an * follows FMAG if FMAG -
AVFM > 0.5.

 2.3.8 Auxiliary RMS sphere output

At times there may be a concern that the final iterative earthquake location is not the best one
possible. If TEST(6) .NE. 0.0, (see 2.2.4) then the RMS residual is calculated at 14 points on
a sphere of radius = TEST(6) centered on the final hypocenter. If the hypocenter is at a

Chapter 2 - 60
 minimum of RMS in space, then all the points on the sphere will have larger RMS values
than the center point. The DRMS is the RMS on the sphere minus the RMS at the center and
will be positive for good locations.
The average DRMS values at the ends of seven diagonals through the sphere are calculated.
These are printed in order of poorest to greatest location control and are specified by their
down-dip azimuths. If TEST(6) is negative and if a point on the sphere has lower RMS than
the center of the sphere, iteration will resume at that point in order to improve the solution.
This is allowed only once per earthquake solution to prevent an infinite loop condition from
arising.
A tabulation is printed listing the number of readings used, the RMS at the center, the minimum
DRMS, the average DRMS, and a quality based upon these values as follows:

Q NUMBER RMS MIN DRMS AVE DRMS

A >6 < 0.2 > 0.3

B >5 < 0.4 > 0.15

C >4 < 0.4 > 0.5

D Others

 2.3.9 Focal mechanism plot

If the number of first motions is greater than or equal TEST(7) a focal-mechanism plot will be
made on the printer. The diagram is an equal-area projection of the lower hemisphere of the
radiation field. The symbol printed is as follows:

+ 1 or more +'s

C 1 compression

B 2 compressions

A 3 or more compressions

X Any combination of compressions and dilatations

- 1 or more -'s

D 1 dilatation

E 2 dilatations

F 3 or more dilatations

Chapter 2 - 61
 ? Indicates that although a first motion was reported,

the station polarity is uncertain.

A +, -, or ? is printed only if a compression or a dilatation does not occupy the position.

If TEST(7) < 0, a second plot will be made showing station codes on the focal sphere.
Use TEST(45) = .1379 for 8.5-inch paper and TEST(45) = .10106 for 11-inch paper.

 2.3.10 Final Tabulation

At the end of each run of a set of earthquakes, a table is printed which gives the number of
earthquakes within each quality specification. There is also a table which shows for each
station the number of times the station was used (N), the average weight (WT), the weighted
average residual (AVE), and the standard deviation of the residuals about their mean (SD).
The TABULATION option (see 2.2.3.4) controls the quality of the events included in this
table.
For P and S arrivals there are sets of two columns in the table. The first bases the weight for the
ith residual from the kth event only on the final weight prior to normalization, WTik, used in
locating the kth event. The second also includes an event-weighting factor, F, based upon the
number of arrivals and their weights used in locating each event. For the kth event:
Fk = SUMWTk * (NRWT - 4)/NRWTk, for NRWTk > 4
Fk = 0.0, for NRWTk < 4
where NRWTk is the number of readings used in locating the kth event, and
SUMWTk is the sum of the weights (WTik) of the arrivals of the kth event.
Inclusion of the factor F in the weights used to compute the average station
residual will bring the average residual into closer agreement with the
modification in station correction that would be obtained from a joint inversion of
all of the events for both location and station correction. Pavlis and Hokanson
(1985) suggested using SUMWTk for this purpose. The addition of the degree of
freedom term, (NRWT - 4)/NRWT should further improve the technique, in that
the residual pattern for an event with few degrees of freedom will be unlikely to
reflect the true station residuals. For data sets consisting of a large number of
events, iterative modification of the station corrections using this table of averages
and then relocating the earthquakes will give approximately the same results as a
joint inversion for both the locations and the station corrections (Pavlis and
Booker, 1983; Pavlis and Hokanson, 1985).
If a station is given zero weight on its STATION record (see 2.2.5), it will be
included in the tabulation even though it has not been used in any of the solutions.

Chapter 2 - 62
 In this case the station code will be proceeded by a 'W'. If a station is assigned a
P-weight-code replacement of 4-8 on the TIME-DEPENDENT STATION
parameter record (see 2.2.5.3), then the summary will include average P-residual
information even though the P arrivals were not used, and the P-residual standard
deviation will be followed by a 'P'. The S-residual standard deviation will be
followed by an 'S' in analogous situations. In either this case or the case where the
station weight is zero, the weight used in the tabulation is based on the P-weight-
code on the ARRIVAL TIME record (See 2.2.6).
WARNING: The station weight and the P- and S-weight-replacement codes are
time-dependent parameters specified in the station list. The final tabulation
assumes, however, that the station weight for a given station was either zero or
non-zero during the entire run, and similarly, that the weight-code replacements
either were or were not within the range 4-8 during the entire run. If these
assumptions are not correct, the tabulation will be in error.

 2.4 Summary record output

The SUMMARY OPTION record described in 2.2.3.2 controls the SUMMARY record output.
The STATION records are generated in the same format as the input STATION records. The
other formats are given below.

 2.4.l Summary record

To save space no decimal points are used. Use the FORTRAN format for reading the summary
record given below.

Item Column Nos. Format for

Reading

Origin Time:

KDATE - year, month, day (e.g. 19981231) 1 8 i8

KHRMN - hour, minute (e.g. 2358) 9 12 i4

KSEC - (seconds) 13 16 F4.2

LAT (degrees) 17 18 i2

N or S 19 A1

LAT (minutes) 20 23 F4.2

LON (degrees) 24 26 i3

Chapter 2 - 63
E or W 27 A1

LON (minutes) 28 31 F4.2

DEPTH (km) [If negative, reset to -00] 32 36 F5.2

PREFERRED MAGNITUDE 37 38 F2.1

NO - Number of P, S, and S-P readings used in the 39 41 i3

solution

GAP - Largest azimuthal separation in degrees between 42 44 i3

stations as seen from the epicenter (deg.)

D1 - Distance to closest station used in solution (km) 45 47 F3.0

RMS (s) 48 51 F4.2

Azimuth of axis 1 of error ellipsoid (deg) 52 54 i3

Dip of axis 1 (deg) 55 56 i2

SE - length of ellipsoid semi-axis 1 (km) 57 60 F4.2

Azimuth of axis 2 of error ellipsoid (deg) 61 63 i3

Dip of axis 2 (deg) 64 65 i2

SE - length of ellipsoid semi-axis 2 (km) 66 69 F4.2

Average XMAG 70 71 F2.1

Average FMAG 72 73 F2.1

Processing state (not used by HYPOELLIPSE) 74 A1

Chapter 2 - 64
 * - More data available to be added
P - Preliminary, but location not finalized
F - Final location determine
G - National Earthquake Information Center (NEIC) solution
A - NEIC solution obtained from USGS/UAGI
N - Not of principal interest
I - Insufficient data to determine a hypocenter

SE - length of ellipsoid semi-axis 3 (km) 75 78 F4.2

Quality - either error-ellipsoid quality or HYPO quality 79 A1

depending upon QUALITY OPTION record. (See
2.2.3.4) [In reformatted NEIC data this column
contains the depth quality indicator.]

MAGTYP - F, X, A, or K to indicate which type of 80 A1

magnitude is entered in columns 37-38. (See 2.2.3.3
and 2.2.17)

NSWT - Number of S-phase arrivals used in solution. 81 82 i2

/ or \ 83 A1
The primary SUMMARY record is always first and
has a "/" in column 83. If an archive file has more
than one SUMMARY record, the second and any
subsequent records will have a "\" in column 83.

First 4 characters of INSTRUCTION record 84 87 A4

Month earthquake was run 88 89 i2

Year earthquake was run 90 91 i2

Event type 92 A1

E or local or regional earthquake S artificial source such as seismic line or shot

blank

T teleseism O other non-earthquake (e.g. sonic boom or

lightning)

R regional (poor coverage; use C calibration signal

Chapter 2 - 65
 solution from another
organization)

N nuclear explosion A volcano tectonic (VT)

G glacial event B volcano long-period (LP)

Q quarry or mine explosion X emergent, low frequency near volcano,

F false trigger V volcano tremor burst or eruption

I Augustine volcano shore-ice H volcano VT-LP hybrid

event

+ continuation of previous
event

For type T, R, or N, do not compute the hypocenter location, but instead compute the azimuth
and apparent velocity across the network.

Once the event type has been placed on the SUMMARY record of an earthquake in archive
format, it will be transferred to succeeding SUMMARY records generated by later runs of
HYPOELLIPSE.

Fixed location indicator, from column 19 of 93 i1

INSTRUCTION record or imposed by SELECT
DELAY option (2.2.3.6)

Sequence number 94 98 A5

S-P time at closest station used in solution. Blank if 99 102 F4.2

either P or S is not used. Set to 9999 if S-P .GE. 100.

ZUP - Computed with GLOBAL OPTION. 103 104 F2.0

ZDN - Computed with GLOBAL OPTION. 105 106 F2.0

Vp/Vs - Computed slope of Ts vs Tp. Only computed if 107 110 F4.2

TEST(49) is not equal 0.

Number of readings weighted out due to Jeffrey's, 111 112 i2

truncation, or boxcar weighting.

DEPTH (km) [Allowed to be negative] 113 117 F5.2

Chapter 2 - 66
  2.4.2 Phase records in input format with corrected arrival times
This option will create a "perfect" set of data, which then may be used to check the
HYPOELLIPSE program. For example, one might want to know how well the program
would work on events in some particular region. Fixed solutions specifying this
epicentral region could be run with SUMMARY OPTION 5 and test earthquakes would be
generated. The "perfect" data will be generated as follows:

Item Column Nos. Format for Reading

Station code 1 4 A4

KDATE 10 15 i6

KHRMN 16 19 i4

SPEC - P-arrival time 20 24 F5.2

SSEC - S-arrival time 32 36 F5.2

Chapter 2 - 67