2003 Herring2003 Article MATLABToolsForViewingGPSVelociy
2003 Herring2003 Article MATLABToolsForViewingGPSVelociy
2003 Herring2003 Article MATLABToolsForViewingGPSVelociy
length of the arrowheads is specified in millimeters, which program. If the time series files are not in the same directory
is useful for judging the size of very small velocities. The from which velview was run, the name of the directory can
number of sites being displayed for each of the velocity be specified in the box next to the ‘‘Time Series’’ label at the
fields is shown at the top of the figure. top of the figure. Tsview is discussed below.
Once the velocities fields are loaded, there are options for Another feature of velview is the ability to select a transect
viewing and manipulating the fields. There are two types of across the velocity field for profiling where the velocities are
zoom available. The ‘‘Vec Zoomin’’ feature allows a click- shown as a function of distance along the profile and are
and-drag selection of an area to be viewed in detail while projected into directions based on the azimuth given in the
the vectors maintain their original scaling. The ‘‘Zoom In’’ box next to the ‘‘AZ’’ button. The azimuth of the projection
feature rescales the vectors during the zoom so that their can be either typed into box, graphically selected by clicking
lengths increase with zooming (this is the standard zoom the ‘‘AZ’’ button when the box is empty, or default to along
feature in Matlab). Both zoom features maintain a stack of the profile direction and normal to the profile direction. The
previous zooms in the zoom out buttons so that previous profile is selected by first drawing the profile centerline and
views can be recovered by successive zoom outs. (If both then by graphical selecting its width. The selected profile
types of zooms are used, the zoom out feature should be box is displayed on the figure and the sites inside the box are
used in the reverse order of the zoom ins.) At anytime, marked with yellow circles. The profile box information is
numerical values can be entered into the arrow length placed on a stack as profiles are selected. The boxes and site
boxes, and the redraw button used to return the vectors to markings on the figure can be removed from the top of the
well-defined lengths. The ‘‘Pan’’ box arrows and minus stack using the ‘‘POP’’ button (most recent removed) or
sign can be used to move the view by half the current from the bottom of the stack (earliest removed) using the
width and zoom out by 50% of the current size. ‘‘PIP’’ button. Only results from the primary velocity field
Manipulation of the velocity fields includes the ability to are shown on the profiles. (The ‘‘SWAP’’ button at the bot-
apply offsets to the motions either by typing the values into tom of the velview window can be used to exchange the
the box next to ‘‘Offset’’ or, when this box is empty, to click- velocity fields thus allowing profiles for each field to be
and-drag a region in which the weighted mean velocities will plotted in separate windows). An example of the profile plot
be zero after the offsets are applied. The offsets are com- is shown in Fig. 2. Part (a) of the figure shows the velview
puted from the primary velocity field and applied to both window used to select the profile and part (b) shows the
primary and secondary (if being displayed) fields. A dif- profile figure itself. For this figure, the azimuth for the
ferential offset between the two velocity fields can be cal- projection of the velocities along the profile was graphically
culated from a click-and-drag region or specified directly in selected to be parallel to the CICE velocity vector.
the box next to the ‘‘Align’’ button. Both the offset and align In Fig. 2b, we have used the ‘‘Fitting Functions’’ window to fit
options display the results in boxes on the figure and as text a function to the profile velocities based on the co-seismic
in the Matlab or terminal window. The boxes on the figure strain accumulation for a locked strike-slip fault (the SS
can be removed by clicking on them. The offsets are applied function) (Okada 1985). When velview is run in Matlab, any
as simple additions to the velocity components. For large valid Matlab function, defined to be a function of distance
areas, these offsets cannot adequately represent the effects of along the profile, can be used in the fitting functions. In the
rotations and translations between two reference frames. In executable version, only specific functions can be used with
this case, the GLOBK utility programs cvframe and velrot or the pre-defined SS function being available in both the
the userÕs own tools should be used to define reference Matlab and executable versions. The green shaded boxes on
frames and to align velocity fields in different frames. the left of the figure show the results of the fit in the direction
Other manipulations available in velview include the removal of the azimuth (labeled ‘‘along track’’), normal to this
of duplicate sites with exactly the same velocities and sigmas direction (labeled ‘‘cross track’’) and for the vertical velocity.
and within the specified distance of each other (given in the Because space is limited in these text boxes, the units are not
box next to the ‘‘RmDups’’ and ‘‘Stats’’ boxes). The removal given. For the weighted root mean square (WRMS) scatter
of duplicates is useful for GLOBK analyses where sites over and the amplitudes of the estimated coefficients (FCN0 and
some intervals of time may have been renamed because of FCN1 in the figure), they are in mm/year. The FCN0 value is
possible changes in position, but the velocities of the multiple the offset of the profile velocities. The normalized root-
site names have been forced to be equal. The ‘‘Stats’’ button mean-square (NRMS) scatter is dimensionless and is the
computes the weighted root mean square (WRMS) differ- ratio of the scatter to the expected scatter based on the
ences in velocities between sites within each velocity field velocity error bars (square root of v2-per-degree-of-free-
that are separated by less than the distance given in the dom). A single fault with a deep locking depth has been used
box next to the button. The ‘‘ID Points’’ list all sites in a to fit the profile in the example but a better fit can be obtained
click-and-drag region and is useful when multiple sites are by fitting to multiple faults with shallower locking depths.
overlaid on each other and clicking on a specific vector is Individual sites may be deleted before the fitting functions
difficult. The list of sites appears both in a box at the top of are estimated and they appear with a cross over the point.
figure (removed by clicking on it) and in the terminal window. The results of the fit and the profile information can be
The velocity vectors plotted by velview can be selected and saved to a file by putting the file name in the yellow ‘‘File
the grey popup menu (example shown in Fig. 1) can be used box’’ and clicking the save button. Information may be
to obtain information about the vector. If position time extracted from this file so that it can be plotted in another
series are available, they can be plotted using the tsview graphics program if desired.
Fig. 2 b the time series and the statistical properties of the time series
(Contd.) residuals with either white or correlated noise assumptions.
If the ‘‘RealSigma’’ (meaning realistic sigma) button is
series of this type can be shown as single time series in unchecked, a white noise model is used. The white noise
tsview. Once a time series is loaded, there are numerous assumption almost invariably generates very optimistic
options for manipulating the time series. estimates of the uncertainties of the parameter estimates.
The primary aims of tsview are to assess the quality of time When the RealSigma box is checked, a time-correlated noise
series and to generate files that can be used to control the model is used to estimate the parameter uncertainties. With
treatment of data from specific sites at specific times when this model, a correlation time of the residuals for each
data are combined in GLOBK. The center set of buttons in the coordinate component is estimated by computing the in-
tsview window control editing and fitting of functions to the crease in the chi-squared-per-degree of freedom of succes-
time series. The function boxes in the lower left hand part of sively longer time averages of the residuals. For a white noise
the window control mainly the output files from tsview. error model, the chi-squared-per-degree of freedom would
Breaks in time series can be treated in multiple ways in not depend on averaging time. With temporal correlations
tsview. The break can be a simple offset in the times series in the time series, chi-squared-per-degree of freedom
or it can be an offset followed by either an exponential or increases as the residuals are averaged over successively
logarithmic functions to represent the time evolution after longer time intervals. The character of the averaged residual
the break. When exponential and logarithmic functions can be viewed in tsview using the ‘‘Average’’ button. The
are used, a change in the secular rate can also be estimated value in the box below the button gives the length of time in
after the break. The main use of these different function days that will be averaged. The statistics of the averaged
forms is to study the behavior of time series after earth- residuals are shown in bold on the display. For the example
quakes. Annual and semiannual components in the time shown in Fig. 3, the realistic sigma estimates of the secular
series can also be estimated. By selecting the ‘‘Linear Only velocity in each coordinate of this site are 3 to 5 times larger
Residual’’ option, the residuals to the fit can be displayed than the white noise estimates. Thirty-day averages of the
with only the linear motions and offsets removed. In this coordinate residuals have scatters of 0.4 mm in the
mode, a thick black line shows the time variations of the horizontal components and 0.8 mm for the vertical
estimated model parameters and the red model uncer- component. These scatters are only a factor 2 smaller for
tainty lines are placed plus and minus one standard horizontal coordinates and a factor 4 smaller for the vertical
deviation about this black line. coordinate than the scatters of daily values. If the residuals
The uncertainties of the parameter estimates in tsview are were white noise, the scatters should have reduced by a
calculated using the sigmas of the coordinate estimates in factor of 5.5 (square root of 30).
Fig. 3 they are used.) The edit file from tsview can be used in
Example of a tsview screen image. The time series shown is for the site GLOBK to remove edited sites from an analysis and to
DHLG which is one of the many sites in Southern California affected
by the October 1999 Hector Mine earthquake (Hudnut et al. 2002). In implement discontinuities in time series through renaming
the GLOBK processing, this site is divided into a pre-Hector mine sites. The lowest box in this part of the figure allows an output
name (DHLG_GLA) and a post-Hector mine name (DHLG_GHT). file from the GLOBK ensum program to be used to sort the
The plot shows positions from both site names appended together list of sites based on the statistics of their fits to linear mo-
with a break in late 1999 (1999/10/16). A logarithmic function is used
to model the behavior of the site position after the earthquake. The
tions with the poorest fit sites appearing at the top of the list.
green horizontal lines show the bounds of 3-times the WRMS scatter
of the residuals. The red, nearly horizontal lines near zero on the plots
show the 1-r limits of the model time series based on the standard
deviations of the parameter estimates with the contribution from the References
overall mean value of the time series removed. (When only a rate is
estimated, this display forms a thin ‘‘x’’ cross shape that crosses zero
Hudnut KW, King NE, Galetzka JE, Stark KF, Behr JA, Aspiotes A,
near the center of the plot.) The full set of parameter estimates and
van Wyk S, Moffitt R, Dockter S, Wyatt F (2002) Continuous
their standard deviations are written to the Matlab workspace and to a
GPS observations of postseismic deformation following the 16
pop-up window in the upper right hand corner of the computer October 1999 Hector Mine, California, earthquake (Mw 7.1),
screen. In this figure we have used the ‘‘Realistic Sigma’’ model Bull Seis Soc Am vol 92, no 4, pp 1403–1422
discussed in the text
King RW (2002) Documentation for the GAMIT GPS analysis
software, MIT Internal Report, 206 pp (http://www-gpsg.mit.-
For users of GAMIT/GLOBK, the results from tsview can be edu/simon/gtgk/GAMIT.pdf)
used in later GLOBK analyses of the data set. The ‘‘Save’’ King RW, Herring TA (2002) Global Kalman filter VLBI and GPS
button in center column will save the results from the time analysis program, MIT Internal Report, 98 pp (http://www-
gpsg.mit.edu/simon/gtgk/GLOBK.pdf)
series being displayed in the files whose names are given in
Okada Y (1985) Surface deformation due to shear and tensile
the bottom part of the left-hand column. (The small W faults in a half space. Bull Seismol Soc Am 75:1135–1154
buttons show the files to be written, and the E buttons Watkins MM, Bock Y, Hudnut KW, Prescott WH (1997) The
indicate which files will be erased before they are written. The Southern California Integrated GPS Network: Status Report, Eos
erase buttons reset automatically to non-erase mode once Trans AGU, AGU Spring Meeting, 29 April 1997, p 105