459
1 INTRODUCTION
RTK(RealTimeKinematic)methodforpositioningis
used in daily life by different consumers for many
purposes.Whentherearesomanymeasurements,it
is essential to know where RTK measurements are
concentrated and which stations are obligatory for
LitPOS performance. LitPOS is a geodetic
infrastructure for
referencing spatial geoinformation,
also it provides the direct linkage to the National
CoordinateSystemandheightdatum.
LitPOS network data processing was done using
Bernese5.2software.Softwarewasslightlyimproved
byeditingscriptsandwritingproceduresforRINEX
files preparation, for downloading other necessary
data, and for getting two
processing solutions from
singleBPE(Berneseprocessingengine)process.
Two different calculation programs were used to
analyzethetimeseriesofcoordinates,whilethedata
received were displayed with the help of the GMT
program.
2 LITPOSNETWORKSTRUCTURE
Surface deformation in Europe can be studied at
relatively large scales by
modern GNSS geodesy if
velocitydatafromahomogeneousanddensenetwork
ofhigh qualitystations isavailablefor a sufficiently
long lapse of time. LitPOS (Lithuanian Positioning
System), is the global position determination system
ofpermanentreferencecontinuouslyoperatingGNSS
stations,becameoperationalinJuly2007(Baniuliset
al.,
2017).LitPOS combines anetwork ofbase GNSS
stationswithdedicatedcommunicationchannelsand
appropriate hardware and software. Its main
objective is to provide the 24/7/365real time precise
positioning service with national‐wide coverage for
On Analysis of LitPOS Stations Time Series and
Velocities
V.Puškorius,E.Paršeliūnas,R.Baniulis,K.Galinauskas&S.Valotka
VilniusGediminasTechnicalUniversity,Vilnius,Lithuania
ABSTRACT: LitPOS (Lithuanian Positioning System), the network of continuously operating GNSS (Global
NavigationSatelliteSystem)stations,becameoperationalinJuly2007.Itprovidesdatabothforreal‐timeand
post‐processing applications.LitPOS stations cover the whole territory of Lithuania. There are 31of LitPOS
GNSSstationsintotalnetwork.AlsoLitPOSnetworkincludes3ASGEUPOS(Poland)and6LATPOS(Latvia)
stations. LitPOS network data processing and analysis of station’s velocities was done using Bernese 5.2,
TSviewandCATREFprograms.
ResultsofanalysisshowthatNorth/EastvelocitiesoftheLitPOSstationsareveryhomogeneouswhat
leadsto
the conclusion that there are no any significant horizontal deformations of Earth’s crust in the territory of
Lithuania.UpcomponentsofLitPOSstationsvelocitiesfitverywelltoverticalmovementsofEarth’scrustas
describedbythelandupliftmodelsofFennoscandia.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 14
Number 2
June 2020
DOI:10.12716/1001.14.02.25
460
thegeodetic,cartographic,cadastralandconstruction
applications. This network is part of National
GeodeticFoundationinfrastructure.Itallowsusersto
evaluate their position by using RTK (Real‐Time
Kinematic), transmitting corrections through
communication channels from points of geodetic
basis. It uses VRS (Virtual Reference Station) and
Ntrip (Networked Transport of
RTCM via Internet
Protocol) corrections methods. LitPOS provides to
usersa possibility toreach centimeter‐levelaccuracy
with single standard rover receiver at any point of
Lithuanian territory. LitPOS network consists of 31
station, located evenly in country’s territory, and
regional management center. The stations are
equipped with Trimble NetR9
GNSS receivers and
Choke ring antennas. Overall running and
management of LitPOS network is executed by
TrimblePivotPlatformsoftware.
Figure1.LitPOSnetworkmap
Bigger part of GPS (Global positioning system)
stations are installed in collaboration with Fire and
RescueDepartmentundertheMinistryoftheInterior
of theRepublic of Lithuania.They are setup in fire
towers. Average distance in between stations is
around50km.Additionally9stationsfromabroadare
incorporatedin
LitPOSnetwork(3fromPolandand6
–Latvian).
LitPOS stations become very important geodetic
points having the combined set of geodetic
parameters:
Coordinates to LitPOS stations are transferred
fromNationalZeroOrderGPSNetworkandEPN
(Europeanpermanentnetwork)stations.
Geopotential heights and normal heights of
National
FirstOrderVerticalNetworkareusedfor
datatransfertoLitPOSstations.
Gravity values of National Zero and First Order
GravimmetricNetworkareused for datatransfer
toLitPOSstations.
HeighttransferfromNationalVerticalFirstOrder
Network by precise levelling to GNSS station
benchmark,andheighttransferfrom
ittoantenna
pierbytrigonometriclevelling(usingtotalstation).
ServicesandProductsprovidedbyLitPOS:
Real‐time services: RTK (Real‐Time Kinematic)
usingVRS(VirtualReferenceStation)technology;
andreal‐timeDGPSservice;
Post‐processing products: RINEX data files
forfurtherprocessing.
Data streams are transferred in real
time from
GNSS stations to regional management center in
separatedatatransferlines.LitPOSnetworkallowsto
determinepositionof aparticular objectinreal time
byprovidingcorrectionsviacommunicationchannels
from fixed points of the network. Users receive
geodetic corrections via GSM and GPRS channels.
LitPOS services are provided
in RTK (+–2cm
accuracy), DGPS (+–0.3m to +–0.5m accuracy), and
GPPS(upto1mmaccuracy)methods.Correctionsto
geodetic coordinates are transferred in RTCM 2.1,
RTCM2.3,RTCM3.1,RTCM3.2,CMR,CMR+,CMRx
irDGPSRTCM2.1,DGPSRTCM2.3,DGPSRTCM2.4
formats.InordertoreceiveLitPOS
corrections,users
have to be registered in LitPOS website and receive
username and password. LitPOS uses GPS and
GLONASSsatellitesystemsfornetworksolution.
Figure2.CoverageofLitPOSstations.
Thestationcoversaround35kmradiusaroundit
(Fig.2),thusthecoverageofthewholecountrycanbe
observable. Covered are by all LitPOS stations,
situated in Lithuania. Not in the distant future, it is
plannedtoinstallupto5morestationsintheterritory
ofLithuania.
Main
characteristics:
Software:BSW5.2update20160108
Network: 30+6+3 LitPOS stations +10 IGS/EPN
fiducialstations
GNSS:GPS
AntennasPCV:absoluteandindividualcalibration
Preciseorbits,etc.:CODE
Troposphericrefraction:VMF
Oceantidemodelmodel:FES2004
Baselinesprocessingstrategy:OBS‐MAX
Ambiquities
resolutionstrategy:QIF
ITRF realisation: IGb08
(EPN_A_IGb08_C1845.SNX)(ITRF2014‐GPSweek
1934(29January2017)
Cut‐offangle:3,10,25
Period:2008‐2014;2015‐2017;2017‐
Products:DailyandweeklySNX(NEQ+COV)
461
Coordinates Time Series: analysis by TSview and
KATREF
3 LITPOSDATAPROCESSING
LitPOS network data re‐processing was done using
Bernese (BSW5.2 update 2016 01 08) software and
applyingprocesscontrolfileNKG_R2Sall.PCF,which
wassetupusing“GuidelinesforNKGGNSSACRe‐
processing No. 1”. Since July
2007, LitPOS provides
data for real‐time and post‐processing applications.
DataarestoredtoRINEXfilesfromreal‐timestreams
using Trimble GPSNet software and TPP (Trimble
PivotPlatform)atLitPOSpermanentnetworkservers.
TheformatofLitPOSnetworkstationsdataisRINEX
2.11. The hourly files with 1‐
second observations
interval were stored. Before starting 2007–2018 re‐
processing of LitPOS network data, RINEX files are
combinedintoonedailyRINEXfile(intervalbetween
observations was set to 30 seconds) using teqc.exe
programandspecialbatchfilecode.Thecombiningof
one‐monthdataof30stationstakesabout10
hoursof
computingtime(Baniulisetal.,2017).
Otherdatalikeasatelliteclockfile(*.CLK),Earth
rotation parameter file (*.ERP), ionosphere file
(*.ION), ephemeris file (*.EPH), Differential Code
Biases corrections file (*.DCB), Vienna Mapping
Function and RINEX files of fiducial stations
necessary for reprocessing were obtained from ftp
servers
using batch file code and wget.exe program
and were putted into appropriate directories.
DownloadingofRINEXfilesof10fiducialstationsof
oneyeartakesabout4hours,andothernecessaryfiles
downloadtakesabout2–3hoursofcomputingtime.
Some batch files have been coded to gathered
mandatory necessary files
in automated way, using
wget.exeprogram(Zurutuzaetal.,2008).AlsoRINEX
files preparation was carried out by using teqc.exe
program and batch file code. During LitPOS
operationalperiodlargeamountofdatawasgathered
what gives a possibility to construct reliable
coordinates time series. Coordinates time series
analysis is
an important part of Geodesy and
Geodynamics studies, especially when continuous
GPS observations are used to explore very low rate
deformations. In this domain, having precise and
robust tools for processing and analyzing position
time series is a prerequisite (Bruyninx et al., 2018).
HavingGNSSpositionlongtimeseries,itis
possible
toanalyzethemsearchingfordiscontinuities,velocity
changes, and outliers. These are caused by various
reasonsandrepresentedintimeserieslikegapsand
offsets (jumps). Especially in the Nordic countries,
snow typically accumulates on the radomes during
wintertime causing systematic outliers (snow peaks)
in time series. The removal
of the snow and other
short and deviating periods were marked manually
with the help of the Tsview tool (Herring, &
McClusky, 2019) and cut off from the daily SNX
solution before velocity. In Table 1 we showed how
muchepochsweshouldtorejectedfromcalculations
togetsmoothtime
seriesandvelocities.
Table1. The difference between the number of epochs
betweentheprimarydataandtheprocesseddata.
_______________________________________________
Station Numberofepochs Epochsaftercutoff
_______________________________________________
ALYT 40093856
BIRZ38983433
DIDZ16931378
DKST37703321
ELEK37553627
JNSK39073661
KAUN 40513868
KEDN 40263731
KELM 39633555
KLAI15831526
KRTN 39813734
MAZK 39663650
MRJM 32332988
NIDA 13731266
PNVZ 40493790
RKSK40563763
RTVS40293813
SAKI37093613
SAUL 39103729
SILT39993764
SLCN 40133858
SVNL 39533906
TAUR 40643670
TELS40323891
UKMG 17091630
UTEN 40353804
VARN 37823607
VEIS37643422
VGTU 38103350
VLNS 37863397
VSTT14601391
_______________________________________________
InGNSSobservationsdatapre‐processingmaybe
required for metadata correction, time windowing,
data splitting or combining and other filtering,
performqualitycheckspriortore‐processingthedata
(Altamimi, Collilieux, Legrand, Garayt, & Boucher,
2007). The re‐processing was executed with the
followingparameters:
Cut‐off
angle:3°,10°,25°;
Troposphericrefraction:VMF;
Preciseorbits,etc.:CODE;
GNSS:onlyGPS;
Ambiguityresolutionstrategy:QIF;
Baselinesprocessingstrategy:OBS‐MAX;
ITRFrealization:
EPN_A_IGb08_C1845.SNX; IGS14 from 1934 GPS
week;
Antennas PCV: absolute and individual
calibration;
Fiducial stations used in re
‐processing: BOR1,
RIGA,ONSA,MDVJ,MAR6,WTZR,VLNS,SWKI,
VIS0,SVTL.
Beforestarting2007–2019re‐processingofLitPOS
network data, RINEX files are combined into one
dailyRINEX file (intervalbetween observations was
setto30seconds)usingteqc.exeprogramandspecial
batchfilecode.Thecombiningofone‐month
dataof
30 stations takes about 10 hours of computing time.
Other data like a satellite clock file (*.CLK), Earth
rotation parameter file (*.ERP), ionosphere file
(*.ION), ephemeris file (*.EPH), Differential Code
Biases corrections file (*.DCB), Vienna Mapping
Function and RINEX files of fiducial stations
necessary for reprocessing were obtained from
ftp
462
servers using batch file code and wget.exe program
andwereputtedintoappropriatedirectories(Lutz,&
Brockmann,2018).DownloadingofRINEXfilesof10
fiducial stations for a year takes about 4 hours, and
othernecessaryfilesdownloadtakesabout2–3hours
ofcomputingtime.
Thepreliminaryanalysisoftime
seriesismadeby
Tsview. Tsview is a MatLab‐script/program for
analysis, editing, and velocity estimation of time
series of daily position estimates from
GAMIT/GLOBK. It provides a graphical user
interphase and it is designed for manual interactive
study of the time series.This program iscalculating
stationvelocitiesusing
stationcoordinatefiles(*CRD)
fromre‐procesedSINEX fileswith Bernesesoftware.
For time series analysis with Tsview we set some
parameters: Outliers (n‐sigma) was set to 3 (an
automatic outlier rejection criterion used in an
iterative detrending); 3 more options:RealSigma,
+Semiannual and +Annual was set; later Detrend
function was used to estimate the velocities with
currentsettings.Ifitisnecessary,weusedBlockedit
functiontoremoveobservationsfromcertainperiod,
for all coordinate components. Function Break is
used to add discontinuities to the time series
(Herring,&McClunky,2009).TheSavebuttonsaves
the results
to the folder that is set in the Tsview
shortcutproperties.Newfilescontainsinformationof
the data editing, breaks, station velocities in 3 axis
anderrors.
Figure4.LitPOSnetworkstationcoordinateschangebased
ondaytimemeasurements.
4 VELOCITIESESTIMATION
The station velocities can be estimated by either
combiningthedailysolutions(withfull covariances)
usingleastsquaresmethodsorestimatingtrendsfrom
the station position time series individually foreach
component. CATREF software has been originally
designedforITRFcombinationsanditiswelladapted
for time
series combination. The combination model
based on the seven parameter similarity
transformation between the frames (Stanaway et al.,
2015).CATREFconsistsofthemainprogramcombins
and several supportive programs for handling the
SNXs
DuringLitPOSoperationalperiodlargeamountof
data was gathered what gives a possibility to
construct
reliablecoordinatestimeseries.Coordinates
time series analysis is animportant part of Geodesy
and Geodynamics studies, especially when
continuousGPSobservationsareusedtoexplorevery
lowratedeformations.Inthisdomain,havingprecise
and robust tools for processing and analyzing
position time series is a prerequisite. Having long
time
series of station coordinates, it is possible to
analyze them searching for discontinuities, velocity
changes,andoutliers(Pospišil,&Hefty,2014).These
arecausedbyvariousreasonsandrepresentedintime
serieslikegapsandoffsets(jumps).Mainreasonsfor
outliers and discontinuities are equipment change,
environmentalconditionslikesnow
coverageorbirds
onantennaradome,earthquakes.
Input data – daily sinex files (10 degree cut off
angle)withNEQ(Normalequation)obtainedfrom
BERNESEusingNKGguidelines.
Timespanofreprocessing2007‐2018.
VelocitieswerecalculatedusingCATREFsoftware
packageandTsviewprogram.
Objective: To estimate
LitPOS network station’s
coordinates horizontal and vertical velocities in
ITRF2014referenceframe;
Figure5.North/EastvelocitiescalculatedwithCATREF
463
Figure6.UpcomponentvelocitiescalculatedwithCATREF
5 DISCUSSIONANDCONCLUSIONS
The inconsistencies of different types were detected
and removed from the LitPOS data to process the
final solutions of LitPOS network based on stations
absolute antenna models, 3°, 10° and 25° cut‐off
angles by Bernese 5.2 software and applying the
initial parameters from NKG (The Nordic
Geodetic
Commission)guidelines.
WeusedtheCATREFsoftwareforthecumulative
combination. The combination model is based on
sevenparameterssimilaritytransformation,andeach
ofthedaily solution istransformed tothecombined
solutionatthefixedepoch.Theresultsobtainedwere
comparedagainsttheresultsofneighboringcountries
obtained
by international organizations (Kenyeres,
2018;Lutz,2018;Zurutuzaetal,2019;Lahtinenetal,
2018, 2019). It could be stated that North/East
velocities of the LitPOS stations are very
homogeneouswhatleadstotheconclusionthatthere
are no any significant horizontal deformations of
Earth’s crust in the territory of Lithuania.
Up
componentsofLitPOSstationsvelocitiesfitverywell
toverticalmovementsofEarth’scrustasdescribedby
thelandupliftmodelsofFennoscandia(Vestøl2006;
2018).
REFERENCES
Altamimi, Z., Collileux, X., Legrand, J., Garyat, B., &
Boucher, C. (2007). ITRF2005: A new release of the
InternationalTerrestrialReferenceFramebasedontime
series of station positions and Earth Orientation
Parameters.JournalofGeophysicalresearche,112, B09401.
doi:10.1029/2007JB004949,2007.
Pospišil,L.,&Hefty,J.(2014).Geodynamicallyactiveareas
in
Central Europe determinated on the bases of GPS
measurements: kinematic models. Acta Montanistica
Slovaca, 19(3), 126‐140.
https://www.researchgate.net/publication/279317049.
Stanaway,R.,Roberts, C., Rizos, Ch., & Crook, Ch. (2015).
DefiningaLocalReferenceFrameUsingaPlateMotion
Model and Deformation Model. In: van Dam T. (eds)
REFAG 2014. International Association
of Geodesy
Symposia, 146, 147‐154. Springer, Cham. DOI:
10.1007/1345_2015_147.
Zurutuza, J. et al. (2019). Present Day Geokinematics of
Central Europe Part 1: the Central European GNSS
ResearchNetwork(CEGRN).JournalofGeodynamics.(in
press)
Lutz,S.;Brockmann,E.(2018).StatusreportWG“EUdense
velocities”. In EUREF 2018 Symposium,
Amsterdam.
(Draft).
Bruyninx C., Altamimi Z., Caporali A., Kenyeres A.,
LidbergM.,StanglG.,TorresJ.A.(2018).Guidelinesfor
EUREF Densifications. http://epncb.oma.be/
_documentation/guidelines/Guidelines_for_EUREF_Den
sifications.pdf(Visitedon2019/04/05).
Lahtinen, S., Jivall, L., & Häkli, P. (2019). ITRF2014
Densification for the Nordic and Baltic Countries.
(Draft).
Herring, T., McClusky, S. (2019) GAMIT/
GLOBK Matlab
Tools. http://www.gpsg
.mit.edu/~tah/GGMatlab/#_tsview. (Visited on
2019/03/20).
Kenyeres,A.etal.(2018)EPNDensificationWG:whereto
go?EUREF2018symposium.Amsterdam.(Draft).
Baniulis, R., Galinauskas, K., Marozas, L., Paršeliūnas, E.,
Petniūnas,M.,&Puškorius,V.(2017).Ananalysisofthe
performance and coordinates times series of
CORS
network LitPOS. In Baltic Geomatic Congress 2017,
Gdansk,Poland.DOI:10.1109/BGC.Geomatics.2017.39.
LahtinenS,HäkliP,JivallL.,etal.(2018)Firstresultsofthe
Nordic and Baltic GNSS Analysis Centre. J Geod Sci
8(1):34‐42.https://doi.org/10.1515/jogs‐2018‐0005.
Vestøl,O.(2006)Determinationofpostglaciallandupliftin
Fennoscandiafromleveling,
tide‐gaugesandcontinuous
GPSstationsusingleastsquarescollocation,JGeodesy,
80,248–258.DOI60610.1007/s00190‐006‐0063‐7.
Vestøl, O.; Ågren, J.; Holger, S.; Kierulf, H.; Tarasov, L.
(2018) NKG2016LU‐A new land uplift model for
FennoscandiaandtheBalticRegion(draft).
