DOI 10.

1590/s1982-21702024000100008

SPECIAL SECTION - Brazilian Colloquiums on Geodetic Sciences

Evaluation of precise point positioning and post-processing

kinematic methods for tide measurement in hydrographic surveys
Felipe Rodrigues Santana1 - ORCID: 0000-0001-6410-7999
Claudia Pereira Krueger1 - ORCID: 0000-0002-4839-1317
Érica Santos Matos Baluta1 - ORCID: 0000-0003-4423-272X
Tulio Alves Santana1 - ORCID: 0000-0002-4429-9409
Kaue de Moraes Vestena1 - ORCID: 0000-0003-1225-2371
1 Federal University of Paraná, Geodetic Sciences Postgraduate Program, Curitiba-Paraná, Brazil.
E-mail: [email protected], [email protected], [email protected],
[email protected], [email protected]

Received in 31st July 2023.

Accepted in 14th March 2024.

Abstract:
The use of the ellipsoidal heights of a vessel for measuring the tide allows the diminution of uncertainties of sounding
reduction in hydrographic surveys. In this article, the Precise Point Positioning (PPP) and Post-Processing Kinematic
(PPK) methods were evaluated by comparing them with data from a tide gauge station while the boat remained
moored nearby. When traversing sounding lines, the PPK solution was used as a reference to validate the PPP
solution. In all, 12 different periods of surveys were analyzed, distributed throughout the year 2021, in the Guanabara
and Ilha Grande Bays, in Rio de Janeiro. The results showed that in comparison with the tide station, when the vessel
was moored, the PPK was able to fulfill the strictest criteria (Exclusive Order) of the IHO (International Hydrographic
Organization) in 100% of the surveys, while the PPP in 50%. Regarding the use of the PPK as a reference, with the
boat both moored and sailing, the PPP met these criteria in 67% for Exclusive Order, 25% for Order 1A and 8% for
the 1B. Reducing the vertical uncertainties of hydrographic surveys in shallow waters is useful for all marine and
coastal vertical positioning applications, such as integrating terrestrial and marine vertical references.
Keywords: Tides; Positioning; PPK; PPP; Hydrography.

How to cite this article: SANTANA FR, KRUEGER CP, BALUTA ESM, SANTANA TA, VESTENA KM. Evaluation of precise point
positioning and post-processing kinematic methods for tide measurement in hydrographic surveys. Bulletin of Geodetic
Sciences. 30: e2024008, 2024.

This content is licensed under a Creative Commons Attribution 4.0 International License.
Evaluation of precise point positioning and post-processing kinematic methods... 2

1. Introduction

With technological development, positioning, land, sea and air navigation have gained another approach
through the junction between systems on a global or regional basis. This is known as the Global Navigation Satellite
System (GNSS) which provides navigation and timing services to GNSS receivers (Krueger 1996).
Despite the popularization of the Global Positioning System (GPS) (Hegarty 2017), there are other systems
with the same purposes, such as GLONASS (Global’naya Navigatision-naya Sputnikovaya Sistema) (Revnivykh et al.
2017), GALILEO (European Satellite Navigation System) (Falcone, Hahn and Burger 2017), BEIDOU (Chinese Satellite
Navigation System), (Yang, Tang and Montenbruck 2017), the regional systems such as the QZSS (Quasi-Zenith
Satellite System) and the IRSS (Indian Regional Navigation Satellite System) (Kogure, Ganeshan and Montenbruck
2017). For maritime positioning, the demand for this technology has grown more and more due to two main
advantages: high accuracy and low cost. Regarding hydrographic surveys, there is the possibility of using different
methods of GNSS positioning, mainly in the search for greater accuracy of solutions, and thus greater safety of
maritime navigation. For vertical positioning, for example, there is the possibility of application in the reduction
of soundings, which consists of discounting the portions referring to the tide and the attitude movements of the
vessel. Traditionally, this sea level measurement is made by tide gauge stations on the coast. However, the tide
measured at these stations will not necessarily be the same in the place where the vessel is located, causing errors,
which are called cotidal, which can exceed 3 m, depending on the complexity of the area (USACE 2013). On the
other hand, according to the Standards for Hydrographic Surveys (IHO 2018) from the International Hydrographic
Organization (IHO), the Maximum Vertical Uncertainty allowed in navigation channels with a depth of 10 m, for
example, is 17 cm. A solution to reduce sounding reduction uncertainties is to use the vessel’s ellipsoidal height
variation as representative of tidal oscillations. In this way, sea level variation is measured in the exact location
where the sounding platform is located and not on the coast, instilling cotidal errors.
In this context, several studies have already been carried out, for example, in Brazil Ramos (2007), who
analyzed the use of the precise differential method in real time in single beam sounding and developed the
SIRBAT-GPS software for sounding reduction; Oliveira et al. (2010), who processed GNSS data with a vessel using
PPP and PPK in the aforementioned software and compared it with the sounding reduction using a tide gauge
station; and Neto (2020), who evaluated the use of RTK in multibeam sounding. Internationally, we can mention
Hocker (2010), who used the PPK method to process GNSS buoy data; in addition to Alkan and Öcalan (2013) and
Abdallah (2016), who used the PPP during a hydrographic survey, the first using Leica Geo-office software (Leica
geosystems 2023) and the second using Bernese (Dach 2015); and, more recently, El-Diasty (2022), who used
the PPK method coupled with an Inertial Measurement Unit to define a separation model between the Chart
Datum (CD) and the ellipsoid. This work complements such studies, as it presents the assessment of the GNSS
positioning carried out at different times of the year and compares two different validation approaches: the first
using a tide gauge station as a reference for the PPP and PPK methods and the second, using the PPK method as
a reference for the PPP.
The reason for using PPK as a reference is that it has a lower uncertainty than PPP. According to Krueger
(1996), PPK is characterized by the simultaneous observation of the satellite signal at least two different stations,
taking one as a base point of known coordinates, which contributes to a significant reduction in errors, especially
those of the satellite clock, orbit and signal propagation in the atmosphere.
On the other hand, in PPP a base station is not used, but the precise positions of the GNSS satellites and clock
corrections are obtained by a network of GNSS receivers distributed globally (Zumberg et al. 1997). In addition,
further corrections to the signal observations are required, for example due to the influence of the signal on the
troposphere and ocean loads. More details can be found in Kouba and Héroux (2001) and Abdallah (2016). As an
evaluation metric, the uncertainty criteria of the Survey Orders defined by the IHO and the requirements of the
Boletim de Ciências Geodésicas, 30: e2024008, 2024
3 Felipe Rodrigues Santana et al.

Canadian Hydrographic Service (CHS 2013) for the fulfillment of these Orders were used. In the next section, the
methodology used in the work will be presented, followed by the results and discussions. It should be noted that the
use of GNSS for measuring the tide must be accompanied by a separation model between the CD and the ellipsoid,
which includes geoid models and Mean Sea Level Topography. Such models will not be addressed in this article, but
more details can be consulted in Santana et. Al (2020).

2. Methodology

A hydrographic vessel was used, equipped with a GNSS receiver (GPS and GLONASS), whose data were saved
in RTCM format and converted to Rinex, through the rinexconv software (Kongsberg 2021). In TABLE 1, the type of
equipment used, and the purpose can be observed.

Table 1: Equipment used and its purposes.

Equipment Purpose
Receiver 3710 DGNSS RTCM for post-processing
Seapath-130 (two antennas and MRU 5) Vessel’s attitude (Roll, Pitch, Heave, Yaw)

Except for the base station equipment, all other sensors were provided by the company Delfos Marítima,
along with the Delfos speedboat, which has a retractable lateral rod, where the transducer, the positioner and the
inertial sensor are installed (Figure 1).

Notes: Delfos hydrographic boat moored next to the tide gauge of Charitas.
Source: Delfos Marítima (2021).

Figure 1: Delfos hydrographic boat

Boletim de Ciências Geodésicas, 30: e2024008, 2024
Evaluation of precise point positioning and post-processing kinematic methods... 4

Terrapos software (Terrapos 2021) was used for the PPP and PPK methods. The version used was the 2.5.9p2,
more information can be obtained on https://field.group/service/terrapos/. The products of Center for Orbit
Determination in Europe (CODE) were used because they have corrections for GPS and GLONASS CDDIS (2021). The CLK
files were used to correct the time delay between the receiver and the satellite. And to correct the pole coordinates,
pole rates and Length-of-day (LOD), the Earth Rotation Parameters (ERP) were used. Precise Ephemerides (SP3 files)
were employed to determine the orbits of the satellites. To remove the first order ionospheric effects for the pseudo
distance equation, the Ionosphere-free combination was used. In Terrapos, the Hydrostatic Zenith Delay is given by
the Saastamoinen model. The troposphere is handled through meteorological lookup tables (UNB3), which contains
the vertical gradients for temperature and water vapor pressure (Børst and Kjørsvik 2023). Regarding the mapping
function, the Global Mapping Function was used. According to Abdallha (2016), the difference between models like
NMD (Niell Mapping Function), GMF (Global Mapping Function) and VMF (Vienna Mapping Function) is less than
1 mm, so it is insignificant for hydrographic surveys. For the antenna calibration, it’s necessary to determine the
phase center variation (PCV) and the phase center offset (PCO) (Luo 2013). The file “igs14.atx” was used to do this,
as well as the “FES2004” was used for the Ocean Tide Loading corrections which were implemented by Le Provostl
and Lyard (1997). In all processing, an elevation cutoff angle of 10 degrees was adopted. The parameters used in the
processing methodology are shown in table 2:

Table 2: Files and Parameters used.

Item Strategies
Frequencies GPS L1 / L2 / L5 and GLONASS G1 / G2 / G3
Elevation cutoff angle 10°
Ocean Tides FES2004
Satellite Ephemeris, Clock and Earth Rotation Parameters CODE final products
Ionosphere Ionosphere-free linear combination with dual-frequency
Troposphere Estimate ZTD
Mapping function GMF
Metrology parameters UNB3
Receiver phase center PCO and PCV values from igs14.atx
Satellite phase center PCO and PCV values from igs14.atx

The areas of study involve hydrographic surveys carried out from May to December 2021, in Ilha Grande Bay
and Guanabara Bay, off the coast of Rio de Janeiro. In table 3, it is presented the distance of the surveys at each
RBMC (Brazilian Network for Continuous Monitoring) GNSS reference station used; the location of the surveys; and
whether the boat was moored by the tide gauge, or sailing.
The data collected in 2021 and made available by Delfos were used to choose the 12 campaigns. Data were
divided into days, modes (moored or sailing) and when there was a gap of more than 10 minutes. The analysis
results are unaffected by the seasonal variation of the tidal phenomenon. This is because the variation in the
antenna’s ellipsoidal altitude can represent, with some uncertainty, all the variations in sea level. However, what
most influences the results is the seasonal variation of the GNSS signal, which is subject to the behavior of the
ionosphere and interference such as multipath, among others. For this reason, the increasing number of samples in
the study makes the results more consistent.
Regarding the base stations, May 2, in Ilha Grande Bay, was the only day when four stations (UBA1, ONRJ,
RJNI, CHPI) were used for the PPK method, at a minimum distance of 83 km. On the other hand, the surveys that
were carried out in Guanabara Bay used only the RJNI and ONRJ stations, so that all survey lines were less than 13
km away from these GNSS reference stations.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

5 Felipe Rodrigues Santana et al.

Table 3: Characteristics of hydrographic surveys.

Distance from RBMC base stations (Km)

Date Period Locals
ONRJ1 RJNI2 CHPI3 UBA14 Mode Approach
2021 (hours) (Bays - RJ)
Max Min Max Min Max Min Max Min
02/05 13h 107 105 116 114 84 83 105.5 103.7 Ilha Grande sailing second
06/05 6h 13 7.95 11.5 0.4 sailing second
first and
10/05 6h
second
first and
11/05 24h
second
first and
12/05 11h moored
second
12.8 12.8 4.22 4.2
Stations were not used in first and
13/05 12.5h Guanabara
the GNSS post-processing second
first and
13/05 12h
second
14/05 11h second
14/05 6h second
01/11 5h 12.7 8.75 4.22 0.4 sailing second
02/11 8h 8.59 1.34 8.53 0.5 second
03/12 5h 12.7 8.14 4.21 0.5 second
Source: The authors
Notes: ONRJ: Observatório Nacional do Rio de Janeiro; 2 RJNI: Niterói; 3 CHPI: Cachoeira Paulista; and 4 UBA1: - Ubatuba.
1

Figures 2 and 3 show the coverage area and the trajectory of the surveys in Ilha Grande Bay and one of
the surveys carried out in Guanabara Bay, respectively. In all of them, the location of the RBMC stations used as a
reference are also indicated. For the PPK, all station coordinates were updated for the time of the survey, using the
weekly solutions of the SIRGAS-CON Network (SIRGAS 2022). The evaluation of uncertainties was carried out by
two approaches: the first, having as reference the sea level variation measured at the tide gauge station with the
speedboat moored, and the second, having as reference the variation of the ellipsoidal heights of the more accurate
method (PPK) with the vessel both moored and sailing. In both approaches, all series were subtracted from their
respective means for comparison purposes and an empirical cumulative distribution function was used to calculate
the 95% confidence interval, as performed by El-diasty (2020).

Boletim de Ciências Geodésicas, 30: e2024008, 2024

Evaluation of precise point positioning and post-processing kinematic methods... 6

Source: The authors.

Figure 2: Hydrographic survey carried out on May 2 (Ilha Grande Bay).

Boletim de Ciências Geodésicas, 30: e2024008, 2024

7 Felipe Rodrigues Santana et al.

Source: The authors.

Figure 3: Hydrographic survey carried out on November 1 (Guanabara Bay).

In the first approach, which adopts the tide gauge station as a reference, it was necessary to resample the
time series of the GNSS data, collected at an interval of 1s, for the same rate as the station, with a rate of 1 minute.
For this, the Interp1 function of Matlab (MathWorks Inc. 2022) was used, with the “spline” model to perform the
interpolation, or extrapolation of the GNSS data, in addition to filling in those instants with missing data. Once all
the series started to be in the same interval, It was necessary to apply a filter to smooth the GNSS data. However, if
it had been applied only the moving mean, the outliers present in the series could contaminate the data processed.
For this reason, it was initially applied a filter for detection, identification and removal of outliers, called Hampel
(Hampel 1971), which was also used by Liu et al. (2014) and Mawrey (2016). This filtering method is carried out by
calculating the median of a window composed of “k” values ​​in the neighborhood and the estimate of the standard
deviation of each sample in relation to the window median, using the absolute mean deviation. If a sample differs
from the median by more than “s” standard deviations, it is replaced by the median (Hampel 1971). The second
filter of the moving mean (Smith 1999), was used to smooth the data, using a “w” window length slide. The values
of “k”, “s” and “w” were calculated to obtain the lowest RMSE between the method to be analyzed and the variation
of the sea level measured at the tide gauge station.
The RMSE was chosen as a metric to select the filter parameters because, according to Isaacs and Srivastava
(1989), RMSE incorporates both the spread error distribution and the bias. Previous works has also used the RMSE
metric, such as, Alsaaq et al. (2016).
At the tide gauge station, a filter was also applied. According to FIG (2014), the sea level measurement data
at tide station must be passed through a low-pass filter to extract only sea-related data. At the same time, OHI
Boletim de Ciências Geodésicas, 30: e2024008, 2024
Evaluation of precise point positioning and post-processing kinematic methods... 8

(2011 p. 281) complements by stating that “the sampling interval should be short enough to measure any seiche
action”, which are the sea level variations caused by the oscillations of the sea waves. After tests were carried out
with filters varying from 3 minutes to 2 hours, it was verified that the filters from 3 minutes to 15 minutes could
maintain the tide signal from the sea. Hence, for the purpose of standardization registration, the 5-minute filter
was employed, which is also recommended by Scarfe (2002) and Ramos (2007). In the second approach, with the
boat both moored and sailing, the PPK was used as a reference for evaluating the PPP. The PPP uncertainties were
measured according to El-Diasty (2010), who suggests taking the root of the sum of squares of the PPP RMS in
relation to the PPK, with the theoretical PPK RMS, calculated by the variance and covariance matrices obtained
in the post-processing (Equation 1).
Since PPP has greater uncertainty than PPK, this method was used as a reference for the RMS calculation,
according to the equations (1), (2) and (3) as indicated by El-Diasty (2010), Elsobeiey (2013), El-Diasty and Elsobeiey
(2015), El-Diasty (2020) and Santana (2021). For the calculation of the 95% confidence level (CL), a cumulative
distribution function was used, with an empirical model, which allows the calculation of the CL independently of
two subsequent data have or not a normal distribution.
n 22 + (E 22
2D √ ∑∑ni=1 ((NPPK
i=1((N −NNPPP
PPK− PPP))i i + (EPPK −EEPPP
PPK− PPP))i i))
RMS2D
RMSPPP_PPK =
PPP_PPK = √  (1)
nn

NPPK  n 22
1D √∑∑ni=1 (UPPK
i=1(U PPK−−UUPPP
PPP))i i
RMS1D
RMSPPP_PPK =
PPP_PPK = √  (2)
NPPK 
NPPK
PPP NPPK  nn
Where:
N
EPPKN
NPPK  NN 
PPP PPP  PPK
PPP
RMSPPP = √RMS 22
√RMSPPP_PPK + RMS 22 
RMSPPK
N  RMS PPP = PPP_PPK + PPK
N
N
EPPK
PPK
PPP :
 the
EPPK 
PPP Northing coordinate of the PPK, andEN 
PPP
PPK  the Northing coordinate of the PPP;
E
E:PPK
ENPPK
U  Easting coordinates of the PPK, and EEPPP
 the  the
 Easting coordinate of the PPP;
PPP
PPP PPP  PPK

EEPPP
𝑈𝑈
U PPK
E:PPP
U  vertical coordinate of the PPK;
 the UEPPP
PPK 
𝑃𝑃𝑃𝑃𝑃𝑃
PPK PPK
2D UPPK  𝑈𝑈U 
𝑈𝑈E𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
RMSU NPPP
PPK 𝑈𝑈PPK
N
𝑃𝑃𝑃𝑃𝑃𝑃
PPK
the
  vertical coordinate of the PPP;
𝑃𝑃𝑃𝑃𝑃𝑃 PPK
𝑃𝑃𝑃𝑃𝑃𝑃

U1D
RMS𝑈𝑈 N
2D 𝑈𝑈2D   : the 2D total horizontal uncertainty
𝑃𝑃𝑃𝑃𝑃𝑃 RMS 2D𝑈𝑈𝑃𝑃𝑃𝑃𝑃𝑃
of−   error2 of the Northing and
the
RMS N N𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃 − E )2 ) Easting position of the PPP
PPK
𝑃𝑃𝑃𝑃𝑃𝑃
PPK
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
PPP PPP ∑n ((N 𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
i=1 PPK N ) + (E
PPP i PPK PPP
i
and RMSPPP_PPK = √
2D
in relation
RMS to
2D
  the PPK;
 2D 
RMS𝑈𝑈𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃 
2D RMS
RMS N 
1D
RMSRMS1D
E E𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
1D
PPK 
𝑃𝑃𝑃𝑃𝑃𝑃
PPP
PPK
PPP_PPK
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃  RMS 𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃 n
  : the 1D vertical uncertainty of RMS
the error
RMS
RMS 2D 1D
E 1D of the Vertical position of the PPP in relation to the PPK.
RMS
RMS
RMS
1D
RMS E𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
EPPK
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
PPP
PPK
PPP_PPK PPP
PPP_PPK
RMS 
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
PPP_PPK
n
After
1D
RMSEPPP calculating
   the RMS of the PPP1Din relation to∑PPK,
i=1(U the − UPPP
PPKtotal )2i was obtained as described in Equation
RMS
RMS
RMS U U
RMS
PPP_PPK
PPK 
PPP_PPK
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
PPK
PPP PPK
PPK  RMS PPP_PPK
RMS
= √
RMSPPP_PPK
PPK   
3, considering the theoretical RMS of the PPK. n
RMS
RMSU
𝑈𝑈RMS
PPK
RMS
PPP_PPK
𝑃𝑃𝑃𝑃𝑃𝑃
PPK
PPP PPK
𝑈𝑈𝑃𝑃𝑃𝑃𝑃𝑃  
PPP 
RMS
RMS PPK
PPP 
𝑈𝑈RMS 2DPPP
   PPP 
2 2
RMS
RMS
RMS
RMS 2D
𝑃𝑃𝑃𝑃𝑃𝑃
PPK
PPP𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
RMSPPP_PPK
RMSPPP = √RMS + RMSPPK  (3)
2D
RMS
RMS
Where:
RMS 1D
RMS 1D
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
PPP
 
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃
1D
RMS 𝑃𝑃𝑃𝑃𝑃𝑃_𝑃𝑃𝑃𝑃𝑃𝑃 :
RMS
RMS RMS (2D or 1D) of the PPP in relation to the PPK;
PPP_PPK
PPP_PPK

RMS
RMS
RMSPPP_PPK
PPK
PPK 
: theoretic RMS of the PPK (2D or 1D), originated in the matrix of variance and covariance obtained
in the post-processing;
RMS PPK 
RMS
RMS and
PPP
PPP

RMSPPP : total RMS (2D or 1D) of the PPP.

In addition, because the PPP and PPK data were captured at the same 1s interval, with aligned epochs, it was
not necessary to interpolate the data, nor to apply any filters, following the methodology of El-Diasty (2020). The
advantage of this approach is that it enables the calculation not only of the vertical uncertainties, but also of the
horizontal ones and even in the trajectories in which the boat is far from the tide gauge station.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

9 Felipe Rodrigues Santana et al.

The criteria chosen for assessing the quality of GNSS data are those recommended by CHS (2013), whose
vertical uncertainty limit (2σ) for water level measurement is 5 cm for Special or Exclusive Order, and 10 cm for other
orders. However, in this work, only for Orders 1A and 1B the limit of 10 cm was adopted; for order 2, a limit of 15
cm was considered, as it has a higher uncertainty tolerance for vertical positioning than orders 1A and 1B, according
to OHI (2018). Furthermore, it should be noted that these values are indicators that, if achieved, increase the
probability of meeting the total vertical uncertainty requirements of each order, as required by OHI (2018). At the
end of the survey, such indicators will be only a portion of all propagated vertical uncertainties such as, for example,
the speed of sound in water, offsets between sensors, the attitude of the vessel, among others (SOUZA 2011).

3. Results and Discussion

3.1 First approach: Tide Gauge as reference

Figure 4 shows the comparison between the PPP and PPK in relation to the tide station, on May 11th in
Guanabara Bay with the boat moored. The appreciation of the histograms allows evaluating that the data may
not have a normal distribution. For the calculation of the RMSE, with a confidence level of 95%, an accumulated
distribution function was used, as used by El-Diasty (2020).

Source: The authors.

Figure 4: Statistical analysis of the PPP and PPK methods in relation to the tide station.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

Evaluation of precise point positioning and post-processing kinematic methods... 10

In Figure 5, the results of the first approach are presented, where each point represents the RMSE value
related to a period. It is observed that in all five analyzed periods, the vertical uncertainties of the PPK were lower
or close to 5 cm. As for the PPP, in 5 days a maximum value of 10 cm was observed, and in the other days the values
were between 5 cm and 10 cm.

Source: The authors.

Figure 5: First approach: PPP and PPK vertical uncertainty in relation to the tide station.

3.2 Second approach: PPK as reference

Figure 6 shows a statistical analysis, using the error histogram, the q-q plot, and the box plot, of the
horizontal and vertical components of the PPP in relation to the PPK for May 11th. It is possible to observe
that the data do not necessarily follow a normal distribution pattern, such as the longitude histogram. This
is confirmed also in the QQ plot, which displays the quantile-quantile plot of the quantile of the vertical and
horizontal positioning errors versus the theoretical quantile values from a normal distribution, represented by
the dashed lines. Whereas the real distribution is plotted using plus sign (‘+’) markers. If the real distribution were
normal, the data plot should be linear.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

11 Felipe Rodrigues Santana et al.

Source: The authors.

Figure 6: Statistical analysis of the PPP method in relation to PPK.

According to IHO (2018), the errors should be estimated at 95% confidence level. Figure 7 shows the cumulative
positioning error for vertical and horizontal components. The blue line represents the vertical positioning error;
the red line the horizontal positioning error; and the green line represents the 95% confidence level of the RMSE.
An error equivalent to 0.053m was found for the vertical component and an error equivalent to 0.039m for the
horizontal component.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

Evaluation of precise point positioning and post-processing kinematic methods... 12

Source: The authors.

Figure 7: Cumulative position error for PPK navigation solution.

As for the results of the second approach (Figure 8) for horizontal uncertainty, all days reached Special/
Exclusive Order. Regarding the vertical uncertainty, of the 12 periods analyzed, in 8 a maximum value lower or equal
than 5 cm was observed for the PPP, in 3 periods values between 5 and 10 cm and in one period values between 10
and 15 cm.

Source: The authors.

Figure 8: Second approach: Horizontal and vertical uncertainty of PPP in relation to PPK.
Boletim de Ciências Geodésicas, 30: e2024008, 2024
13 Felipe Rodrigues Santana et al.

3.3 Discussion

In Figure 9, it is indicated, in which Order of Survey the PPP, or PPK fits according to the criteria of CHS (2013).

Source: The authors.

Figure 9: Results of the first and second approach.

In the first approach, with the boat moored and having the tide station as a reference, the PPP method was able
to reach the special order in 50% of the periods, while the PPK method in 100%. From the results found in the second
approach, with the boat both moored and sailing and using the PPK method as a reference, it was verified that the
horizontal component of the PPP fulfilled the estimated requirements for the special order in 100% of the cases, while
for the vertical component, this occurred 67% for the Exclusive Order, 25% for the 1A Order only and 8% for the 1B
Order only. A possible justification for the differences in results when comparing the two approaches (Figure 2) may
be linked to the use of filters to compare data from the tide gauge station, in addition to the fact that the tide gauge
station sensor is not susceptible to PPK uncertainties. Another fact that could contribute to raising the RMS of the PPP
in relation to the PPK was considering all the trajectories of the boat, including passages under, or close to bridges.
The results of the PPP stand out for being an alternative for hydrographers in soundings, since in all bathymetries
a post-processing is necessary. Furthermore, the strictest order for a single beam survey is 1b (IHO, 2018), which
results showed that it can be met in 100% of the surveys using the PPP method. Regarding the PPK, despite being
able to measure sea level with the requirements of the Special Order, it requires a greater logistical effort to maintain
reference stations. What can be minimized when using the IBGE (Brazilian Institute of Geography and Statistics)
RBMC stations, as presented. In any case, the use of post-processed GNSS data represents an alternative method
for sounding reduction that is more effective, economical, and accurate.
When choosing the second way to check GNSS accuracy, it is assumed that the PPK’s theoretical uncertainty
is not related to the PPP’s uncertainties. This fact is not necessarily true, as the quality of the PPK in relation to
the PPP depends on the vessel’s proximity to the reference GNSS station and the quality of its data. All of this
would contribute to affirming that the first approach would be the most adequate for this type of evaluation, in
comparison with the second approach, commonly used in the literature. Although, the mean difference of the first
and second approach was 0.8 cm, and the number of periods that complied the Special and 1A order was the same,
in the periods when the vessel was moored.
Boletim de Ciências Geodésicas, 30: e2024008, 2024
Evaluation of precise point positioning and post-processing kinematic methods... 14

The authors suggest, for future works, the comparison between both methods, but with the boat traversing
trajectories close to a tide gauge, which would exempt the influence of cotidal errors. This would give the possibility
of evaluating the impact of the vessel’s movement on the accuracy of the methods. The option of using the Inertial-
Aided PPK (IA-PPK) as a positioning reference is also recommended, to reduce the uncertainty of the reference
method, as used by El-Diasty (2020).

4. Conclusion

The use of ellipsoidal heights for measuring sea level can reduce vertical uncertainties in hydrographic surveys.
This data can be improved in quality using post-processing techniques, with or without a base station. The results
presented for the PPK were able to fulfill the requirements for the Exclusive Order in 100% of six periods of survey
when the vessel was moored, using the Tide Gauge as a reference. For the PPP, this value was 50% having the PPK as
reference for 12 periods, with the vessel moored or sailing, it could be inferred that 100% of the PPP dataset could
comply with the 1b order, out of this number, 92% of periods adhered to Order 1A and 67% would also be capable
of adhering to the Special Order. The vertical uncertainty of the PPP in relation to the PPK ranged between 4.1 cm
and 13.2 cm, which consistent with Abdallah (2016), who found a value of 12.10 cm. Which means that 100% of
the PPP dataset would fulfill the 1b order, according to the CHS criteria, standing out for being an alternative for
hydrographers in soundings.
The mean difference of the first and second approach was 0.8 cm, with the vessel, however more investigations
are needed to evaluate the difference of two approaches with the vessel sailing and using the IA-PPK as reference.
Additionally, using the ellipsoid as a vertical reference is useful for other marine positioning applications.

ACKNOWLEDGEMENT

To the Brazilian Navy, for financing this study through Ordinance No. 279 of the Navy Military Personnel
Directorate (DPMM). To Professor Regiane Dalazoana and Professors Silvio Freitas and Ítalo Ferreira for their valuable
suggestions. To the company Delfos Marítima, in the person of Comte Aluízio, and the company’s employees, Pedro,
Ítalo, Rogério, Alex. To Even Brøste and Narve S. Kjørsvik, from the company Terratec, in Norway, for providing the
Terrapos program license. To Aleksander Hammernes, from Kongsberg, for implementing the Seapath workflow in
the Terrapos software, which greatly facilitated the data processing for this work.

AUTHOR´S CONTRIBUTION

Felipe Rodrigues Santana: Conceptualization, methodology, drafting, data collection, data analysis, writing,
Literature review, editing; Claudia Pereira Krueger: conceptualization, methodology, revision, data analysis, editing;
Érica Santos Matos Baluta: conceptualization, methodology, revision, editing, data analysis; Tulio Alves Santana:
writing, revision, editing; Kaue de Moraes Vestena: revision, data analysis.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

15 Felipe Rodrigues Santana et al.

REFERENCES

Abdallah, A. T. M. 2016. Precise Point Positioning for Kinematic Applications to Improve Hydrographic Survey. Thesis
(Doctorate of Engineering Sciences) - Institute of Engineering Geodesy (IIGS), University of Stuttgart, 2016. Available
in: https://elib.uni-stuttgart.de/handle/11682/9043. [Accessed 31 jul 2021].
Alkan, R. M.; Öcalan, T. 2013. Usability of the GPS precise point positioning technique in marine applications. Journal
of Navigation, v. 66, n. 4, p. 579–588, 2013.
Alsaaq, F., Kuhn, M., El-mowafy, A. and Kennedy, P. 2016. Filtering methods to extract the tide height from Global
Navigation Satellite Systems (GNSS) signals for Hydrographic applications. November, 8–10. Available in: https://
core.ac.uk/download/pdf/220107518.pdf. [Accessed 01 dec 2021].
Børst E. and Kjørsvik N.S. 2023. Doubt about Ocean Loading Displacement on Terrapos + Doubt about Terrapos
Setting. Message received by [email protected] 12 dec 2023.
Canadian Hydrographic Service (CHS). 2013. Standards for Hydrographic Surveys. Survey Management Guidelines.
Available in: <https://ppa.gc.ca/sites/default/files/2018-09/CHS%20Standards%20for%20Hydrographic%20
Surveys.pdf>. [Accessed 13 jul. 2020].
Crustal Dynamics Data Information System (CDDIS DAAC). 2020. Global Navigation Satellite System (GNSS) Overview.
Available in:<https://cddis.nasa.gov/Techniques/GNSS/GNSS_Overview.html>. [Accessed 29 jul. 2020].
Dach, R.; Walser, P. 2015. Bernese GNSS Software Version 5.2 Tutorial. Available at: <https://users.aalto.fi/~mvermeer/
TUTORIAL-X.pdf>. [Accessed 14 dec. 2023].
Delfos Marítma. 2021. Photo of the hydrography boat moored at Clube Naval Charitas. Personally available to the
Author.
El-Diasty, M. 2010. Development of a MEMS-Based INS/GPS Vessel Navigation System for Marine Applications. PhD
Dissertation, Publication Number NR64919, York University, Toronto.
El-Diasty, M. 2020. Evaluation of KSACORS-based network, GNSS-INS integrated system for Saudi coastal hydrographic
surveys. Geomatics, Natural Hazards and Risk, 11(1), 1426–1446. DOI: 10.1080/19475705.2020.1799081
El-Diasty, M.; Elsobeiey, M. 2015. Precise Point Positioning Technique with IGS Real-Time Service (RTS) for Maritime
Applications. Positioning, 06(04), 71–80. DOI: 10.4236/pos.2015.64008
El-Diasty, M.; Kaloop, M.R.; Alsaaq, F. 2022. Chart Datum-to-Ellipsoid Separation Model Development for Obhur
Creek Using Multibeam Hydrographic Surveying. J. Mar. Sci. Eng., 10, 264. DOI: 10.3390/jmse10020264
Elsobeiey, M. 2013. Performance analysis of low-cost single-frequency GPS receivers in hydrographic surveying.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives,
42(4W5), 67–71. DOI: 10.5194/isprs-archives-XLII-4-W5-67-2017
Falcone, M.; Hahn, J.; Burger, T. 2017. Galileo. In: Teunissen, P. J. G.; Montenbruck, O. (Ed.) Springer Handbook of
Global Navigation Satellite Systems. Springer, Berlin. p. 247-272.
Fédération Internationale Des Géomètres (FIG). 2014. Ellipsoidally Referenced Surveying for Hydrography. N. 62.
Copenhagen, Denmark. Available at: : <https://www.fig.net/resources/publications/figpub/pub62/Figpub62.pdf>.
[Accessed 20 jul. 2020].
Hampel, F. R. 1971. A general qualitative definition of robustness. Annals of Mathematics Statistics, 42, 1887–1896.
Hegarty C.J. (2017) The Global Positioning System (GPS). In: Teunissen P.J. G.; Montenbruck, O. (Ed.) Springer
Handbook of Global Navigation Satellite Systems. Springer, Berlin. p. 247-272.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

Evaluation of precise point positioning and post-processing kinematic methods... 16

Hocker, B.; Wardwell, N. 2010. Tidal datum determination and VDatum evaluation with a GNSS buoy. 23rd
International Technical Meeting of the Satellite Division of the Institute of Navigation 2010, ION GNSS 2010. Anais...
. v. 3, p.2076–2086.
International Hydrographic Organization (IHO). 2011. Manual on Hydrography (C-13). Monaco. Available at: <https://
www.iho.int/iho_pubs/IHO_Download.htm>. [Accessed 14 dec. 2023].
International Hydrographic Organization (IHO). 2018. IHO Standards for Hydrographic Surveys. Principauté de
Monaco. Available at:<https://iho.int/en/miscellaneous-publications>. [Accessed 14 dec. 2023].
Isaacs, E.; Srivastava, M.R. 1989. An Introduction to Applied Geostatistics. Oxford University Press, New York, 592.
Kogure, S.; Ganeshan, A. S.; Montenbruck, O. 2017. Regional Systems. In: TEUNISSEN, P. J. G.; MONTENBRUCK, O.
(Ed.) Springer Handbook of Global Navigation Satellite Systems. Springer, Berlin, 2017. p. 305-338.
Kongsberg, K. 2013. 3710 receiver differential GNSS receiver instruction manual. Available at:<https://www.
manualslib.com/manual/1422390/Kongsberg-3710.html>. [Accessed 20 jun. 2020].
Kouba, J. and Heroux, P. 2001. Precise Point Positioning Using IGS Orbit and Clock Products. GPS Solutions, 5, 12-28.
DOI: 10.1007/PL00012883.
KRUEGER, C. P. 1996. Investigações sobre aplicações de alta precisão do GPS no âmbito marinho. 288 f. Thesis (Doctor
in Geodetic Sciences) – Earth Sciences Sector, Federal University of Paraná, Curitiba, 1996.
Le Provost, C., Lyard, F. (1997): Energetics of the M2 barotropic ocean tides: an estimate of bottom friction dissipation
from a hydrodynamic model. Progress in Oceanography, 40(1-4).
Leica Geosystems. 2023. Leica Geo Office – one integrated Office Software. Available at: <https://leica-geosystems.
com/rugbycl/archive-data/software/leica-geo-office>. [Accessed 20 jul. 2023].
Luo, X. 2013. Mathematical Models for GPS Positioning. In GPS Stochastic Modelling; Springer Theses; Springer:
Berlin/Heidelberg, Germany, pp. 55–116; ISBN 978-3-642-34835-8.
Mawrey, R.S (2016). Developing na IoT Analytics System with MATLAB, Machine Learning, and ThingSpeak. In
MathWorks. Available on: https://www.mathworks.com/company/newsletters/articles/developing-an-iot-
analytics-system-with-matlab-machine-learning-and-thingspeak.html. [Accessed 20 jul. 2023].
Neto, W. P. A. 2019. Utilização do sistema GNSS RTK para o ajuste de marés em batimetria automatizada multifeixe,
136 f. Dissertation (Master in Transport Engineering) - Transport Engineering Sector, Instituo Militar de Engenharia,
Rio de Janeiro, 2019.
Oliveira Junior, A. M.; Arroyo, E. N. S.; Ramos, A. M.; Arentz, M. F. R. 2010. Seabed Mapping on an Earth Centered
Earth Fixed (ECEF) Geocentric Reference Frame. Cooperative Validation with US Navy and Brazilian Navy in Guanabara
Bay, Rio de Janeiro. In: ION GNSS 2010 CONFERENCE. Portland OR USA. September 21 – 24.
Ramos, A. M. 2007. Aplicação, Investigação e Análise da Metodologia de Reduções Batimétricas Através do Método
GPS Diferencial Preciso. 227 f. Dissertation (Master’s in Geodetic Sciences) – Earth Sciences Sector, Federal University
of Paraná, Curitiba, 2007.
Revnivykh, S. Bolkunov, A. Serdyukov, A. and Montenbruck, O. 2017. GLONASS. In: Teunissen, P., Montenbruck, O.
(Eds). Springer Handbook of Global Navigation Satellite Systems. Springer, Cham, pp. 219-245.
Santana, F. R. 2021. Determinação e validação de modelos de separação de superfícies com referência ao elipsoide
pelo GNSS e GNSS/INS. f. Dissertation (Master’s in Geodetic Sciences) – Earth Sciences Sector, Federal University of
Paraná, Curitiba.
Santana, F. R.; Krueger, C. P.; Santana, T. A.; Nascimento, G. A. G.; Oliveira Junior, A. M. 2020. Nautical Charts with
SEP Models: Historic Evolution and Perspectives for the Brazilian Hydrography. Revista Brasileira de Cartografia, [S.
l.], v. 72, p. 1299–1328, DOI: 10.14393/rbcv72nespecial50anos-56616.

Boletim de Ciências Geodésicas, 30: e2024008, 2024

17 Felipe Rodrigues Santana et al.

Scarfe, B. 2002. Measuring Water Level Corrections (WLC) using RTK GPS, The Hydrographic Journal, nº 104.
Available in: <https://web.archive.org/web/20070324014031/http://www.hydrographicsociety.org/Articles/
journal/2002/104-3.htm>. [Accessed 12 dec. 2023].
Sistema de Referência Geodésico para as Américas (SIRGAS). 2022. Coordenadas semanales de las estaciones SIRGAS-
CON. Available in: <https://sirgas.ipgh.org/red-gnss/coordenadas/coordenadas-semanales/?msclkid=f2693357bb3
c11ec9509f0d33ce8af20>. [Accessed 1 jun. 2022].
Smith S. W. 1999. The Scientist and Engineer’s Guide to Digital Signal Processing. 2nd ed. San Diego: California
Technical Publishing.
Souza, A. V. 2011.Análise dos parâmetros que compõem a equação da incerteza vertical propagada da profundidade
reduzida. Master’s thesis in Geodetic Sciences, Earth Sciences Sector, Department of Geomatics, Federal University
of Paraná, 2011. 167p
Terrapos. 2021. USER’S MANUAL. Personally made available to the Author. Available in: <https://field.group/service/
terrapos/>. [Accessed 14 dec. 2023].
The MathWorks Inc. 2022. MATLAB version: 9.13.0 (R2022b), Natick, Massachusetts: The MathWorks Inc. Available
in: <https://www.mathworks.com/help/matlab/ref/interp1.html>. [Accessed 14 dec. 2023].
US ARMY CORPS OF ENGINEERS (USACE). 2013. EM 1110-2-1003 Engineering and Design Hydrographic Surveying.
Department of the Army, Washington.
Yang, Y.; Tang, J.; Montenbruck, O. 2017. Chinese Navigation Satellite Systems. In: Teunissen, P. J. G.; Montenbruck,
O. (Ed.) Springer Handbook of Global Navigation Satellite Systems. Springer, Berlin, 2017. p. 273-304.
Zumberge, J. F.; Heflin, M. B.; Jefferson, D. C.; Watkins, M. M.; Webb, F. H. (1997). Precise point positioning for the
efficient and robust analysis of GPS data from large networks. Journal of Geophysical Research: Solid Earth, 102(B3),
5005–5017. DOI: 10.1029/96jb03860.
Additional Information
Data Availability: The data used to generate the manuscript can be find on https://anonymous.4open.science/r/
BCG-2023-article-B117/README.md

Boletim de Ciências Geodésicas, 30: e2024008, 2024