The Egyptian Journal of Remote Sensing and Space Sciences 25 (2022) 551–559

Contents lists available at ScienceDirect

The Egyptian Journal of Remote Sensing and Space Sciences

journal homepage: www.sciencedirect.com

Research Paper

Planetary boundary layer height retrieval using GNSS Radio Occultation

over Egypt
Mohamed Zhran a,⇑, Ashraf Mousa b
a
Public Works Engineering Department, Faculty of Engineering, Mansoura University, El-Mansoura, Egypt
b
National Research Institute of Astronomy and Geophysics, Helwan, Egypt

a r t i c l e i n f o a b s t r a c t

Article history: Global Navigation Satellite System (GNSS) Radio Occultation (RO) is an excellent remote sensing tech-
Received 25 October 2021 nique for sounding the atmosphere. GNSS satellites send signals through the atmosphere of the Earth,
Revised 24 March 2022 and a GNSS receiver mounted on a Low Earth Orbit (LEO) satellite receives them. Through GNSS RO,
Accepted 25 March 2022
refractivity will be estimated, which is considered the main parameter in the determination of planetary
boundary layer height (PBLH). This paper presents the use of GNSS RO for the detection of PBLH over
Egypt.
Keywords:
PBLH is an important parameter to characterize the lower troposphere and plays a significant role in air
GNSS Radio Occultation
Planetary boundary layer height
quality adjustment. GNSS RO can be used to detect PBLH because of its high spatial and temporal cover-
ECMWF model age. The estimated PBLHs from GNSS RO are compared with PBLH estimated from the European Centre
Low Earth Orbit for Medium-range Weather Forecasts (ECMWF) model. This paper introduces the variation of the PBLH
over Egypt for the period from 2006 to 2019. GNSS RO PBLH distribution is consistent with that of
ECMWF. The maximum value of PBLH (4.75 km) occurs in the summer and its minimum value
(0.43 km) occurs in the winter. PBLH shows higher values around 15:00 LT (late afternoon) in Egypt.
Our results show that GNSS RO is a promising technique for sensing the atmosphere’s dynamics.
Ó 2022 National Authority of Remote Sensing & Space Science. Published by Elsevier B.V. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction straight line because of the change in density of the neutral atmo-
sphere before the receiver on LEO picks up the signals. The effect of
Planetary boundary layer height (PBLH) is the height of the the atmosphere can be expressed by a bending angle (a) as a func-
planetary boundary layer (PBL), and it is a necessary factor for tion of the impact parameter (a).
studying climate and weather. PBL is the lowest tropospheric The excess phase is the basic measurement of radio occultation,
region. The PBLH varies from a few hundred meters to a few kilo- and through the Doppler shift, one can get bending angles on L1
meters. Free troposphere located above PBL as shown in Fig. 1. and L2 with knowledge of the precise GNSS satellite orbit determi-
PBLH affects meteorology and air pollution and is used in numeri- nation. The bending angle is converted to a refractivity profile
cal weather prediction models (Garratt, 1992; Stull, 1988; Arya, through Abel inversion with the assumption of the atmosphere’s
2001). spherical symmetry (Sokolovskiy, 2001). Then, ignoring scattering
PBLH can be detected using Global Navigation Satellite System and moisture and applying the ionospheric correction to remove
(GNSS) Radio Occultation (RO), which gives atmospheric vertical the ionospheric contributions, the dry temperature profile can be
profiles with high vertical resolution. GNSS RO is an excellent derived from the refractivity as shown in Fig. 2. The GNSS RO
remote sensing manner for sounding the atmosphere. remote sensing technique has the advantage of determining PBLH
GNSS RO depends mainly on the refraction where the signals over both deserts and seas.
are transmitted by GNSS satellites and received by a receiver on A refractivity gradient will be used to detect PBLH, which is
Low Earth Orbit (LEO) satellites. GNSS signals propagation is obtained from the differentiation of refractivity. GNSS RO data is
affected by the atmosphere. GNSS signals are refracted from a processed using version 9 of the Radio Occultation Processing
Package (ROPP V.9). This paper introduces the distribution of PBLH
over Egypt. The paper presents studying the spatial and temporal
Peer review under responsibility of National Authority for Remote Sensing and
Space Sciences. variability of PBLH. We validate the PBLH obtained from GNSS
⇑ Corresponding author. RO with the PBLH obtained from the European Centre for
E-mail address: mohamedzhran@mans.edu.eg (M. Zhran).

https://doi.org/10.1016/j.ejrs.2022.03.013
1110-9823/Ó 2022 National Authority of Remote Sensing & Space Science. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 1. Planetary boundary layer height definition.

Fig. 2. GNSS – LEO satellite occultation principal.

Medium-range Weather Forecasts (ECMWF) operational daily Fig. 3. Distribution of Radio occultation events for PBLH from 2006 to 2019 over
analysis. Egypt.

2. Data sources
ECMWF provides a global set of atmospheric analyses data con-
taining atmospheric quantities on a temporal and spatial grid.
Fourteen years of data over Egypt with a latitude and longitude
These parameters are obtained with the ECMWF data assimilation
range from 22° N to 32° N and 25° E to 37° E are used. GNSS RO
system from various types of measurements (satellite data, ship,
data from the Constellation Observing System for the Meteorology,
and aircraft reports) together with a global weather forecast model
Ionosphere, and Climate (COSMIC) mission is the primary data set
(Technical Attachment of ECMWF, 1994;
used in this study for the determination of PBLH. The data were
Steiner, 1998). The ECMWF operational analyses are widely
obtained from the COSMIC RO Data Analysis and Archive Center
used for several applications in atmospheric research and in
(CDAAC; https://cdaac-www.cosmic.ucar.edu/cdaac/index.html). numerical weather prediction. This data is used for validation of
RO profiles include fourteen years from January 2006 to December the PBLH calculated from RO. However, this dataset does not con-
2019. Fig. 3 shows the RO events distribution which can be used to tain any direct information on the PBLH. The co-located ECMWF
determine PBLH in this period. The number of profiles that can be background profiles over the time span of January 2016–December
solved to detect PBLH during the previous period is about 712 RO 2017 will be used for validating the results from RO data. These
profiles out of 21,296 profiles, i.e., 97 % of the profiles cannot be background profiles are at the same times and locations as the
solved for determination of PBLH. This failure is due to the fact that observed data and are independent of RO profiles.
GNSS RO signals cannot be delivered to the terrain because of the
variability of water vapor in the atmosphere and the distortion of
the signals. 3. PBLH detection using refractivity gradient
Cosmic RO receivers have the advantage of the signal capability
to penetrate deeper into the lower troposphere because cosmic GNSS RO signals can penetrate clouds and precipitation
satellites use the open-loop (OL) tracking technique (Sokolovskiy (Solheim et al., 1999), which is very necessary at the top of PBL.
2001; Ao et al., 2009). The high vertical resolution of GNSS RO profiles approximately
552
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

100 m (Gorbunov et al., 2004) can be used to detect PBLH. These the upper boundary in the Abelian integral being infinity as seen
properties make GNSS RO a promising technique for estimating in Eq.2. Now the refractive index will be converted to refractivity
PBLH (Guo et al., 2011; Ao et al., 2012; Xie et al., 2012; Chan and through Eq.3:
Wood, 2013; Ho et al., 2015).
The PBLH determination mainly depends on the refractivity gra- N ¼ ðn  1Þ  106 ð3Þ
dient. A refractivity gradient will be produced through the differ- Refractivity depends mainly on pressure, temperature, and
entiation of the atmospheric refractivity profile. PBLH is assumed water vapor (Smith and Weintraub, 1953), and variations of these
to be the height where the minimum refractivity gradient occurs parameters will affect refractivity and will lead to the variability of
(e.g., Ao et al., 2008; Ao et al., 2012; Ao et al., 2019; Xie et al., PBLH over space and time. PBLHN, as the height of minima of the
2012; Ho et al., 2015). vertical derivative of the refractivity, can be given as:
Processing steps for the determination of PBLH from GNSS RO  
@N
observations are summarized in the following block diagram PBLHN ¼ arg min ð4Þ
(Fig. 4). @h
As shown in the previous block diagram, GNSS RO processing Fig. 5 shows an example of the refractivity profile obtained from
starts with the excess phase to compute the bending angle on L1 Abelian transformation as expressed by Eq.2 and Eq.3 at a specific
and L2 through Doppler shift measurements. The excess phase rep- event.
resents the difference between the actual ray path and the vacuum Fig. 5 illustrates the location of the maximum refractivity as
path length. By applying the ionospheric correction, the neutral about 270 N-units at a height of 946 m, and the refractivity
bending angle will be estimated. Vorob’ev and Krasil’nikova decreases as the height increases. This event consists of 2866 tan-
(1994) proposed the linear combination of the two bending angles gent points.
ða1 anda2 Þ at the same impact parameters as presented in Eq.1:
4. Results and discussion
2
f1 a1 ðaÞ  a2 ðaÞ 2
f2
ac ðaÞ ¼ 2 2
ð1Þ 4.1. Analysis of GNSS RO PBLH
f1  f2
a1 ðaÞ and a2 ðaÞ are the uncorrected bending angles of L1 and L2 The main purpose of this study is to investigate the spatial and
signals. Using the Abelian transformation, the refractive index will temporal variation of PBLH over Egypt. The spatial variation of
be retrieved as shown in Eq.2 (Kursinski et al., 1997): PBLH is introduced. PBLH is plotted according to their latitude
and longitude from 2006 to 2019, as shown in Fig. 6.
" # Fig. 6 also illustrates that PBLH varies with latitude and longi-
Z
1 1
ac ðaÞ tude without a constant trend. Fig. 7 depicts the contour of the
nðr t Þ ¼ exp pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi da ð2Þ
p a1 a2  a1 2 PBLH for Egypt using the GNSS RO data. This figure clearly shows
the spatial distribution of zonal mean PBLH over Egypt from
Equation (2) is called an Abelian transformation and is used to 2006 to 2019. The minimum and maximum values of the zonal
convert bending angle to refractive index (Fu, 2011). Statistical mean PBLH for the period from 2006 to 2019 are 1.1 km and
optimization must be applied to profiles of bending angle due to 2.6 km, respectively.

Fig. 4. Block diagram for GNSS RO observations processing for deduction PBLH.

553
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 5. Refractivity profile at an event.

PBLH at four seasons (averaged 3-months seasonal mean (DJF, Fig. 7. Contour of PBLH over fourteen years.
MAM, JJA, and SON)) is derived from the year 2006 to 2019 using
GNSS RO as shown in Fig. 8.
Fig. 8 indicates that during winter, PBLH changes mainly
between 1.2 km and 1.7 km, with a small area to the west showing The analysis of the results shows that the highest value of PBLH
a higher value of about 2.7 km. On the other hand, during the is about 4.75 km and occurs in the summer at 15:00 LT, while the
spring season, PBLH varies mainly from 1.4 km to 2.9 km, with a lowest value of PBLH that happens in the winter is about 0.43 km
small area in the southeastern part of the Red Sea of about at 21:00 LT over Egypt.
0.9 km. During the summer season, PBLH changes mainly between To investigate PBLH seasonal variability, we calculated mean
1.4 km and 2.9 km, with a higher value of 3.9 km in a small area in PBLH at specific times (03:00, 9:00, 15:00, and 21:00 LT) in differ-
the southeastern Red Sea. On the other hand, during the autumn ent seasons. Fig. 10 illustrates that PBLH’s mean value in spring is
season, PBLH changes mainly between 1.4 km and 2.4 km. Summer higher than its mean value in autumn. PBLH shows higher values
and spring show the same range of variation in general. around 15:00 LT in Egypt, and it is similar in all seasons as shown
For studying the seasonal variation of PBLH at different specific in Fig. 10.
local times (03:00, 09:00, 15:00, and 21:00 LT), for example, PBLH Fig. 10 shows that the highest PBLH occurs in the afternoon and
at MAM for this different local time is shown in Fig. 9. Local Time the lowest PBLH happens at night. As shown in the previous figure,
(LT) for Egypt equals Universal Time (UT) plus 02:00. PBLH significantly varies with time. PBLH increases from morning
Around dawn, during spring, the PBLH changes between 0.9 km to afternoon and decreases towards night.
and 3.7 km with a small area to the south, the PBLH reaches 4.2 km.
During the spring season around the morning, PBLH varies from 4.2. Comparison of PBLH from GNSS RO and ECMWF model (the
0.7 km to 2.7 km with a small area in the northern part of the Gulf background)
of Suez. Around the afternoon in the spring season, PBLH generally
varies from 1.2 km to 3.2 km. At night in the spring season, PBLH After retrieving the PBLH from the GNSS RO, the accuracy of the
changes from 1.1 km to 3.1 km. The previous figure indicated that PBLH needs to be evaluated. The PBLH calculated from the GNSS
PBLH in desert regions increases during the daytime. RO data is compared with co-located profiles obtained from

Fig. 6. Spatial distributions of PBLH over 14-years.

554
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 8. Seasonal zonal averages of PBLH over Egypt from GNSS RO data from 2006 to 2019. Units in [Km].

ECMWF for 2016 and 2017. The numbers of profiles used in the PBLH from ECMWF analysis in 2016 and 2017, respectively. Fig. 12
comparison are 220 and 142 profiles for 2016 and 2017 respec- shows the comparison between the monthly mean PBLH from RO
tively. Comparisons of PBLH determined from GNSS RO data and and the monthly mean PBLH from ECMWF analysis.
the ECMWF model in 2016 and 2017 are shown in Fig. 11. In general, the RO tends to overestimate the PBLH compared to
Fig. 11 ensures that PBLH determined from GNSS RO and its the ECMWF model. As it is clear from Fig. 12, there is a good match
value from ECMWF is well matched. The overall average difference with the monthly average PBLH in 2016 and 2017 as seen in Table 1
is about 0.235 km and 0.14 km in 2016 and 2017, respectively. The and Fig. 12. The standard deviation of the monthly mean PBLH
linear Pearson correlation coefficient between PBLH calculated from RO is shown in Fig. 13, where the means are represented by
from GNSS RO and PBLH estimated from the ECMWF model is the circles and the standard deviations are represented by the error
about 0.77 and 0.65 in 2016 and 2017, respectively. It is to be noted bars.
that GNSS RO PBLH distribution is well-matched with that of The mean deviation between PBLH from RO and PBLH com-
ECMWF. puted from ECMWF analyses is referred to as bias. The overall aver-
Here the values of monthly mean PBLH are investigated in 2016 age difference is about 0.235 km and 0.14 km in 2016 and 2017
and 2017. Table 1 summarizes the monthly mean PBLH from RO respectively, with a standard deviation of the mean of 0.56 km
and ECMWF in 2016 and 2017. and 0.59 km in 2016 and 2017 respectively. Fig. 14 shows the scat-
High correlation coefficients of 0.803 and 0.744 were observed ter around the mean difference. In most cases, the scatter around
between the monthly mean PBLH from RO and the monthly mean the mean is within the standard deviation. From the above analy-

555
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 9. MAM Seasonal zonal averages of PBLH over Egypt from GNSS RO data at 3:00, 9:00, 15:00, and 21:00 LT (Local Time) from 2006 to 2019. Units in [Km].

sis, it can be concluded that there is very good agreement between

the RO and the ECMWF model, and this indicates that RO is a
promising remote sensing technique for studying the atmosphere
and determining PBLH.

5. Conclusions

PBLH is a principal guideline for studying climate and weather.

The high vertical resolution of GNSS RO profiles can be used to
detect PBLH. GNSS RO can detect PBLH with high spatial and tem-
poral resolution over land and sea. From the refraction of the sig-
nals transmitted by an occulted GNSS and received by a LEO, one
can get the bending angle. Through Abel transformation bending
angle is converted to refractivity assuming spherical symmetry of
atmosphere. Refractivity is the cornerstone of the detection of
PBLH. We generate a PBLH based on the refractivity gradient from
GNSS RO data over the years 2006 to 2019. PBLH is often high in
the afternoon and low at night. The spatial and temporal variation
Fig.10. Seasonal variation of mean PBLH over Egypt from GNSS RO data at 3:00, of PBLH over Egypt will help in modeling the atmosphere and con-
6:00, 15:00, and 21:00 LT (Local Time) from 2006 to 2019. Units in [Km]. trolling air pollution. The analysis of the results illustrates that
556
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 11. Comparison of PBLH determined from GNSS RO data and ECMWF model in 2016 and 2017.

Table 1
Monthly mean PBLH from RO and monthly mean PBLH from ECMWF analysis.

Year 2016 Year 2017

Month Monthly Monthly Monthly Monthly

Mean RO Mean ECMWF Mean RO Mean ECMWF

PBLH (km) PBLH (km) PBLH (km) PBLH (km)
January 1.72 1.41 1.47 1.33
February 1.23 1.28 1.52 1.41
March 1.49 1.21 1.19 1.09
April 1.51 1.47 1.24 1.44
May 1.59 1.25 2.01 1.43
June 1.64 1.50 1.61 1.58
July 1.73 1.38 1.92 1.63
August 1.81 1.46 1.90 1.55
September 1.83 1.65 1.36 1.03
October 1.21 0.94 1.85 1.60
November 1.36 1.05 1.66 1.45
December 1.64 1.41 1.26 1.17

Fig. 12. Comparison between monthly mean PBLH from RO and monthly mean PBLH from ECMWF analysis for years 2016 and 2017.

PBLH is at its maximum in the summer season and its value is time of day. The PBLH calculated from RO shows good agreement
about 4.75 km at 15:00 LT. Results also showed that the minimum with the results of ECMWF. The overall average difference between
value of PBLH was in winter, about 0.43 km at 21:00 LT, and this RO and ECMWF results is about 0.235 km and 0.14 km in 2016 and
proves that PBLH has a large time variation with the season and 2017 respectively. The results of this paper ensure that the GNSS

557
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

Fig. 13. The standard deviation of monthly mean PBLH from RO for years 2016 and 2017.

Fig. 14. The overall average difference for year 2016 and 2017.

RO technique is an excellent method for remote sensing the atmo- Arya, S. P., 2001. Introduction to Micrometeorology Second Edition.
Chan, K.M., Wood, R., 2013. The seasonal cycle of planetary boundary layer depth
sphere and can be used to detect PBLH over both land and sea with
determined using COSMIC radio occultation data. J. Geophys. Res.-Atmos. 118
good coverage and high temporal resolution. (22), 12,422–12,434. https://doi.org/10.1002/2013JD020147.
Fu, E., 2011. An Investigation of GNSS Radio Occultation Atmospheric Sounding
Declaration of Competing Interest Technique for Australian Meteorology. Ph.D thesis. Engineering and Health
RMIT University.
Garratt, J.R., 1992. The atmospheric boundary layer. Cambridge Atmospheric and
The authors declare that they have no known competing finan- Space Science Series, Volume 416. Cambridge University Press, Cambridge, UK,
cial interests or personal relationships that could have appeared p. 444.
Gorbunov, M.E., Benzon, H.-H., Jensen, A.S., Lohmann, M.S., Nielsen, A.S., 2004.
to influence the work reported in this paper. Comparative analysis of radio occultation processing approaches based on
Fourier integral operators. Radio Sci. 39 (6), n/a–n/a. https://doi.org/10.1029/
2003RS002916.
Guo, P., Kuo, Y.-H., Sokolovskiy, S.V., Lenschow, D.H., 2011. Estimating atmospheric
boundary layer depth using COSMIC radio occultation data. J. Atmos. Sci. 68,
References 1703–1713.
HO, S.-P., L. Peng, R. A. Anthes, Y.-H. Kuo, and H.-C. Lin, 2015. Marine boundary layer
Ao, C.O., Chan, T.K., Iijima, B.A., Li, J.-L., Mannucci, A.J., Teixeira, T., Tian, B., Waliser, heights and their longitudinal, diurnal and inter-seasonal variability in the
D.E., 2008. Planetary boundary layer information from GPS radio occultation southeast Pacific using COSMIC, CALIOP, and radiosonde data. J. Climate, 28,
measurements, ECMWF GRAS SAF Workshop on Applications of GPS Radio 2856–2872, https://doi.org/10.1175/JCLI-D-14-00238.1.
Occultation Measurements (16–18 June 2008). Reading, UK. Kursinski, E.R., Hajj, G.A., Hardy, K.R., Schofield, J.T., Linfield, R., 1997. Observing
Ao, C.O., Hajj, G.A., Meehan, T.K., Dong, D., Iijima, B.A., Mannucci, A.J., Kursinski, E.R., earth’s atmosphere with radio occultation measurements. J. Geophys. Res. 102,
2009. Rising and setting GPS occultations by use of open-loop tracking. J. 23429–23465.
Geophys. Res. 114, D04101. https://doi.org/10.1029/2008JD010483. Smith, E.K., Weintraub, S., 1953. The constants in the equation for atmosphere
Ao, C.O., Waliser, D.E., Chan, S.K., Li, J.-L., Tian, B., Xie, F., Mannucci, A.J., 2012. refractive index at radio frequencies. Proc. IRE 41, 1035–1037.
Planetary boundary layer heights from GPS radio occultation refractivity and Sokolovskiy, Sergey V., 2001. Tracking tropospheric radio occultation signals from
humidity profiles. J. Geophys. Res.-Atmos. 117 (D16), n/a–n/a. https://doi.org/ low Earth orbit. Radio Sci. 36 (3), 483–498. https://doi.org/10.1029/
10.1029/2012JD017598. 1999RS002305.
Ao, C.O., Wang, K.N., Kubar, T.L., 2019. Characterizing the Vertical Stratification of Solheim, Fredrick S., Vivekanandan, Jothiram, Ware, Randolph H., Rocken, Christian,
the Planetary Boundary Layer from Space with GNSS Radio Occultation vol, 1999. Propagation delays induced in GPS signals by dry air, water vapor,
2019.

558
M. Zhran and A. Mousa Egypt. J. Remote Sensing Space Sci. 25 (2022) 551–559

hydrometeors, and other particulates. J. Geophys. Res.-Atmos. 104 (D8), 9663– Vorob’ev, V.V., Krasil’nikova, T.G., 1994. Estimation of the accuracy of the
9670. https://doi.org/10.1029/1999JD900095. atmospheric refractive index recovery from Doppler shift measurements at
Steiner, A. K., 1998. High resolution sounding of key climate variables using the frequencies used in the NAVSTAR system. Phys. Atmos Ocean 29, 602–609.
radio occultation technique. PhD thesis. University of Graz, Austria: Institute for Xie, D. L. Wu, C. O. Ao, A. J. Mannucci, and E. R. Kursinski, 2012. Advances and
Geophysics, Astrophysics, and Meteorology. limitations of atmospheric boundary layer observations with GPS occultation
Stull, R.B., 1988. An Introduction to Boundary Layer Meteorology; Kluwer Academic over southeast Pacific Ocean. Atmos. Chem. Phys., 12, 903–918, doi:10.5194/
Publishers: Norwell. MA, USA; p, p. 666. acp-12-903-2012.
Technical Attachment of ECMWF, 1994. The Description of the ECMWF/WCRP Level
III-A Global Atmospheric Data Archive. ECMWF.

559