Arabian Journal of Geosciences (2021) 14: 13

https://doi.org/10.1007/s12517-020-06251-4

S. I. GEOLOGY OF AFRICA

Edge detection of aeromagnetic data as effective tools for structural

imaging at Shilman area, South Eastern Desert, Egypt
Ahmed Mohammed Eldosouky 1 & Hassan Mohamed 2

Received: 15 April 2020 / Accepted: 13 November 2020 / Published online: 6 January 2021
# Saudi Society for Geosciences 2021

Abstract
Aeromagnetic data over an extending part of Shilman area, South Eastern Desert (SED) of Egypt have been processed for
enhancing the edges of magnetic sources to give a clear understanding of the structural framework. Various edge detectors
including 1st vertical derivative (1st VD), total gradient (TG), and Theta were utilized to define the boundaries of the magnetic
origins. These methods are applied to a synthetic model firstly. The interpretation exhibits that the N-S and NW directions are
dominant in 1st VD map. The NW structural trend is dominant in the TG map while NW and N-S are the main structural trends
controlling the Theta map of Shilman area. The correlation between the utilized techniques indicated that the Theta method is
more effective in detecting the edges of magnetic sources of Shilman area than those of 1st Vd and TG methods. These results
display that the Theta method is a powerful tool for enhancing the source edges of magnetic data.

Keywords Aeromagnetic . 1st vertical derivative . TG . Theta . Edge detection

Introduction structures like lineaments, faults, rock boundaries, and ore ma-
terials. Recently, numerous methods have been proposed to
Magnetic data considers one of the common advantageous promote the representation process of magnetic anomalies and
available tools that help in the identification of the surface the description of the patterns of the source bodies despite their
and subsurface geologic structures. The aeromagnetic analysis depths. One of the primary improvement filters is the reduction
is utilized to help in explaining the problems of provincial to the magnetic pole (RTP) which was first proposed by
geologic imaging and structure (Sehsah et al. 2019; Baranov (1957) to solve the asymmetrical configuration of
Eldosouky 2019; Eldosouky et al. 2020c), representation of anomalies by representing the anomalies with vertical magne-
rock boundaries and edges (Pham et al. 2019; Eldosouky et al. tization. The common improvement methods that have been
2020b; Eldosouky and Saada 2020), positioning of the pre- routinely employed in the magnetic data are vertical and hori-
sumable fields of rock differentiation, and mineralization zontal derivatives (Arisoy and Dikmen 2013) which have been
(Eldosouky et al. 2017, 2020a; Ekwok et al. 2019). practiced to indicate the edges of magnetization frames (Oruç
Detection of edges has been extensively used in the geo- 2011). These methods are robust in the understanding that they
physical investigation, principally in mapping geologic can overcome the low-quality data and need moderate compu-
tational work (Pilkington and Tschirhart 2017). Some conven-
tional methods of this group are the total gradient (Roest et al.
This article is part of the Topical Collection on New Advances and 1992) and the total horizontal gradient (Cordell and Grauch
Research Results on the Geology of Africa 1985). These techniques have been practiced to delineate caus-
ative origins in aeromagnetic data (Bastani and Pedersen 2001),
* Ahmed Mohammed Eldosouky besides being applied as constituents to develop other routines
[email protected];
[email protected] (Cooper 2009). Later on, various methods have introduced for
detecting edges of magnetic sources such as Theta (Wijns et al.
1
2005; Eldosouky 2019) and horizontal tilt derivative method
Geology Department, Faculty of Science, Suez University,
Suez 43518, Egypt
(Cooper and Cowan 2006).
2
In the existing study, we employ 1st VD, TG, and Theta to
Geophysics Department, Egyptian Nuclear Materials Authority, New
Cairo, Egypt
aeromagnetic data of Shilman area, SED, Egypt, to detect
13 Page 2 of 10 Arab J Geosci (2021) 14: 13

Fig. 1 Synthetic model. (a)

Horizontal locations for the
source bodies; (b) vertical loca-
tions for the source bodies; and (c)
three-dimensional perspective
view for model

edges of magnetic sources and to give a well understanding of detection of the edges constitutes a primary step in the process
the structural framework of the study area. of magnetic data analysis. The first vertical derivative (1st
VD) method has been utilized broadly to map the susceptibil-
ity variations.
Methodology 1st VD filter (Foss 2011) is commonly performed to
gridded data using the fast Fourier transform method.
Various techniques are concerned with the edge detection of Certain VD of the magnetic field can be measured by multi-
aeromagnetic anomalies caused by geologic structures. Now, plying the amplitude spectra of the domain by a factor of the
Arab J Geosci (2021) 14: 13 Page 3 of 10 13

form: Table 1 Location for each buried prism (synthetic model)

1 h
2 1=2 in Model parameters Prism (1) Prism (2) Prism (3)

VD ¼ e þ r2 ð1Þ
i X (km) 3 7 18 22 33 37
where i = the VD order, e, r is the wavenumber corresponding
Y (km) 1 21 1 21 1 21
to the x, y directions, respectively. 1st VD has become almost
Z (km) 8 12 4 12 1 12
an essential element in magnetic improvement projects. This
I 90o 90o 90o
derivative of the anomaly with regard to the depth of the
D 0o 0o 0o
source gives a commending resolution of the anomaly itself.
M 10 A/m 10 A/m 10 A/m
Estimation of the 1st VD is used for enhancing near-surface
bodies, overcoming lower ones, and providing sufficient res-
olution of closely spaced bodies (Foss 2011(.
Total gradient (TG) is a modernized gradient method, Figure 1b shows the change of depth with each source
which is linked to magnetic domains by the derivatives. The body. The depth to the top of each block from the left is 8
amplitude of the TG can be calculated from the three orthog- km, 4km, and 1 km, respectively. Model-1 and location of the
onal components of the magnetic data employing the follow- observation points are presented in 3-D view (Fig. 1c).
ing expression, as Roest et al. (1992) explained: Figure 2 shows the calculated magnetic field data; the mag-
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
     ffi netic effect of each prism is gradually changing according to
∂M 2 ∂M 2 ∂M 2 the depth of each body.
J ðx;yÞ ¼ þ þ ð2Þ
∂x ∂y ∂z

where J(x, y) is the TG amplitude at x, y.

Shilman study area
TG is applied to define the geometry of the magnetic
The chief rock types in Shilman area are described by
sources, Roest et al. (1992) said.
ophiolites, granodiorite, metasediments, and metavolcanics
Cos (θ) map (Theta) was introduced by Wijns et al. (2005)
(Fig. 3). The ophiolitic rocks are intruded by granodiorite
for mapping the edges of magnetic sources. It is the proportion
and being covered by the sequence of arc metavolcanic–
between total horizontal derivative (THD) (Blakely and
metasediment (Kröner et al. 1987). The ophiolites include
Simpson 1986) and TG (Roest et al. 1992) and is given by:
talc-carbonates, peridotite, metagabbros and amphibolites
 
THD (Abd El-Naby et al. 2000; Abdeen and Abdelghaffar 2011).
Theta ðCosθÞ ¼ ð3Þ
TG The most significant serpentinite mass occurs at Shilman area
generating a thrust nappe, which is associated with a consid-
Theta technique can imagine edges according to amplitude erable metagabbro source. Older granites occur to the eastern
and strike and amplitude (Oruç and Keskinsezer 2008; of Gebel Shilman while younger ones appear in small locali-
Eldosouky 2019). ties at the northwestern part. The central part of Shilman area
is occupied by the NW trending metavolcanics. Small traces
of metasediments can be observed in the northwestern part of
Synthetic data the Shilman area (Abd El-Naby et al. 2000; Abdeen and
Abdelghaffar 2011) as presented in Fig. 3.
A dimensional forward model is used to calculate the magnet- The Shilman area exposed the remains of an overthrusted
ic field anomaly and application of applied edge detectors. hot body of back-arc oceanic crust, which produces a meta-
The model has three magnetized bodies (Fig. 1a). All buried morphic sole below an allochthonous ultramafic body, Abd
bodies assigned with the same length and width, 20 km and 4 El-Naby et al. (2000) concluded.
km, respectively. The location of each prism for the model is Figure 4 shows the reduced to pole (RTP) data of the
presented in Table 1. The ambient field has an inclination of Shilman area (Aero-Service 1984). The RTP map shows var-
90o and a declination of 0o (Table.1). One kilometer grid in- iations in the magnetic intensities from 250 (highs) to − 260
terval is used to calculate the surface magnetic anomaly. nT (lows) nT with a strong NW structural trend.
The model region extends 40 km and 22 km along eastern
and northern direction, respectively. Total magnetic field at
each 943 data points collected over the model region with
Results and discussion
41 × 23 grid of 1 km interval along x and y directions is
calculated (Fig. 1). The magnetization intensity of the normal
The first scenario includes the application of edge detectors
fault is assumed to be constant 10 A/m (Table 1).
(ED) to a synthetic model with three prisms located at various
13 Page 4 of 10 Arab J Geosci (2021) 14: 13

Fig. 2 The total magnetic anomaly produced by model (1) using GLQ method; the inducing field has direction I = 90o and D = 0o

altitudes (Fig. 2). Figure 5 shows the application of 1st VD, not provide a definite result for further deep-seated sources.
TG, and Theta to the magnetic anomaly of the model. The While we can notice that the Theta method gives a clear and
detection of the source edges (DSE) method calculates the definite detection for the edges of shallow and deep prisms
edges of 1st VD, TG, and Theta grids from the RTP data as (Figs. 5c and 6c).
shown in Fig. 6. As depicted in Figs. 5a, 5b, 6a, and 6b, the 1st Afterwards, the ED (1st VD, TG, and Theta) are imple-
VD and TG can imagine the edges of shallow sources and do mented to the RTP grid of Shilman area to map the boundaries

Fig. 3 Location and geology of

Shilman area (after Eldosouky
et al. 2017)
Arab J Geosci (2021) 14: 13 Page 5 of 10 13

Fig. 4 RTP map of Shilman area

of various magnetic sources (Fig. 7). DSEs are calculated and Shilman area producing the N-S direction is recent, while the
represented in Fig. 8. As represented in Figs. 7 and 8, the NW trend is the predominant tectonic framework at Shilman
Theta method (Figs. 7c and 8c) reveals the effects of both deep area (Eldosouky and Elkhateeb 2018). Moreover, the Theta
and shallow sources than the 1st VD and TG (Figs. 7a, 8a, 7b, map (Fig. 7c) gives more detailed structural and anomalous
and 8b) and gives more detailed information about rock features than those present in the RTP map (Fig. 4).
boundaries and edges (faults and contacts). The delineated Note also the appearance of powerful edges characterized
structural edges map (Fig. 8) also points out that the N-S by the magnetic structures. So, with the correlation between
direction is dominant in 1st VD map (Fig. 9a), and NW and the Theta filter and 1st VD and TG of Shilman area, it is clear
NE directions represent the foremost dominant structural that the source edges highlighted by Theta method improve
trends in Shilman area (Fig. 9b). Figure 9c shows that the the main magnetic structures making these structures more
NW, N-S, and NNE are the main predominant structural ele- discernible. Moreover, the edges in the Theta map appear to
ments that control Shilman area. These tectonic frameworks be more interconnected and also many structural trends are not
match well with the recognized ones mentioned by Eldosouky depicted by 1st VD and TG methods which are now sharply
et al. (2017). This indicates that the tectonic regime affecting distinguished (i.e., NNE direction). Theta maps (Figs. 7c and
13 Page 6 of 10 Arab J Geosci (2021) 14: 13

Fig. 5 (a) First vertical derivative

(1st VD ) map of the synthetic
model (Fig. 2); (b) total gradient
(TG) map of the synthetic model
(Fig. 2); and (c) Theta map of the
synthetic model
Arab J Geosci (2021) 14: 13 Page 7 of 10 13

Fig. 6 (a) Detection of source

edges of 1st VD (Fig. 5a); (b) de-
tection of source edges of TG
(Fig. 5b); and (c) detection of
source edges of Theta (Fig. 5c)
13 Page 8 of 10 Arab J Geosci (2021) 14: 13

Fig. 7 Applied edge detectors to RTP data (Fig. 4): (a) 1st VD map of Shilman area; (b) TG map of Shilman area; and (c) Theta map of Shilman area

Fig. 8 (a) Detection of source edges of 1st VD of Shilman area (Fig. 7a); (b) detection of source edges of TG of Shilman area (Fig. 7b); and (c) detection
of source edges of Theta of Shilman area (Fig. 7c)

Fig. 9 (a) Rose diagram of Fig. 8a; (b) rose diagram of Fig. 8b; and (c) rose diagram of Fig. 8c
Arab J Geosci (2021) 14: 13 Page 9 of 10 13

8c) also show a cleaner model for imaging edges and struc- Bastani M, Pedersen LB (2001) Automatic interpretation of magnetic
dike parameters using the analytical signal technique. Geophysics
tures of Shilman area compared to the 1stVD and TG filters
66(2):551–561
(Figs. 7a, 8a, 7b, and 8b). Blakely RJ, Simpsom RW (1986) Approximating edges of source bodies
from magnetic or gravity anomalies”. Geophysics 51:1494–1498
Cooper GRJ (2009) Balancing images of potential-field data. Geophysics
74(3):L17–L20
Conclusion Cooper GRJ, Cowan DR (2006) Enhancing potential field data using
filters based on the local phase. Comput Geosci 32(10):1585–1591
The existing study deals with the structural imaging based on Cordell L, Grauch VJS (1985) Mapping basement magnetization zones
edge detection of a synthetic model and magnetic sources of from aeromagnetic data in the San Juan Basin. New Mexico. In:
Hinze WJ (ed) The utility of regional gravity and magnetic anoma-
Shilman area, SED of Egypt. Detecting the edges of magnetic
lies maps. Soc Explor Geophys 181–197. https://doi.org/10.1190/1.
sources represents a significant stage in structural imaging 0931830346.ch16
using magnetic data. The synthetic scenarios displayed that Ekwok SE, Akpan AE, Ebong DE (2019) Enhancement and modeling of
the Theta method gives edge definition better and presents aeromagnetic data of some inland basins, southeastern Nigeria. J Afr
Earth Sci 155:43–53
more intuitive pictures with improved data than 1st VD and
Eldosouky AM (2019) Aeromagnetic data for mapping geologic contacts
TG filters. at Samr El-Qaa area, North Eastern Desert, Egypt. Arab J Geosci 12:
The edge detection techniques were implemented to RTP 2. https://doi.org/10.1007/s12517-018-4182-2
data of Shilman area (1st VD, TG, and Theta) to imagine the Eldosouky AM, Elkhateeb SO (2018) Texture analysis of aeromagnetic
edges and structures. The results revealed that the N-S trend is data for enhancing geologic features using co-occurrence matrices in
Elallaqi area. South Eastern Desert of Egypt 7(1):155–161. https://
dominant in the 1st VD maps and the NW and NE directions doi.org/10.1016/j.nrjag.2017.12.006
were dominant in the TG maps. Eldosouky AM, Abdelkareem M, Elkhateeb SO (2017) Integration of
The inferred structures from Theta map show that the remote sensing and aeromagnetic data for mapping structural fea-
Shilman area is affected by NW, N-S, NNE, and NE tectonic tures and hydrothermal alteration zones in Wadi Allaqi area, South
Eastern Desert of Egypt. J Afr Earth Sci 10:1016
regimes. Application of Theta filter revealed more anomalies Eldosouky AM, Saada SA (2020) Source edge detection (SED) of aero-
and depicted further continuous linear edges than with the magnetic data synthetic examples and a case study from Haimur
other filters. In comparison with the 1st VD and TD edge area, south Eastern Desert, Egypt. Arab J Geosci 13, 626. https://
detectors, Theta has demonstrated excellent promise in pro- doi.org/10.1007/s12517-020-05653-8
Eldosouky AM, Sehsah H, Elkhateeb SO, Pour AB (2020a) Integrating
viding images with more extraordinary resolution and clear aeromagnetic data and Landsat-8 imagery for detection of post-
responses over the edges of magnetic sources in Shilman area. accretionary shear zones controlling hydrothermal alterations: The
The detection of source edges using Theta map filter gave Allaqi-Heiani Suture zone, South Eastern Desert. Egypt. Advances
sharp responses from shallow and deep sources of Shilman in Space Research 65:1008–1024. https://doi.org/10.1016/j.asr.
2019.10.030
area than other methods. As a result, in alike areas, it is ad- Eldosouky AM, Pham LT, Hassan P, Pradhan MB (2020b) A compara-
vised to use Theta filter for structural imaging with aeromag- tive study of THG, AS, TA, Theta, TDX and LTHG techniques for
netic data. improving source boundaries detection of magnetic data using syn-
thetic models: A case study from G. Um Monqul, North Eastern
Desert, Egypt. Journal of African Earth Sciences 170:103940
Compliance with ethical standards Eldosouky AM, Elkhateeb SO, Ali A, Kharbish S (2020c) Enhancing
Linear features in Aeromagnetic data using directional horizontal
Conflict of interest There is no potential conflicts exist in the article. gradient at Wadi Haimur area. South Eastern Desert, Egypt,
Carpathian Journal of Earth and Environmental Sciences 15(2):
323–326
Foss C (2011) Magnetic Data Enhancements and Depth Estimation. In:
References Gupta H.K. (eds) Encyclopedia of Solid Earth Geophysics.
Encyclopedia of Earth Sciences Series. Springer, Dordrecht.
Abd El-Naby H, Frisch W, Hegner E (2000) Evolution of the Pan-African https://doi.org/10.1007/978-90-481-8702-7_104
Wadi Haimur metamorphic sole, Eastern Desert, Egypt. J Kröner A, Greiling R, Reischmann T, Hussein IM, Stern RJ, Kruger J,
Metamorph Geol 18:639–651 Duur S, Zimmer M (1987) Pan-African crustal evolution in the
Abdeen MM, Abdelghaffar AA (2011) Syn- and postaccretionary struc- Nubian segment of Northeast Africa. In: Kröner A (ed)
tures in the Neoproterozoic Central Allaqi-Heiani suture zone, Proterozoic Lithosphere Evolution. American Geophysical Union,
Southeastern Egypt. Precambrian Res 185:95–108 Washington D.C, pp 235–257
Aeroservice Co (1984) Final operational report of airborne magnetic/ Oruç B (2011) Edge detection and depth estimation using a tilt angle map
radiation survey in the Eastern Desert, Egypt. Aeroservice for from gravity gradient data of the Kozaklı-Central Anatolia region,
EGPC, Houston, Texas, April 1984 .Six volumes. An internal report Turkey. Pure Appl Geophys 168:1769–1780
.No.3609 Oruç B, Keskinsezer A (2008) Structural setting of the northeastern Biga
Arisoy M, Dikmen U (2013) Edge detection of magnetic sources using Peninsula (Turkey) from tilt derivatives of gravity gradient tensors
enhanced total horizontal derivative of the Tilt angle. Bull Earth Sci and magnitude of horizontal gravity components. Pure Appl
Appl Res Cent Hacet Univ 34:73–82 Geophys 165:1913–1927
Baranov V (1957) A new method for interpretation of aeromagnetic Pham LT, Oksum E, Do TD (2019) Edge enhancement of potential field
maps: pseudo-gravimetric anomalies. Geophysics 22:359–383 data usingthe logistic function and the total horizontal gradient. Acta
13 Page 10 of 10 Arab J Geosci (2021) 14: 13

Geod. Geophys 54(1):143–155. https://doi.org/10.1007/s40328- Sehsah H, Eldosouky AM, El Afandy AH (2019) Unpaired ophiolite
019-00248-6 belts in the Neoproterozoic Allaqi-Heiani suture, the Arabian-
Pilkington M, Tschirhart V (2017) Practical considerations in the use of Nubian shield: evidences from magnetic data. J Afr Earth Sci 156:
edge detectors for geologic mapping using magnetic data. 26–34. https://doi.org/10.1016/j.jafrearsci.2019.05.002
Geophysics 82(3):J1–J8. https://doi.org/10.1190/geo2016-0364.1 Wijns C, Perez C, Kowalczyk P (2005) Theta map: edge detection in
Roest W, Verhoef J, Pilkington M (1992) Magnetic interpretation using magnetic data. Geophysics 70(4):L39–L43
3-D analytical signal. Geophysics 57:116–125