G eophysi cal M appi ng of G hana U si ng Ad vanced Cartog raphi c Tool G M T

. Ghana is a country exceptionally rich in geologic mineral resources with contrasting topographic relief and varied geophysical setting. This paper evaluated the geological and geophysical setting of Ghana with a special focus on the impact of the geologic setting and topography on gravity. Specifically, it assessed how variations in geology, topography, landscapes and the environment control the geophysical parameters and how these vary among the major regions of the country – the Volta Basin, Northern Plains, Ashanti-Kwahu (Kumasi) and Coastal Plains in the Accra surroundings. Previous studies utilizing traditional Geographic Information System (GIS) approaches have documented the geologic evolution of Ghana evolved as a part of the West African Craton. As a contribution to the existing research, this paper presents a regional analysis of Ghana by integrated mapping of geology, geophysics and topography of the country. The technical approach of this research focuses on utilizing the console-based scripting cartographic toolset Generic Mapping Tools (GMT) integrated with QGIS for processing and mapping the datasets: General Bathymetric Chart of the Oceans (GEBCO), Earth Gravitational Model 2008 (EGM-2008), gravity grids. The theoretical background is based on the geologic research of West Africa supported by high-resolution data. The paper defines a conceptual cartographic framework for integrated geologic and geophysical visualization in a regional-scale mapping project on Ghana.


Introduction
Progress in geologic and geophysical analysis relies on advanced cartographic visualization. This deep ontological, conceptual and technical connection between the Earth science disciplines has led to a variety of developed cartographic software, algorithms and methods which help geologic and geophysical data modeling and mapping. The epoch ofbig geospatial data presents new technical challenges in contemporary cartography which aims at effective and precise data processing and visualization with minimized handmade routines and increased speed of mapping. Therefore, the use of scripting consolebased mapping in addition to a menu-based Geographic Information System (GIS), as presented in this paper, opens up new ways in the geological and geophysical mapping of Ghana.
Through this research, we were able to use the powerful functionality of both GMT and Quantum GIS (QGIS) to reflect the geophysical and geologic setting of the country from various data sources based on integrated approaches for accurate semiautomated mapping of Ghana, which is one of the significant advantages ofthis paper. Thus, this paper presented the combined cartographic method using GMT and QGIS for handling geospatial data to represent topographic, geologic and geophysical phenomena in Ghana with a twofold goal: The cartographic goal is to present a series of the five new thematic maps on Ghana prepared by the sequential use of the QGIS and GMT techniques: a topographic raster layer representing terrain relief and geophysical layers of geoid and gravity grids, an overlay and geologic vector map prepared in the QGIS environment.
The geophysical goal is to analyze the gravity anomaly fields, geoid and topographic elevations to demonstrate the effects of the geology on the surface topographic and geophysical setting. The comparative analysis of maps aims to illustrate the correlation between the topography, geophysics and geology of Ghana.

Mapping Software
Recently, geologic research has focused on advanced solutions for data analysis and mapping (Sylvester and Attoh 1992, Attoh 1998, Lindh and Lemenkova 2021, Allibone et al. 2002, Lemenkov and Lemenkova 2021b. However, previous studies used traditional methods of plotting for data visualization with limited functionality to perform mapping by automated approaches, such as Generic Mapping Tools (GMT). Among various methods in geosciences, data analysis and mapping are the most widely used approaches (Klaučo et al. 2013, Lemenkov andLemenkova 2021c). As for GIS, it remains a major method in mapping (Owusu-Nimo et al. 2018, Osei et al. 2021, Lemenkova et al. 2012, Suetova et al. 2005, Lemenkova 2021c, Gohl et al. 2006, Schenke and Lemenkova 2008. In contrast to the traditional GIS mapping methods, the GMT technique is less subjective and more functional, as it can perform mapping using repetitive scripts. However, due to the increased functionality of the scripting techniques for data processing, in particular in relation to the large datasets with global coverage, using a GMT cartographic toolset might be an effective way of cartographic data visualization (Wessel et al. 2019).
Although GIS software based on the conventional Graphical User Interface (GUI) can be used for mapping large amounts of geospatial data due to an easy interface and simpler functionality (Appiah 1991, Boher et al. 1992, Verutes et al. 2012, Eisenlohr and Hirdes 1992, Engstrom et al. 2013, there is an inherent potential for human-induced errors because of the workflow routine of mapping. Therefore, it is uncertain whether GIS will function effectively when processing multi-source large spatial data as is often the case in geologic mapping of Africa (Black and Fabre 1983, Loh and Hirdes 1996, 1999, Perrouty et al. 2012). The traditional GIS has a straightforward way of mapping through GUI and an extended menu with a variety of functions for vector and raster types of data. At the same time, a complex mapping project functions best when adopting all the available data formats which might include native .shp formats finely readable by a compatible and open source QGIS.
In contrast to the traditional GIS, GMT provides a console-based machine learning approach for spatial processing (Lemenkova, 2021a(Lemenkova, , 2021d. Flexibly combining these mapping approaches enables taking the advantages of both and increases mapping functionality. The sophisticated scripting of GMT is particularly well suited to cartographic visualization due to the extended functionality of the individual modules controlling map elements: finely adjusted colour palettes, legend placement and depiction, graticule ticks, clipping and translucency, compatibility with the Geospatial Data Abstraction Library (GDAL), hierarchical annotations using a variety of fonts, operating with various data formats, etc. For this reason, GMT is widely used in geophysics (Gauger et al. 2007, Lemenkova 2021a, 2019b, 2019c, Gorman et al. 2008).
Ghana was ranked the 1st country in Africa according to gold deposits in 2019 with 130 t of yearly gold production (Mining.com 2019). The orogenic gold deposits of Ghana were formed as a result of the longterm geologic evolution of the region which included the meta-sedimentary successions of Ghana form part of the Birimian Paleoproterozoic West African Craton (Chudasama et al. 2016, Kalsbeek et al. 2020, Davis et al. 1994, Feybesse et al. 2006, Block et al. 2016.
Established gold belts ofGhana include the regions of Ashanti Fault, Kumasi Basin and Sefwi-Bibiani. The orogenic gold mineral deposits in southern Ghana were extensively studied and resulted in regular reports on the distribution and origin of gold (see inter alia Adu-Baffour et al. 2021, Harcouët et al. 2007, Perrouty et al. 2016, Dzigbodi-Adjimah and Nana Asamoah 2009, Hilson 2002, Benshaul-Tolonen et al. 2019.
Other natural resources of Ghana include renewable energy potential presented by a variety ofsources (Ankrah and Lin 2020). The most significant is the hydropower sector, presented by the hydroelectric Akosombo (Volta) Dam on the Volta River which enabled the completion of construction of a plant for producing aluminium from bauxite at the supplemented Kpong Dam. Other hydro-energy resources of Ghana include the Lower Pra River Basin and Volta tributaries (Arthur et al. 2020). In addition, solar, geothermal and wind energy may be developed to become important energy sources in Ghana (Essandoh-Yeddu 1997, Danso et al. 2021, Nuru et al. 2021.
Central challenges in this study include the twofold aspects of cartographic data processing.
First, the present study is a data-driven project based on the open-source datasets covering Ghana. Recent technical progress in data capture and organized geological surveys resulted in the available materials, such as high-resolution GEBCO (Figure 1 (Figures 4 and 5), that enable the use of reliable Earth observation datasets. The accuracy of mapping is therefore ensured through the high-quality raw datasets. Using high-resolution EGM-2008 data facilitates detection of variation in geoid undulations with unprecedentedly high details. High-resolution satellite-derived gravity grids provide a source for fine-resolution mapping of geophysical fields and analysis of variations with respect to the topography of Ghana. Since the geoid reflects variations in the land masses that correspond well to the topography of the terrain with unique patterns, it can be compared with the topographic and gravity maps.
Second, this study employs GMT scripting methods for mapping several maps. Automatic data processing and plotting are the major challenges in digital cartography. Automated visualization is an important task because of the increased speed of data processing, accuracy and precision of visualization and fewer human-induced errors. Besides, the advantages of scripts consist in code repeatability so it can be reused in similar studies. This enables rapid processing of large volumes of data. Automated GMT techniques facilitate the cartographic workflow and enable focusing on geographical analysis. Thus, a comparative analysis of maps reveals similar geological and geophysical features on the visualized maps through an analysis of isolines contouring fields and distribution of objects. By contrast, GIS mapping can result in human-induced mistakes and accidental errors during plotting, because traditional mapping is a time-consuming and subjective process. Therefore, accurate automated mapping is a challenging problem in modern cartography, demonstrated in the Methodology section. Thus, using GMT presents a breakthrough in contemporary cartography applied in geologic, topographic and geophysical analysis.

Analysis of Geophysical Features over Ghana
The geologic mapping was performed using QGIS software (QGIS.org 2021). First, a GIS project was generated in a QGIS environment, where the input vector layers were uploaded (Persits et al. 1997) in the European Petroleum Survey Group (EPSG) standard EPSG:4326 -World Geodetic System (WGS) of 1984 WGS-84 projections, and the layout ( Figure 2) was generated using the Layout Manager options. The Digital Chart of the World (DCW) was used as a clipping mask layer for Ghana as a comprehensive reliable vector cartographic layer (Goff 1994).

Mapping geologic units
The geologic provinces of Ghana visualized in Figure 2 include the outcrops of rocks from the following geologic period units: Ordovician Cambrian (OCm), Precambrian and Cambrian (pCm), Cretaceous (K), Quaternary Tertiary (QT), Quaternary eolian (Qe) and Tertiary (T). The eolian fraction of the late Quaternary (Qe) sediments points at climate variations and change in the tropical region ofGhana since the Quaternary enabling the reconstruction of the regional climate changes based on the geological records ofthe past sediments.
The disposition of the geologic units shows the tectonic and stratigraphic correlations between the Precambrian and Lower Paleozoic Volta River Basin and the Pan African orogenic belt of West Africa. Moreover, the location of the Early Proterozoic Birimian Supergroup of Ghana is largely associated with gold mineralization in Ghana.

Mapping geologic provinces
The geologic provinces of Ghana include the Gulf of Guinea, Nigerian Massif, Taoudeni Basin, Volta River Basin and the West African Shield. The Volta Basin presents the structural geologic core of the region, while the Nigerian Massif is distributed on the east of the country and the West African Shield on the west, respectively. A three-phase geodynamic evolution of the Volta Basin ofGhana (light green colour in Figure2) was structured by Affaton (1990) divided into three groups: i) deposition of craton-margin sandstones (Bombouaka Supergroup), ii) sedimentation marginal to a Pan-African oceanic domain (Pendjari-Oti Supergroup), iii) development of a foreland basin where molassic sediments of the Tamale Supergroup were poured. Deynoux et al. (2006) discussed the geologic properties of the 3 major units (megasequences) of the Volta Basin (green colour in Figure 2): 1) Bombouaka, 2) Pendjari-Oti, 3) Tamale, serving as marker horizons to constrain inter-basin correlations between the geologic units of Ghana. However, the period of the sequential evolution from the rift-to collision-related sedimentation in these provinces ofGhana differed regionally. The GMT consists in a console-based scripting approach, which differs from traditional GIS, such as QGIS. Parameters are set in special GMT modules to allow the cartographic elements on all plots to be visualized based on the settings (flags) in the shell script for each element (colour palette, grid specification, annotations, layers translucency and order of appearance). The specific commands used for mapping Figures 1, 3, 4 and 5 are provided with explanations below.

No.
GMT module GMT code Br.

Geophysical mapping
Mapping geoid and free-air gravity in Faye's and Bouguer corrections followed a similar scheme using the general GMT modules such as 'pscoast', 'makecpt', 'grdimage', 'pstext', 'psbasemap' and 'psscale' using the same Mercator projection. The codes using for plotting geophysical maps (Figures 3, 4 and 5) are presented in Table 2. Here the 'grdconvert' module was applied for the EGM-2008 data re-formatting using code No. 1, following the example (Lemenkova, 2020a). The isolines on the geoid map ( Figure 3) were plotted with an interval of 0.25 m by code No. 2. Plotting isolines provides information on the distribution of fields over the area enabling the assessment of the proximity of values, which are determined by the geologic and geophysical setting of the region, including rock density, topography and tectonic lineaments. Adding coastlines, borders and a river network was performed using the GMT 'pscoast' by code No. 3. The free-air gravity in Faye's reduction (Figure 4) was extracted as a subset of an img file in the Mercator projection by code No. 4. This file was then clipped using the coordinate extent of the study area and saved as netCDF format by code No. 5. The same procedure was repeated for the Bouguer correction ( Figure 5).
The extremes of the gravity grid were then assessed using the 'gdalinfo' GDAL utility using code No. 6. The GDAL Library is used both for handling and reformatting geospatial data (Lemenkova, 2020b(Lemenkova, , 2021b. The results of the evaluation revealed the following data: Minimum=−66.055, Maximum=92.523, Mean= 7.462, StdDev=20.665. The same process was repeated for the Bouguer grid ( Figure 5) with the following outcome: It takes values from Minimum=−126.055 to Mean=0.267,StdDev=12.482. According to the data scope, both grids (Figures 4 and 5) were visualized using the 'jet' and 'seis' colour palettes. The GMT logo was plotted on all the maps using code No. 7.
To summarize the methodology, the GMT console-based scripting method for cartographic data processing is to a certain extent similar to programming such as in Python or R (Lemenkova, 2019d(Lemenkova, , 2019a. It adds the lines of code using modules processed by the GMT syntax, and then executes the scripts from the console and presents the graphical output of the map. The final maps are saved by the 'psconvert' module using the Post-Script format into the graphical standard output (JPG, TIFF).

Results and Discussion
There are several aspects of the results of this work that should be pointed out separately: i. Modelling the/a geoid shows the variability of the geoid undulations over the region of Ghana with increased values in the SW. ii. Modelling free-air gravity in Faye's and Bouguer reductions mirrors the large topographic structures (e.g. the long Kwahu Plateau extending diagonally in the southern Ghana). iii. The topography is reflected in the spatial distribution of the geologic units and basins (Volta Basin, White Volta, Black Volta, Oti River). Hence, the developed maps were used to study the relationships and comparisons between the geophysical fields, geologic units and provinces, and topographic structures of the country based on the input of high-resolution data. For each of these maps, data were processed using the Mercator projection for compatibility of the grids.
It must be pointed out that in an effort to study the geologic and geophysical setting of Ghana visualization of geoid and gravity anomalies must be embedded into a more comprehensive framework using more detailed data. In this study, using cartographic representation methods, geology is reported to affect both the geophysical and topographic setting of the country by observing that the differences in those values are not universal and tend to be dependent on regional and local geology instead. A series of investigations on the impact of geology and geologic development on the geophysical setting and topography can further be continued using this study as a basis to explore the existence of bias phenomenon on comparative analysis of the topography and geology of the west African region. Both GMT and GIS in this study have shown that both scripting and traditional techniques consistently tend to perform well in the cartographic workflow, however, the GMT demonstrates more automation in techniques.
The free-air gravity maps visualized in Figures 4 and 5 for Faye's and Bouguer reductions, respectively. In each case, the correlation between the fields and topographic values of the Ghanian relief was noted. The grid for the free-air gravity was used in Faye's reduction (Figure 4).
In the area where the Kwahu Plateau extends between 0.15°W−1.0°W, 6.00°N−7.20°N, the free-air gravity values are above 40 mGal (dark red colour in Figure 5), that is, higher compared to the plain surrounding areas (between −5 to 1 mGal, coloured yellow in Figure 5). Selected depressions in the topography of the Volta Basin coincide and are in correspondence with the lower values ofgravity (−30 to −20 mGal, green colours in Figure 5). The correspondence between the submarine relief in the coastal area of the Gulf of Guinea also demonstrates comparability between the elevated relief and higher free-air gravity values.
The comparison between Faye's and Bouguer reductions gives the following results. The data range for Faye's gravity (grav_27.1.img) extends from the minimum=−66.055 to a maximum=92.523, with a mean of 7.462 and standard deviation (StdDev)=20.665. For the same grid extent, the vertical gravity in the Bouguer reduction (curv_27.1.img) demonstrates the following data range: minimum=−126.055, maximum= 157.265, mean=0.267 and the standard deviation (StdDev)=12.482. It indicates that Faye's gravity reduction shows a more compact data range and general extent which is higher for the Bouguer gravity due to the differences in the data processing algorithm.
The free-air Faye gravity anomalies are strongly correlated with topographic height and can be achieved by the interpolation of Bouguer anomalies and sequential transformation to the free-air (Faye's) anomalies.
Local linear correlation with the topographic heights ofthe reliefin Ghana are demonstrated in both grids. Numerical values are modelled, compared and visualized for the interpolation of both Bouguer and Faye's gravity anomalies on Ghana. Higher values (over 60 mGal) are recorded for the elevated areas both in the submarine and terrestrial relief, while strongly negative values ( 80 mGal) are noted in the water areas of the GulfofGuinea. The majority ofthe Ghana terrain is covered by the values −10 to −20 mGal (aquamarine colour in Figure 4) with clearly visible region ofthe extension of the Kwahu Plateau (25−30 mGal, orange colour in Figure 4). The negative values in gravity are caused by the isostasy effect, as rock density of the mountains is lower, compared with the surrounding Earth's mantle. Hence, positive gravity values may indicate metallic ores and help in geologic prospecting.

Conclusions and Recommendations
The presented paper proposes that the described combination of the traditional GIS (with an example of QGIS) and scripting GMT methodology for an integrated thematic mapping of Ghana, starting with topographic relief visualization and then adding a series of geophysical and geologic maps, is not only a functional alternative to the GIS methods for cartographic multiformat data processing and map producing, but may also be significantly faster due to the console-based shell scripting applied for mapping. doprinos razvoju kartografskih metoda. Osim toga, nizom tematskih karata naglašava se korelacija između geofizičkih, geoloških i topografskih svojstava Gane uz upotrebu nekoliko skupova podataka (gravitacijske mreže, geoidni model, geološki slojevi i topografski raster) kako bi se prikazala raspodjela geofizičkih polja na području Gane na temelju usporedne analize kartografskih podataka obrađenih integriranim pristupom dvaju geomatskih alata otvorenog koda: skupa kartografskih alata za skriptiranje na GMT konzoli i softverskog proizvoda QGIS.
In particular, using GMT syntax to write codes for data modelling, mapping and cartographic visualization appears to require an approach similar to programming (Lemenkova 2020e, Lemenkov andLemenkova 2021a). These GMT code lines are needed to control the specific settings in a script for each element visualized on a map for each raster grid as opposed to the standard GIS methodology (Klaučo et al. 2014(Klaučo et al. , 2017 presented by the GIS where the settings were defined for the maps as undertaken in the QGIS part of this research. The available techniques for the cross-validation and calculation ofthe results received in geological mapping are described in previous studies and can be applied in research for comparison of results in similar studies (Malvić et al. 2019, Ivšinović andMalvić 2020).
This paper demonstrates a contribution to the development of mapping methods in cartography. Moreover, through a series of thematic maps it highlights the correlations between the geophysical, geological and topographic setting of Ghana using several datasets (gravity grids, geoid model, geologic layers and topographic raster) to show the distribution of the geophysical fields over the terrain relief of Ghana based on comparative cartographic data analysis processed by an integrated approach of the two open source geomatic tools: the GMT console-based scripting cartographic toolset and the QGIS software product. The recommendations for future studies can be summarized as follows: 1. Using GMT scripting is beneficial for enlarged regions of Ghana using the available geological datasets or fieldwork data. Other GMT modules can be applied as well (Lemenkova 2020c(Lemenkova , 2020d. QGIS or other open-source GIS may be considered for mapping, e.g. System for Automated Geoscientific Analyses GIS (SAGA GIS) or the Environment for Visualizing Images GIS (ENVI GIS). 2. Geologic data quality should be assessed carefully.
This includes controling the data reliability and origin, format and resolution, spatial extent and relevance. Before handling datasets and making decisions about their acceptability for mapping, data quality has to be examined. It should normally include the inspection of settings using metadata and the relevant comments often accompanying a dataset. Moreover, data analysis includes controling the dataset generation time or the timeline ofthe geologic fieldwork and relevant technical specifications. 3. Retrospective analysis ofthe results is advised for the comparison of maps with previously made ones. This should be based on different maps of Ghana of an earlier origin and made by geological surveys in order to control the quality of the new maps using previously made ones. 4. Expanding the research towards a multidisciplinary project on the geology ofGhana includes both natural and social aspects of mineral resource exploration. Future research might also consider collaborative works with local geologic communities and expand projects towards emerging or geologically promising areas of Ghana. Additional studies can include environmental data assessment on climate issues (floods, droughts), social development and benefits for the local population from geologic exploration (e.g. increased employment, social facilities, mobility and possibilities for people). 5. Statistical data would update and bring new issues in geologic research on Ghana, integrating natural and social sciences. The combination of statistical data and mapping could be achieved by the use ofR ofPython languages that have a variety oflibraries for tabular data processing and visualization. To conclude, it is important to note the significance of scripting methods in contemporary cartography. Since the first scripting applications in cartographic routine, the emphasis has been placed on increasing both the quality and the speed of mapping achieved by machine learning algorithms in data processing. The first issue concerns the precision and aesthetics of maps. The second one concerns the automatization that increases the speed of data processing compared to the traditional GIS.
Hence, research on the geology and geophysics of Ghana should consider an integrated mapping methodology as a challenge for expanding technical cartographic tools into other geologic studies of Ghana in similar projects. While this emphasis is justified by the importance of automatization and machine learning in Earth sciences as technical tools, it also arises from the reason that the interest towards the geology of Ghana is consistently growing due to the exceptional richness ofthe country's mineral resources.