Original Article

Assessment of the Geomorphological Diversity of Komati Gorge,

Mpumalanga Province, South Africa

Mukhodeni Mudau1,*, Edmore Kori1

1Department of Geography and Environmental Sciences, University of Venda, Limpopo Province, South Africa

Corresponding Author:

Mukhodeni Mudau Department of Geography and Environmental Sciences, University of Venda, Limpopo

Province, South Africa Email: [email protected]

Abstract

The emerging broad science of geodiversity defined in terms of geomorphological diversity assesses geomorphological features of territory by comparing them in an extrinsic and intrinsic way. This paper uses SRTM (Shuttle Radar Topographic Mission) data and GIS (Geographical Information System) techniques to assess the geomorphological diversity of Komati Gorge, in Mpumalanga Province of South Africa. Factors used to assess geomorphological diversity are relative height, in- solation, hydrography, geology, soil erodibility, ruggedness, slope position, and land use/land cover. Each factor was normalized to five classes by applying nat- ural breaks, and all were weighted before overlaying. The weighting reveals that hydrography, ruggedness, relative heights, and geology carry more weight, respec- tively. Slope position, insolation, land use/land cover, and soil carry the least weight in that order. The final geomorphodiversity map reveals that the south-western parts of Komati Gorge have medium to very high geomorphological diversity. The north-eastern parts have low to medium geomorphological diversity. This indicates that factor-specific research can add more information to geomorphodiversity re- search and education.

Keywords: Geodiversity; Geomorphological diversity; GIS; Landform inventory; Geodatabase

Article information

Received: 2022-04-08 Accepted: 2022-07-11 DOI: 10.30486/GCR.2022.1955504.1103 How to cite: Mudau M & Kori E (2022). Assessment of the Geomorphological Diversity of Komati Gorge, Mpumalanga Province, South Africa. Geoconservation Research. 5(1): 195-208. doi: 10.30486/ gcr.2022.1955504.1103 Geoconservation Research e-ISSN: 2588-7343 p-ISSN: 2645-4661 © Author(s) 2022, this article is published with open access at http://gcr.khuisf.ac.ir This work is licensed under a Creative Commons Attribution 2.0 Generic License.

Introduction

The earth is complex and dynamic, with many processes occurring beneath, in, and above the crust. Geomorphological processes such as weath- ering and erosion determine the surface features. Geomorphological diversity refers to the natural range of geological, geomorphological, and soil features, processes currently acting on rocks, landforms, soils, and topography of a defined area (Kori et al. 2019; Kot 2018; Panizza 2009). Geomorphodiversity, which falls within the broad geosciences field of geodiversity, is a critical and specific assessment of forms or surface features of a defined place on Earth (Kori et al. 2019; Paniz- za 2009). Geomorphodiversity cannot be univo- cal, having extrinsic and intrinsic aspects (Paniz- za 2009). Extrinsic factors are morphostructural, morphoclimatic, morphographic, morphogenetic, and morphometric characteristics, and intrinsic aspects include the diversity and complexity of landforms, assessing the total range of phenom- ena, and comparing them to each other. Both ap- proaches, whether individually or combined, start with a geomorphological map (Panizza 2009).

An intrinsic and extrinsic diversity assessment explains why the environment is characterized by an interaction of rocks, water, air, and hu- mans. This geomorphodiversity study critically and specifically assesses the geomorphological features of Komati Gorge by comparing them intrinsically and extrinsically. The geomorpho- logical features are then mapped to visualize the characteristics and behavior of the morphology of the area, the morpho-characteristics of the Komati Gorge. The study is based on the identi- fication of geomorphological elements that inde- pendently characterize the landscape of a terrain (Panizza 2011).

Materials and Methods Study Area

The study was conducted in the Komati Gorge (Fig. 1), a river valley situated in the northeastern part of South Africa, near the communities of Car- olina and Machadodorp in Mpumalanga Province. Komati Gorge is in the upper reaches of the pe- rennial Komati River, situated at latitude 25o 52’ 4,19” S and longitude 30o 18’ 15,60” E.

Figure 1. Location of the study area, Komati gorge

The Komati River originates from the Transvaal plateau near Breyton in the Drakensberg range west of the small-town Carolina in the Mpumalanga Province in South Africa (Mahlathi 2018; Nkomo and van der Zaag 2004). It flows into Swaziland and then re-enters South Africa where it is joined by the Crocodile River at the border with Mozambique to form the Incomati basin. The Incomati River then flows north-eastwards to the Indian Ocean at Maputo Bay in Mozambique (Nkomo and van der Zaag 2004). The Komati River is approximately 480 km long from its source to the confluence with the Crocodile River. The total catchment area of the Komati River is approximately 50,000 km2 (Nko- mo and van der Zaag 2004).

From upstream to downstream, the Incomati basin comprises Precambrian granites and gneiss of the Archean, and Cretaceous and Karroo lavas of the Mesozoic (Romano 1965). The area is dominated by wide sandstone krantzes, with exposures of the Kromberg Formation and upper Hooggenoeg For- mation of the Onverwacht Series (Dann 2000). It also consists of a riparian zone, bluff habitat, thorn bushveld, and Highveld grassland.

Data and Sources

Relevant factors for the assessment of geomor- phological diversity of the Komati Gorge were selected following the definition of geodiversi- ty. The topographic Position Index (TPI), Topo- graphic Wetness Index (TWI), and insolation are relevant choices for geodiversity assessment (Na- jwer and Zwoliński 2014). Geology, soil, relative height, erodibility, ruggedness, and land use/land cover were then selected based on their correlation with the first three factors. The eight factors can characterize geodiversity in the fullest possible way (IAG/AIG 2019; Kori et al. 2019).

Geomorphometric parameters, namely soil erod- ibility, development ratio, drainage density, drain- age texture, stream length ratio, and ruggedness were calculated using mathematical formulae

1 to 6. The mathematical formulae were used

in combination with the 30-m raster size SRTM DEM (Digital Elevation Model) aerial photo- graphs (Price et al. 2011). The photographs were obtained from National GeoSpatial Information (NGI) and data extracted using System for Au- tomated Geoscientific Analysis Geographic In- formation System (SAGA GIS), Environmental System Research Institute Aeronautical Recon- naissance Coverage Geographic Information Sys- tem (Esri ArcGIS) 10.2, and Esri ArcGIS Pro 1.2. The SRTM downloader tool in the ArcGIS 10.2 map server was used to download the 30-m raster size SRTM DEM from the NGI.

A 30 arc-second raster size Harmonized World Soil Database layer was obtained from the International Institute for Applied Systems Analysis (IIAS) on- line database. The erodibility of the soil was ob- tained from calculating the soil erodibility factor (K factor) using the Bouyoucos method (Bouyoucos 1935). This method was chosen for its ability to ef- ficiently match field data in addition to its simplic- ity to use in GIS (Kori et al. 2019). The erodibility of the soil accounts for the influence of the intrinsic soil properties (Wang et al. 2001). The intrinsic soil properties may characterize and serve as an indica- tor for soil taxonomy (Svoray and Shoshany 2004). The intrinsic soil factors affect and are also affected by the variations in soil properties. Soil erodibili- ty is used in quantifying geomorphic diversity be- cause it significantly influences landform and land- scape development (Kori et al. 2019).

Geological data was downloaded from the South African Geosciences online database. Hydrogra- phy systems shapefiles were obtained from the De- partment of Water Affairs and Forestry (DWAF). Land use/land cover shapefiles for 2020 were ob- tained from the Department of Rural Development and Land Reform (DALRRD).

Creation of Factor Maps

The geologic, hydrologic, soil and land use/land cover data were analyzed using SAGA GIS, Arc- GIS 10.2, and ArcGIS Pro. The hydrography fac-

tor map is the result of the addition of three maps showing the diversity of rivers, lakes, and the TWI. The TWI map was created from DEM in terrain analysis tools (hydrology) in SAGA GIS. It was based on a modified catchment area calcu- lation which does not think of the flow as a very thin film (IAG/AIG 2019; Kori et al. 2019). The lakes were assessed based on shoreline Develop- ment Ratio (DL) (Minár et al. 2005) computed using formula 1. Calculations were done in an at- tribute table of the hydrography layer in ArcMap. Rivers were assessed using the average slope at about 100 m long. The tool for creating points on lines in ArcMap was used to divide the rivers into 100 m sections.

𝐷𝐿 = 𝐿 2 √𝜋𝐴 (1)

where DL = Shoreline Development Ratio; L = Length; 𝜋 = 3.14; A = Area.

The geology factor map was created in ArcMap from the geology shapefiles. Lithological diver- sity information was extracted from the attribute table (Table 1) of the geology shapefile and com- bined with the rock hardness using the raster cal- culator of ArcMap to produce the final geology factor map. The geology shapefiles provided attri- butes of each polygon, which indicated the pres- ence of a dominant rock type and the presence of other rocks. The soil factor map was produced in ArcMap based on the topsoil erodibility (K) fac- tor from the obtained soil layers. The topsoil was considered because it is more sensitive to erosion. The (K) factor was calculated using a formula by

Bouyoucos (Bouyoucos 1935) method. Erodibility (K) = [(sand + silt) / (clay)] / 100 (2)

The erodibility layer was then added to the soil layer to produce the soil factor map. Producing a map based on soil erodibility (K) is important be- cause of the intrinsic characteristics of a soil to be eroded (Kori et al. 2019; Oparaku et al. 2016).

The land use/land cover factor map was created based on the expert classification in ArcGIS Pro

1.2 from the downloaded shapefiles. The geomor- phodiversity analysis of hydrological, geological, soil, and land use/land cover data was done by overlaying appropriate hydrological, geological, soil, and land use/land cover maps as thematic maps to a shaded relief to achieve composite vi- sualization. This was to get morphographic, mor- phometric, structural, and morphogenetic data of features. The visualization of topographic attri- butes and their analysis is a useful tool for under- standing the geomorphological features and pro- cesses acting on them. The first three maps were then reclassified into five geodiversity classes us- ing Jenks’ natural breaks classification method in spatial analyst tools (reclass) in ArcMap and land use/land cover in ArcGIS Pro and saved (Jenks 1967).

Insolation was calculated in both SAGA GIS and ArcGIS Pro 1.2. In SAGA using terrain analysis tools (lightning), the DEM was automatically an- alyzed to get the map for the analysis area for the

Table 1. Soil erodibility

whole year. Ruggedness was computed using the topographic roughness index in SAGA GIS. The roughness factor expresses the amount of elevation difference between adjacent cells of a DEM (Kori et al. 2019). The landscape roughness is a measure of the irregularities of a topographic surface.

The slope position factor map classifies the land- scape into cliffs, scree slopes, transportation mid- slopes, foot slopes, and open valleys (Kori et al. 2019). A slope position map was obtained from the DEM using TPI in raster tools in QGIS. The TPI was chosen for slope analysis because it is the difference between a cell elevation value and the average elevation of a neighboring area around the cell (De Reu et al. 2013). The TPI was computed considering a 300-m radius circle neighborhood in tandem with the 30-m resolution of the DEM.

Relative height is a measure of the elevation of a place in its surroundings. It shows the diversity of local heights, which reflects the energy of the re- lief and is a measure of elevation of a place. DEM was used to show the diversity of relative heights, which reflects the energy of the relief (Zwoliński 2008; Zwoliński 2009). The values were calculat- ed for each grid cell using the spatial analyst tools to analyze the neighborhood (Focal Statistics) in a 3 × 3 grid cells moving window in ArcMap.

The drainage texture parameters describe the characteristics of the basin including the length of overland flow, constant of channel maintenance, frequency, drainage density, drainage texture, infiltration number, and intensity of the drain- age network (Biswas 2016; Chethan and Vishnu 2018). Drainage texture is classified according to standard metrics (Table 2). The drainage density is the ratio of the total length of streams of all orders to the area of the basin. The value range of the drainage density is approximately a minimum of 0 and a maximum of 5000 m-1. It is calculated for all the river channels and interpreted according to the type of landscape. Drainage density is given by:

Dd = L / A (3)

where Dd = Drainage density; A = Area of basin; L = Total length of the stream.

Table 2. Five classes of drainage texture (Source: (Magesh et al. 2013))

The relative height factor map in raster format was added to ArcMap and standardized into five geo- diversity classes using Jenks’ natural breaks clas- sification method (Jenks 1967) in spatial analyst tools (reclass) and saved. Insolation and rugged- ness factor maps in raster format were added to ArcGIS Pro and standardized into five geodiver- sity classes using Jenks’ natural breaks classifica- tion method (Jenks 1967) in spatial analyst tools (reclass) and saved. Jenks natural breaks classi- fication method is a data classification method designed to optimize the arrangement of a set of values into “natural” classes. This method seeks to minimize the average deviation from the class meanwhile maximizing the deviation from the means of the other groups. The features are divid- ed into classes whose boundaries were set where there are relatively big differences in the data val- ues. It is also known as the goodness of variance fit (GVF), which is equal to the subtraction of the sum of Squared Deviations for Class Means (SDCM) from the sum of Squared Deviations for Array Mean (SDAM). The slope positions factor map was reclassified into five geodiversity classes by the table in the processing toolbox (raster anal- ysis) in QGIS and saved.

The Final Geomorphological Map

The geomorphological map is the most prominent output from this study. The eight-factor maps were combined as raster layers to produce a geomorpho- logical map. Before overlaying the eight factors, the Multi-Criteria Evaluation (MCE) using the

Analytic Hierarchy Process (AHP) criteria were employed to calculate relative weights for each factor map (IAG/AIG 2019). Table 3 shows the ranking values of each factor in percentage. The AHP applies a pairwise comparison to provide a basis for the criteria used in decision analysis and determines values for each of the criteria (Cay and Uyan 2013). A matrix is then generated because

of pairwise comparisons and criteria weights are reached because of the calculations. The ordered weighted averaging tool from grid analysis in pro- cessing tools in QGIS was used to combine the factor maps. Geomorphological characteristics of the factor maps were added as tabular attribute in- formation linked to the vector information as at- tribute data.

Figure 2. Komati Gorge hydrography factor map

Figure 3. Komati Gorge geology factor map

Table 3. Geomorphological factor ranking values

Results and Discussion Factor maps

The hydrography factor reveals the interplay of high elevation areas, waterways, and open val- leys. Low hydrography occupies 50% of Komati Gorge, revealing that Komati Gorge has few sur- face water resources (Fig. 2). There are large areas of hard and very hard rocks (Fig. 3; Table 4) be- cause the area is rich in rocks such as sandstones, cherts, and basaltic rocks. Geology affects soils and waterways. Rock hardness (Table 4) indicates density and resistance to breaking. The hardness

and strength of the rocks are necessary parameters in this study as they are a function of the individ- ual rock type. The hard, impermeable rock types resist rivers in the Komati Gorge to create many water paths and cut deep, which means that the water bodies of this area are mainly shallow.

The soil factor map (Fig. 4) shows that Komati Gorge has highly erodible soils. About 17 % of the soils have low credibility and occupy the south section of the gorge only. Soil factors affect other factors such as river networks and slope angles, which are directly influenced by soil erodibility. Rivers follow easily erodible terrain while highly erodible soils create plains. Slopes and river chan- nels in the Komati Gorge conform to this general anticipation. The land use/land cover factor map (Fig. 5) reveals how and what the land in Komati Gorge is used for. The map shows that much of the land (over 90%) is used for either one of the uses in Table 5. The land use/land cover map strongly corresponds with the relative heights, hydrogra- phy, soil, and insolation.

Figure 4. Komati Gorge soil factor map

Table 4. Rocks found in Komati Gorge

Figure 5. Komati Gorge land use/ land cover factor map

The relative height factor map (Fig. 6) reveals the interplay of many factors and the dominance of rough surfaces such as mountains, hills, cliffs, and open valleys, an indicator of high local differenc- es in elevation between cells. The high diversity of local heights reveals the high energy of the relief of Komati Gorge. This factor affects other factors such as the ruggedness of the area, which decreases with an increase in elevation. Over 90 % of Koma- ti Gorge experiences high to very high insolation (Fig. 7). Since the area is dominated by high-lying areas, it can be concluded that these are the ones receiving such high insolation, and the remaining 6 % low-lying areas would receive less insolation. This may be due to the shading effect of the high el- evation areas. The insolation factor directly affects other factors such as soil, geology, and hydrology. The soil of the high-lying areas is more substantial- ly developed than soils in low-lying areas. Rocks in areas of high insolation are rich in iron.

Over 70% of Komati Gorge has low ruggedness

(Fig. 8), namely areas at high elevations. Only 25

% is found at low elevations with high rugged- ness, along the mountain edges and river valleys. The ruggedness factor strongly correlates with slope position and relative height factors. The un- even distribution of river valleys and cliffs is clear (Fig. 9). Slope positions further classify the land- scape into five morphological classes.

The final geomorphodiversity map (Fig. 10) re- veals the variable influence of different factors on the geomorphology of Komati Gorge. The edges of the area have very high geomorphodiversity while the inner part has very low geomorphodiversity. Over 70% of the area has medium to very high and 21% has low geomorphodiversity. The area of high geomorphodiversity is based on over 90% of dif- ferent land uses. These areas are located at high el- evations, receive 94 % of insolation and constitute valleys and rivers. They also had highly erodible soil, and very negligible (2 %) ruggedness.

Figure 6. Komati Gorge rela- tive height factor map

Figure 7. Komati Gorge insolation factor map

Figure 8. Komati Gorge ruggedness factor map

Table 5. Komati Gorge land use/land cover

Figure 9. Komati Gorge slope positions factor map

Figure 10. The variable influence of different factors on the geo- morphology of Komati Gorge

Concluding Remarks

Geomorphological mapping, a now flourishing science, provides a firm basis for the investiga- tion of geomorphological features. The geomor- phological map produced provides insights into geological processes that shaped the Gorge. The map revealed that the southwest parts of Komati Gorge have high geomorphological diversity. The northeast parts of Komati Gorge have low to me- dium geomorphological diversity. Hydrography, ruggedness, relative heights, and geology carried more weight and had the most noticeable influ- ence on the geomorphological diversity of Komati Gorge. Slope position, soil erodibility, insolation, and land use/land cover carried the least weight and were observed as minor influencers of the geomorphological diversity of Komati Gorge.

Acknowledgments

Mudau Mukhodeni wishes to acknowledge the Department of Ecology and Resources Manage- ment. I would like to thank my supervisor, Mr

E. Kori, for his good communication, providing a good learning environment, and for guiding me throughout my learning experience.

Author contributions

Mudau Mukhodeni drafted the paper and did the GIS extractions and calculations. Edmore Kori corrected and proofread the paper before submis- sion.

References

Biswas SS (2016). Analysis of GIS-based morphomet- ric parameters and hydrological changes in Parbati River Basin, Himachal Pradesh, India. Journal of Geographic Natural Disasters. 6(2): 1–8.

Bouyoucos GJ (1935). The clay ratio as a criterion of susceptibility of soils to erosion. Journal of the American Society of Agronomy. 27: 738–741.

Cay T and Uyan M (2013). Evaluation of reallocation criteria in land consolidation studies using the An- alytic Hierarchy Process (AHP). Land use policy. 30(1): 541-548.

Chethan BJ and Vishnu B (2018). Determination of the Geomorphologic parameters of the Thuthapuzha River Basin in Central Kerala, India, Using GIS and Remote sensing. Journal of Current Microbiology and Applied Sciences. 7(9): 1245-1260.

Dann JJ (2000). The 3.5 Ga Komati Formation, Barber- ton greenstone belt, South Africa, part I: new maps and magmatic architecture. South African Journal of Geology.103(1): 47-68.

De Reu J, Bourgeois J, Bats M, et al. (2013). Applica- tion of the topographic position index to heteroge- neous landscapes. Geomorphology. 186: 39-49.

IAG/AIG (2019). Geodiversity assessment in mountain areas: Computer training tutorial. Krakow-Zako- pane, Poland: Workshop of the IAG/AIG Working Group on Landform Assessment for Geodiversity.

Jenks GF (1967). The data model concept in statistical mapping. International Yearbook of Cartography. 7: 186-190.

Kori E, Onyango Odhiambo BD and Chikoore H (2019). A geomorphodiversity map of the Sout- pansberg Range, South Africa. Landform analysis. 38: 13-24.

Kot R (2018). A comparison of results from geomor- phological diversity evaluation methods in the Pol- ish Lowland (Toruń Basin and Chełmno Lakeland). Danish Journal of Geography. 118(1): 17-35.

Magesh NS, Jitheshlal KV, Chandrasekar N, and Jini KV (2013). Geographical information system-based morphometric analysis of Bharathapuzha river basin, Kerala, India. Applied Water Science. 3(2): 467-477.

Mahlathi S (2018). The net economic benefits from eradication of invasive alien vegetation: The case of the Inkomati Catchment Mpumalanga Province in South Africa. Dissertation, University of Pretoria.

Minár J, Mentlík P, Jedlicka K, and Barka I (2005). Geomorphological information system: idea and options for practical implementation. Geograficky Casapis. 57(3): 247.

Najwer A and Zwoliński Zb (2014). The landform geo- diversity assessment method– a comparative anal- ysis for Polish and Swiss mountainous landscape. EGU General Assembly. 1201.

Nkomo S and van der Zaag P (2004). Equitable wa- ter allocation in a heavily committed international catchment area: the case of the Komati Catchment. Physics and Chemistry of the Earth, Parts A/B/C. 29(15-18): 1309-1317.

Oparaku LA, Aho IM and Iwar RT (2016). Relative vulnerability to gully erosion of three geological sediments: A texture-based assessment. Journal of Environment and Earth Science. 6(2).

Panizza M (2009). The geomorphodiversity of the Do- lomites (Italy): a key of geoheritage assessment. Geoheritage. 1(1): 33-42.

Panizza M (2011). The Geomorphology of the Dolo- mites from a geomorphodiversity viewpoint. Geogr. Fis. Dinam. Quat. 34: 25-32.

Price SJ, Ford JR, Cooper AH and Neal C (2011). Hu- mans as major geological and geomorphological agents in the Anthropocene: the significance of arti- ficial ground in Great Britain. Philosophical Trans-

actions: Mathematical, Physical and Engineering Sciences. 369(1938): 1056-1084.

Romano M (1965). A contribution to the hydrological survey of the Incomati River drainage basin (Riv- ers Komati and Crocodile; Rep. Of South Africa and Swaziland). South African Journal of Science. 61(7): 268-275.

Svoray T and Shoshany M (2004). Multi-scale analysis of intrinsic soil factors from SAR-based mapping of drying rates. Remote Sensing of Environment. 92(2): 233-246.

Wang G, Gertner G, Liu X and Anderson A (2001). Uncertainty assessment of soil erodibility factor for revised universal soil loss equation. Catena. 46(1): 1-14.

Zwoliński Z (2008). Designing a map of the geodiver- sity of landforms in Poland. IAG and AIGEO In- ternational Meeting Environmental Analysis and Geomorphological Mapping for a Sustainable De- velopment, Addis Ababa, Ethiopia (pp. 18-22).

Zwoliński Z (2009). The routine of landform geodi- versity map design for the Polish Carpathian Mts. Landform Analysis. 11: 77-85.