Abstract : Despite being well established in continental watersheds, morphometric analysis is still incipient in oceanic islands. Our goal is to analyze the morphometry and morphologically classify the island of Trindade, located in the South Atlantic. The parameters we used were altitude, slope gradient and aspect, LS factor, Topographic Wetness Index (TWI) and the Geomorphons classification. Trindade island has an altimetric amplitude of 600m, with 34% of its area between 0-100m and 21.2% in the 100-200m range. The island’s lower parts are gravel and sand beaches and alluvial fans. Ridgelines such as the Desejado and Verde peaks represent heights above 500m and occupy 2% of the island’s area. Slope gradients vary from strongly undulated (20-45%) to hilly (45-75%), with both classes occupying half the island’s area. Flatter areas (with slope gradients between 0-3% and 3-8%) occupy less than 4% of the island. The slopes more exposed to weather conditions (rainfall and winds) are the ones facing east and southeast, therefore being the most prone to evolutionary relief processes. The western and northwestern faces are steeper and more sheltered from the weather. The Geomorphons classification has shown that 57% of the island is composed by straight slopes, 19% are spurs (slightly convex slopes in plan and profile), and 18.6% are hollows. The TWI shows that the areas prone to saturation are relief concavities in medium to lower thirds of the slopes.
Keywords: Oceanic IslandsOceanic Islands,HydrologyHydrology,Topographic Wetness IndexTopographic Wetness Index,GeomorphonsGeomorphons.
Resumo : Apesar de bem estabelecida em bacias hidrográficas continentais, a análise morfométrica ainda é pouco desenvolvida em ilhas oceânicas. Este trabalho tem como objetivo analisar a morfometria e classificar morfologicamente a ilha da Trindade, localizada no Atlântico Sul e, verificar sua relação com o ITU. Foram utilizados os parâmetros de altimetria, declividade, fator LS, Índice Topográfico de Umidade (ITU) e a classificação Geomorphons. A ilha da Trindade tem amplitude altimétrica de 600m, sendo que 34% da área da ilha tem cota entre 0-100m e, 21,2% tem entre 100-200m. As áreas mais baixas da ilha são praias de cascalho, areia e leques aluviais. Divisores de água como os picos do Desejado e Verde correspondem às porções acima de 500m e ocupam cerca de 2% da área da ilha. A declividade varia de fortemente ondulada (20-45%) a montanhosa (45-75%), com ambas as classes ocupando mais de metade da ilha. As porções mais aplainadas (declividade entre 0-3% e 3-8%) ocupam menos de 4% da área. Verifica-se que as faces das vertentes mais expostas da ilha às intempéries climáticas (vento e chuva) são as orientadas a leste e sudeste e, portanto, são as mais trabalhadas por processos escultores do relevo. As faces oeste e noroeste são as mais escarpadas e protegidas das ações de intempéries climáticas. A classificação por Geomorphons mostrou que 57% da área da ilha é composta por encostas retilíneas, 19% são cristas secundárias (encostas levemente convexas em plano e perfil) e 18,6% são concavidades. O ITU aponta para setores com tendência a saturação em concavidades do relevo e terços médios e inferiores das encostas, com predominância na porção leste.
Palavras-chave: Ilhas oceânicas, Hidrologia, Índice Topográfico de Umidade, Geomorphons.
Articles
Morphometric and Morphological Analysis of Trindade Island, South Atlantic, Brazil
Análise Morfométrica e Morfológica da Ilha da Trindade, Atlântico Sul, Brasil
Gabriel Silva Nogueira gbl.snog@gmail.com
Julia Frederica Effgen juliaeffgen@gmail.com
Eberval Marchioro ebervalm@gmail.com
Received: 17 July 2019
Accepted: 22 January 2020
Ever since Robert E. Horton’s article (1945), morphometric analysis has been an important relief analysis tool. After the 1950s, new parameters were presented by Strahler (1952; 1957) aiming to insert new variables for morphological and morphometric analysis. In Brazil, the studies of morphology and morphometrics began with Christofoletti (1969).
Since then, the quantitative and morphological analysis and interrelations between the physiography and the relief evolution dynamics were focused on geographical studies. After the pioneering work of Christofoletti (1969), other authors researched on the theme with a variety of goals, such as evaluation and analysis in watershed (TORRES et al., 2011; MOURA, 2013; SPANGHERO; MELIANI; MENDES, 2015; FERNANDEZ, 2016), support to territorial zoning and environmental sustainability (SANTOS; SOBREIRA, 2008; CLEMENTE et al., 2011; UMETSU et al., 2012), characterization and morphometric analysis (BARROS; STEINKE, 2009; STIPP et al., 2010; CARVALHO et al., 2014).
Those methodologies were developed primarily in continental watersheds, being still incipient in volcanic oceanic islands, such as the Trindade Island, in the South Atlantic region. Therefore, this work aims to analyze the morphometrical and morphological aspects from Trindade Island, as well as verify its relation with the Topographic Wetness Index (TWI).
The Trindade Island (20°29’S; 20°30’W), 1200km away from Vitória, Espirito Santo State’s capital, and 1870km from Fernando de Noronha Island, Figure 1.
It is a military island controlled by the Brazilian Navy and since 1957 it is solely occupied for the use of the military and the scientific community, with a scientific base called Posto Oceanográfico da Ilha da Trindade (POIT - Trindade Island's Oceanographic Post).
With perennial and temporary rivers (Figure 2) the island’s characteristics favor the geological, geomorphological, biological, oceanographic and hydrosedimentological studies.
Due to its geographic isolation, Trindade Island favors the concentration of many endemic and endangered species. The island’s singular characteristics have been attractive to scientists since the 19th century, including works of authors such as Charles Darwin and his expedition on the HMS Beagle ship (MURPHY, 1915).
Due to its geographic isolation, Trindade Island favors the concentration of many endemic and endangered species. The island’s singular characteristics have been attractive to scientists since the 19th century, including works of authors such as Charles Darwin and his expedition on the HMS Beagle ship (MURPHY, 1915).
Such studies contributed to the creation of the island’s management plan (a Brazilian document to rule the use and occupation of land in preservation areas), in accordance with pertinent laws. According to the current legislation, the management plan must be concluded until five years after the creation of protected areas. Therefore, on the decree No. 9.312 from March 9th, 2018, the Conservation Unit and Natural Monument of Conservation of Trindade Island was created with the purpose of preserving I- remnants of the Tropical Forest ecosystem; II - scenic beauty and III - Natural resources and marine biodiversity on submerse mountain chains (BRASIL, 2018).
The classic geological description of Trindade was made by (ALMEIDA, 1962). He recognized five volcanogenic successions in distinctive ages: the Trindade Complex (TC) and the Desejado (DF), Morro Vermelho (MV), Valado (VF) and Paredão (PF) formations, composed of nephelinitic to phonolitic effusive-pyroclastic deposits, dykes and necks (Figure 3).
Geochronological data placed the peak of the volcanic activity between 3.9 and 2.5 Ma (i.e. Lower Pliocene to Lower Pleistocene), however, the volcanic activity in Trindade was ceased completely at ca. 0.25 Ma, event registered in the uppermost volcanic deposits of the Paredão Volcano and that represents the last volcanic activity in the Brazilian territory which corresponds to the largest volume of volcanic deposits preserved in the island (PIRES et al., 2016). The Paredão Volcano is the last place in Brazil where debris of an extinct volcano can be recognized (CORDANI, 1970) (Figure 4). The geological environment of the island consists of basic and alkaline rocks, as well as Pleistocene Holocene sediment deposits (CASTRO, 2010).
Trindade Island has (mainly) a Tropical Oceanic climate (MOHR et al., 2009). The average annual temperature is 25.3°C and the prevailing winds are from the East (Figure 5).
According to Pedroso et al. (2018), November is the rainiest month (with mean precipitation of 215mm), followed by October (mean precipitation of 186mm). January and February are the driest months, with precipitation averages of 64 mm. The orographic rainfalls are important on the island (called Pirajá by the locals, which means fast rain in the indigenous language).
The predominant soil in Trindade is the Neosoil. With approximately 55%, it can vary between Litolic and Regolitic. Followed by Haplic Cambisol with 30% and Folic Organosol with 10%, being the remaining 5% of the area of the island composed by rocky outcrops (SÁ, 2010).
With the description from Sá (2010), Machado et al. (2017) indicating that in Trindade, due to peculiarities such as the cloud forest, the materials from unusual origin, such as young volcano rocks and bioclastic carbonate sediments, are influenced by orographic humidity; and the specific interactions between substrates and biological activities with avifaunal colonization, ontogenesis and phosphatization, are the potential soils producers.
Until the 17th century, 80% of the island was covered by forests (ALVES, 1998). These forests were alive and robust in 1700, when the island was visited by the English astronomer Edmund Halley (1656-1742) (MURPHY, 1915). However, in two visits in 1881 and 1889, Sir Edwar Knight (1852-1925) described a dead forest (ALVES; SILVA, 2016 apud KNIGHT, 1884; 1892). The reasons of degradation can be safely attributed to successive attempts at occupations by humans, including the insertion of exotic animals.
Publications and manuscripts of Trindade, Alves and Silva (2016) argue that the occupation and devastation of Trindade Island occurred before the first records of flora and fauna. The authors also add that the human occupation that introduced the goats, pigs and other exotic animals have caused negative environmental impact.
An extensive bibliographic review of works developed on the island was carried out for the elaboration of this article, as well as scientific expeditions, necessary given the geographic isolation of the place. The scientific expeditions occurred in 2018. In April, with approximately 3 days on the island and, between May and June, in a period of 45 days for validation of the maps.
The shapefiles were obtained through vectorization of Nautical Chart No. 21, Ilhas ao Largo - Ilha da Trindade in raster format, made available by the Brazilian Navy (1:15.000).
In 2018, to improve the accuracy of topographic data, a topographic survey was carried out on the island with the Global Position System (GPS) - Real Time Kinematic (RTK).
To validate the maps, we did a scientific expedition. For inaccessible areas (due to topographic issues) we used orthophoto mosaics with a scale of 1:6.000, UTM projection - 26S (WGS-84) from the year 2012 made available by the Brazilian Navy (2017).
The maps were created with the software SAGA GIS 6.4.0, available for free download at <https://sourceforge.net/projects/saga-gis-br/> and ArcMap™ 10.3.2 from ESRI™. The SAGA GIS software was chosen because it is an open source software. Figure 6 shows a flowchart of all GIS processes performed for this paper.
To generate the Digital Elevation Model (DEM), geoprocessing was performed with the pixel at 5 meters (contour lines), scale of 1:10.000, using the Topo to Raster tool in 3D Analyst from ArcMap™. The final maps have a scale of 1:10.000.
The DEM was imported into SAGA GIS and with the Terrain Analysis tool the processing of data was generated (Figure 7). After processing, they were exported to ArcMap™, where the maps were made.
The elevation map was generated from the DEM. The classes were sliced in 25m intervals until 250m, after which the classes were sliced in 100m intervals.
The slope map was classified accordingly with the proposed by Embrapa (1999), the Brazilian Agricultural Research Corporation (Table 1).
The Map of Slope Length and Steepness factor (LS-factor) is part of the universal soil loss equation (USLE), made in SAGA GIS software, with the methodology proposed by Desmet and Govers (1996). They developed an algorithm to calculate the L factor, based on the equation of Foster and Wischemeier (1974), in which they considered the flux accumulation as an estimate of slope length.
The L factor represents the ratio of soil losses between slope length and length of the plot of 25m of the same soil, rainfall, hillslope degrees, cover land and management factors. The S factor deals with the ratio of soil losses between any slope and a 9% slope, and with an equation for possible adjustments to local characteristics (BESKOW et al., 2009).
The classification for the LS factor map was done by Natural Breaks, using 5 classes, as follows: 0 - 4.186 (Very low); 4.186 - 8.152 (Low); 8.152 - 11.897 (Average); 11.897 - 16.303 (High) e 16.303 - 56.180 (Very High).
To obtain the Slope Orientation map we used the Raster Surface - Aspect tool from ArcMap™. The results were classified at every 27,5° interval from degree 0 as north.
Geomorphons is a morphology classification method proposed by Jasiewicz and Stepinski (2013). The textural similarity analysis is performed on the DEM, considering the variations of grey level between a central cell (center pixel) and the neighboring cells. Considering the central cell, it assumes positive values as “higher”, negative values as “lower” and 0 as “same”.
To calculate the zenith and nadir angles of a central cell, eight elevation profiles starting at the central cell and extending along the principal directions up to the “lookup distance” L are extracted from the DEM (Figure 8).
To process the DEM and generate the .jpg geomorphons file, we used the online tool available at <http://sil.uc.edu/geom/app>. The code is available for use in GIS software tools. The parameters used to prepare the map were: 10 cells of search radius (totaling 50m radius) and leveling of 0.3° for the flat areas.
The TWI is defined by the natural logarithm of the ratio between the specific contribution area (As, in units of orthogonal to front line of flux (m2 m-1)) and the tangent of slope (β, in radians), (TWI = In (As/tan β)) where As corresponds to the specific area contribution. The equation shows that the TWI is a function of β. Therefore, the constraints arise: 0º < β < 90°, because if β = 0, tan β = o and 1/tan β is infinite and if β = 90°, 1/tan 90 = 0, In 0 is infinite.
The TWI provides an index of relative saturation in watersheds, and hydrological similarity between areas with same index value is assumed (BEVEN; KIRKBY, 1979).
In the elaboration of the TWI map for Trindade, we sliced the results in qualitative segments using the Natural Breaks classification from ArcMap™ in 5 classes of saturation possibilities: Very low; Low; Average; High and Very High.
The altimetric amplitude of Trindade Island (Figure 9) is roughly 600 meters, with 34% of the island’s area ranging between 0-100m and 21.2% with altitudes ranging from 100-200m, as shown in Table 2. The interval between 200-300m corresponds to 16.8% of the island, and only 2.1% of Trindade heights are above 500m.
Figure 10 shows the waterheads with the peaks Desejado, Verde, São Bonifácio and Trindade. This sequence was named by Almeida (1962) as Sequência do Desejado.
The areas with lower altitudes correspond to regions classified by Angulo et al. (2018) as sand and gravel beaches and alluvial fans, concentrated mainly in the East and Southeast parts of the island, such as Tartarugas` Beach , Paredão Beach, among others.
The following relief classes are predominant: highly undulated (31.2% of the island’s area) and mountainous (28.8% of the island’s area), totaling 60% of the island’s relief (Table 3). The cliffs occur in 24.3%, while the flat, gently undulated and undulated are in 13.4% of Trindade Island’s area (Figure 11).
The mountainous and the cliffs classes were in all slopes on the Northwest and West (Figure 12) occurring in some parts of upper and middle slopes from the East.
The highly undulated and mountainous classes are present mainly on slopes to the West part of Trindade, especially on slopes facing the Northwest (Figure 13). These characteristics suggest that these slopes are protected from winds, which come from the East through most of the year, as well as protected from air masses from the South, which occur between April and September.
The slope map shows that slopes facing the Northeast, East and Southeast (Figure 12) have lower inclinations in relation to winds coming from the East, which altogether transport more humidity to Trindade during most of the year, causing more weathering and erosion of the slopes.
Table 4 shows the area of Trindade presented as m² and as a percentage of LS Factor results. The highest percentage of the LS Factor was classified as Low, represented by clear green (4.186 - 8.152), totaling 30% of Trindade, followed by the average class in yellow (8.152-11.897), with 26% of area, and the class very low, in dark green, with 22% of Trindade’s area. The classifications high (11.897 - 16.303), in orange, had 17% and very high (16.303 - 56.180), in red, had 5% of the total area.
In all thirds of the slopes facing the Northwest and West, there is predominance of higher LS Factor results, suggesting short-length Slopes and steeper inclinations, when compared to slopes facing the East (which had the lower results of LS Factor, suggesting long slopes and with mild inclinations) (Figure 15 and Figure 16).
The high values of LS Factor found in the Northwest and West indicate high susceptibility to soil erosion as they have steeper inclinations and short lengths of slopes when compared to the East slopes.
Table 5 presents the slope orientations of Trindade Island. The slopes facing the Northeast and Southeast directions have predominance over the others, with the percentage of 24.8% and 16.1%. The slopes with lower results are the North-facing ones with 6.5% followed by the Southeast, with 7.7% and Northwest, with 7.8% of the total area (Figure 17).
The slopes facing East can be associated with greater morphogenesis processes, as it is explained by the East wind’s hot and humid prevalence throughout the year, coming from the Equator line.
The island’s northwestern portion has slopes protected from winds coming from the East during most of the year, and also the air masses coming from the South between June and September. Therefore, the slopes were mostly of the Cliff class (inclination higher than 75%) (Figure 11) and the results from the LS Factor are High and Very High (Figure 15).
The West, Southwest and South facing slopes also have particular characteristics, due to these parts receiving Polar air masses coming from the south mainly between June and September, when the South Atlantic Subtropical Anticyclone moves towards the Equator (ANGULO et al., 2018). These parts had a predominance of slopes classified as Mountainous (45% - 75%) and Cliff (>75%), and also had the LS Factor classified as Average and High, indicating smaller morphogenesis in comparison to the Northeast part and, greater when compared to the East part.
The morphological class with the highest percentage was Slope with 51.1%, followed by the Spur class, 19.1% and Hollow, with 18.5% of the total area (Table 6). The lowest percentage was accounted for the Flats and Peaks (each with only 0.3%). The Shoulders had 0.1% and the Pit hardly anything (Figure 18).
The Hollow zones concentrate the fluvial courses, most of them temporary. Among the perennial rivers on Trindade, the Captação river stands out, being the one that is used for human consumption. Also, the Hollow class in some slopes have the form of amphitheatres, mainly on the East part, and with minor occurrence on the West part, being inexpressive on the Northwest side (Figure 19).
The presence of alluvial fan deposits on the outlets of those perennial streams corroborates the association of hollows and fluvial streams, as shown by Angulo et al. (2018). Therefore, the erosion and sediment transportation processes carried out by streams in Trindade stand out (Figure 20).
The Spurs (Figure 21) occur in 19.1%, corresponding to buttresses or secondary water divisors. They are important because they allow the formation of small headwater drainages as described by Marques et al. (2017).
The Peaks and Ridges occur in only 2.7% of the area of Trindade (Figure 22). Their importance lies in determining flux direction, functioning as topographic divisors. The Flat areas occur in just 0.3% of the island and are associated with the coastal plain, sand and gravel beaches and alluvial fan deposits.
The topographic wetness index (TWI) shows the following percentages of the total area: very low (17.5%); low (33.6%); average (30.6%); high (16.9%) and very high (16.9%), as shown in Figure 23.
The TWI High and Very High classes stand out at the East part of Trindade, following the lower results of the LS-Factor in this region, when compared to the West and Northwest parts (Figure 24). Also, in this part, it has concave slopes with the formation of areas of convergence fluxes, as shown in dark blue on the map. This is beneficial for the formation of perennial streams like the one of the Captação River.
The fluvial streams are concomitant with areas of the Hollow morphological class, areas of convergence of fluxes and are represented on the map in dark blue color, such as the watershed of the Captação River.
The Northwest part of Trindade had mainly a TWI classification of very low, with the exception of some areas concomitant with morphological aspect of the Hollow class, indicating the fluvial streams. This land is mainly composed by steep and abrupt rocky walls (Figure 25) a factor that is not favorable to the occurrence of fluvial streams.
In this part, the slope class that stood out was the Cliff type and the LS-Factor results were classified as High and Very High, reinforcing the indications of the TWI. These slopes are protected from the hot and humid winds coming from the East for the majority of the year and also from the air masses coming from the South between June and September.
As pointed by Clemente et al. (2018), the morphological class Peak, in Trindade, is a pedoclimate, colder and more humid. These and the Ridge are lumped together in the central parts of Trindade, with the vast Desejado Sequence standing out (ALMEIDA, 1962). Furthermore, they are areas where the TWI results had mainly classifications of the type Very low and act as water flow dividers.
The analysis made from thematic maps corroborates the relationships between the morphometry and morphology of Trindade Island and to associate them with its evolution over time. As an example, we observed that the East part has higher morphogenesis when compared with the other parts of the island.
The higher morphogenesis at the eastern portion of Trindade is evidenced by the lower values of slope inclination, indicating a flat and gently undulated relief, as well as with the lower results of LS-Factor when compared to the western portion, which had higher values of slope inclination. This relief has its slopes undulated and mountainous along with higher results of LS Factor. We claim that the higher morphogenesis in the East part of the island is due to the climate (i.e. winds, rainfall and temperature), more prominent during the majority of the year and with winds coming from the East.
In the Geomorphons classification, the slopes from the East part are peculiar for having lots of concave areas that are denominated Hollow, which favors water flow concentration. The West part has some convergence zones of water flow, with less expression than in the East part but with greater altimetric amplitude, given that most of its relief is made of cliffs. On the Northeast part of the island, the slopes show cliffs and smaller occurrence of convergence flux zones.
The topography wetness index’s higher classes are found mainly in the eastern portion of Trindade, associated with headwater drainages, in hollow types of morphology. These indexes reinforce the higher possibility of water in the East part of the island instead of the other parts, an important fact to geomorphologic evolution. However, Trindade lacks complementary studies about hydro-sedimentology in order to assess the quantitative and qualitative hydric availability.
To the Laboratory of Environmental Systems Modeling (LAMOSA); to the Extension Department of Ufes (Proex-Ufes); to the Brazilian Navy and its militar for the logistics and support, PROTRINDADE, Secirm, Capes for providing scholarships and to the Post-Graduate Program in Geography of the Federal University of Espirito Santo.
