Abstract
Transport network infrastructure interacts with the earth’s surface because they often share common spaces (e.g. river valleys), such that transport is an anthropogenic pressure that can affect geomorphological processes and outcomes. Since having its profound effect worldwide, the systematic study on the effect of transportation infrastructures (TIs) on the alteration of geomorphological forms and processes has been less focused than on any other anthropogeomorphic driver. The present review provides a multidimensional overview based on the available literature and data on the effect of TIs in changing hillslope and fluvial geomorphology to sustain a peaceful harmony between the transport network and its surrounding landscapes. The study underlines the effect of major TIs like trails, roads, railways, tunnels, causeways, waterways and airports on the alteration of different geomorphological processes on hillslope and fluvial landscapes like the movement of earth material, geomorphic connectivity, slope instability, sediment production, gully initiation and surface runoff. For instance, the global level proximity analysis shows ~40% of landslides happen within the 500 m of any major roads only, while at the regional scale it becomes ~65% irrespective of the degree of seismicity. Due to the fast development of TIs, the mountain regions are more prone to slope instability because of the alteration of surface hydrology by increasing runoff, road and ditch guided concentrated flow, rills and gully formation by reducing drainage area to cross the critical threshold limit. The plain regions are primarily facing the problem of fluvial (dis)connectivity because of the close proximity between river and transport networks and undersized causeways. For sustainable TIs development, factors like the practice of bio-engineering for roadside slope management, de-culverting, 100-year flood return for causeway construction, mapping of river corridors, road water harvesting should be incorporated for less effect on hillslope and fluvial geomorphology.
Keywords
Introduction
The Anthropocene reshapes the ideas of anthropogenic alteration in geomorphological forms and processes with more intensive and quantitative ways (Brown et al., 2017; Tarolli and Sofia, 2016). Technological advancement assists the human to interact with nearly all natural processes of the earth and leaves their traceable imprint on it. Such notions expand the field of exploration from first-order geomorphic processes (e.g. hilltop removal, large reservoirs) to second and lower-order alterations (e.g. soil erosion, deposition) through a wide range of anthropogenic stresses (Brown et al., 2017). Scientific involvement to investigate the anthropogenic alteration of earth surface and processes was initiated by Marsh (1864) in the ‘Man and Nature’, a milestone work on human-induced modification of the physical world with a vast scope of knowledge on anthropogeomorphology including the projection on the effect of waterways (e.g. Suez Canal) on the basin characteristics. In the beginning phase, Gilbert (1917) has also significantly enriched the concept based on the effect of upstream intensive gold mining on in-stream sedimentation problems in the river systems of the Nevada region (USA).
Since the beginning of the 20th century, several research works have been emphasised the influence of anthropogenic activities like construction of dams and reservoirs, channelisation, land use land cover (LULC) changes, urbanisation, mining (in-stream and out-stream), water lifting on the hydro-geomorphological alterations. However, limited attention has been given on the effect of transportation systems (TS) and their infrastructures as a driver of anthropogeomorphology, while its direct interaction with topography has been observed since the early Roman civilisation in 400 B.C. (David et al., 2011). Roman roads were the most durable structure built during the fourth to fifth Century B.C. and the technique led to a significant change in topography and land cover by deep excavation (>1 m) for five-layered paved road and by 60 m wide deforestation along the road, respectively (David et al., 2011). The major kick-off in the development of transport networks was started since the industrial revolution and maximum expansion has been started after World War II (David et al., 2011). The constructional works on earth surface related to the TIs like underground metro systems, subways and tunnels are known as ‘shallow anthroturbation’ (Zalasiewicz et al., 2014), which are mainly extended from 1 to 10 m below the ground level but generates huge excavated earth materials. For example, from the world’s longest railway tunnel with two tubes in Switzerland that is, the Gotthard Base Tunnel (57 km) about 28.20 million tons of earth material has been excavated from below the Swiss Alps (Amberg Engineering, 2018). Hooke (1994) has also considered road-building a central human activity for significant earth movement across the world which becoming a major source of sediment to the nearest river system. The effectiveness of transport infrastructure (TI) to change the earth’s surface and necessity for further research on this field could be assessed from annually ~$2.1 trillion investment on transport sector development across the world (Global Facility for Disaster Reduction and Recovery (GFDRR), 2017). Conversely, ~27% of the global road and railways are exposed to at least one geo-hazard and generate globally expected annual damage of ~$3.1–$22 billion of which ~34% comes through the river flooding only (Koks et al., 2019).
Without mentioning the transport sector, Knighton (1984) has initially categorised two groups of anthropogenic changes in the river system: (a) direct or channel-phase changes and (b) indirect or land-phase changes. Recently, Rhoads (2020) has also skipped the effect of TS on river dynamics. However, early concepts about the fluvio-geomorphic effect of river engineering for TIs like construction of the bridge was came from several engineering literatures in the 1970s (Bradley, 1970; Jansen et al., 1979; Klingeman, 1973; Neill, 1973; Shen, 1971; Task Committee, 1978) and have been also pointed about the problems like scouring, backwater effect, river bed degradation, instability of channel, risk assessment of flooding etc. (see Table 1 in Gregory and Brookes, 1983). Further investigations show that installation of the bridge could double the downstream channel width, carrying capacity and width–depth ratio than the upstream (Douglas, 1985; Gregory and Brookes, 1983). The construction of transport networks on the adjacent lands along the rivers are also significantly disturbed the floodplain geomorphology and lateral connectivity (Blanton and Marcus, 2009; Roy, 2022; Snyder et al., 2002). The interaction between stream lines and road networks is also profoundly observed on the hillslope (Jones et al., 2000). The position of transportation routes on the hillslope determines the degree of interaction between roads and streams, in particular the direct interactions are more common in the middle and lower slope zones than near to the ridges (Jones et al., 2000; Montgomery, 1994).
Major and minor changes in fluvial and hillslope geomorphology during and after airport construction.
In the last phase of 20th century, some important studies have been also investigated the effect of road construction on the surface and sub-surface flow distribution (Megahan, 1972), soil erosion and landslide (Swanson and Dyrness, 1975), mass movement due to weathering effect on the exposed bedrocks by road cutting (Clayton, 1983), effect on topography and stability (Anderson, 1983), formation of gully (Mati, 1984), production of sediment (Reid and Dunne, 1984), runoff generation, channel initiation and slope instability (Montgomery, 1994; Wemple et al., 1996), responses on the pattern of peak flow (Jones and Grant, 1996). While, studies on the respective fields have been intensified from the beginning of 21st century with significant advancement in the methodology and survey techniques used for TS research from the manual land-based geomorphic survey (Gregory and Brookes, 1983) to Unmanned Aerial Vehicles (UAV) or drones based high-resolution topographic change monitoring and sustainable management (Achillopoulou et al., 2020; Greenwood et al., 2019; Outay et al., 2020; Tarolli, 2014; Figures 1a and 1b).
(a) Timeline of the important milestones in the process of transportation geomorphology development and (b) annual frequency of published literature since 1972.
The application of drones on 3D landslide modelling during road construction helps to make an easy and precise estimation of the hillslope geometry with about 95% accuracy in comparison to the land-based total station survey (Gupta and Shukla, 2018). Using high-resolution terrain data, Sofia et al. (2016) have developed the Slope Local Length of Auto-correlation (SLLAC) metric to automatically quantify the effect of road network on hillslope geomorphology. The construction of TI as an effect on process geomorphology is a relatively new field and facing a lack of research interest (Tarolli and Sofia, 2016), although the presence of road seriously influences the degradation of landscape in form of landslides, soil erosion, altering riverscape and these are all emergent issues in the United Nation’s sustainable development goal practices. As per Sidle and Ziegler (2012), sustainable land management is not possible without preserving the impact of roads and trails on soil erosion and landslide. Therefore, more insights on systematic consideration of new and upcoming threats to geomorphology (e.g. climate change, expansion of transportation system) is now essential to reduce disturbances in the development of infrastructures and environmental sustainability. With this, the primary aim of this review is to make a systematic insight on the multi-dimensional effects of transportation system infrastructures on hillslope and fluvial geomorphological forms and processes in particular to sustain a peaceful harmony between the transport networks and concerned landscapes and to address the generated problems at the planning stage.
Design of the study
An inclusive bibliographic review has been carried out to shortlist and organise the published works like research articles (e.g. Blanton and Marcus, 2009; Jones et al., 2000; Montgomery, 1994), review articles (e.g. Zgłobicki et al., 2021), book chapters (e.g. Pijet-Migon and Migon, 2018), reports and technical manuals from scientific organisations (e.g. Castro, 2003), which are primarily dealing with the impact of transportation infrastructure on the modification of fluvial and hillslope geomorphology. The search operation has been led with different keywords related to the transportation infrastructures and fluvial and hillslope geomorphological forms and process combinedly presence in the title, abstract and/or keywords of the manuscripts available in the database of ScienceDirect, Wiley Online Library, EndNote, Google Scholar, JSTORE (Figure 2). A total 120 works of literature have been shortlisted, whereas numerous conceptual literatures on fluvial and hillslope connectivity (e.g. Poeppl et al., 2020; Wohl, 2017; Wohl et al., 2019) are also followed to established the concerned fact of research (Supplemental 1). The period of publication is extended from 1972 to 2022, however significant development has been observed since 2000 only, because up to 1999 the average annual publication was one only which becomes four during the 21st century (Figure 1b). The tabulated literatures are consisting with 90 research articles (~75%) and 11 review articles (9.2%), which are either indexed in Scopus or Web of Science (WOS) or Science Citation Index (SCI) (Figure 2). In terms of the number of contributions, the ‘Earth Surface Processes and Landforms’ (Wiley) and ‘Water Resources Research’ (Wiley) journals are leading with 10 and 6 published articles, respectively (Supplemental 1). In addition, two books, five chapters, eight reports/manual and one commentary and an electronic thesis have been also followed in the current review. The country-level contribution shows that about 34% of the literature is coming from the United States of America only and their study area are mostly concentrated on the mountain region of the Pacific Northwest and/or Western United States (e.g. Oregon, Idaho, Colorado) (Figure 3a). Countries from Southeast Asia are contributing 25% of the literature in this field whereas India and China are also individually contributing 12 (10%) and 7 (5.8%) articles, respectively. European countries are also contributing 22 (~18%) articles, whereas 15 (~12.5%) articles are addressing the problem globally.
Framework of method followed in the current study for shortlisting the relevant literature.
(a) Country-level distribution of reviewed literature source, the yellow-coloured countries have contributed one article each and grey colour indicates no literature and (b) Work tendency on the effect of transport infrastructure on different sub-fields of fluvial geomorphology.
In Figure 3b, the shortlisted literature portrays that nearly half of the works (~40%) have investigated the role of TIs on sediment generation and/or soil erosion by enhancing sediment delivery zone from the deforested area for the construction of trails, paved and unpaved roads within the forest area for biking and hiking activity, logging and management, respectively. In addition, ~17% of works also deal with the changing flow regime after road surface induced concentrated runoff entered into the surface and channel hydrology, which also helps to estimate the soil erosion, gully formation, sediment generation from the road surface. About 23% of works are concerned with the alteration of channel morphology by TIs, while ~24% work has been given special attention on the lateral and longitudinal (dis)connectivity due to alignment of transport network within the floodplain area and in-stream crossing structures. The construction of roads on hillslopes also significantly alters the stability of slope units in the highland and causes sediment dynamics within the river basin (Sidle and Ziegler, 2012). A major focus with ~20% of works has been also noticed on this aspect, whereas ~7.5% of which particularly has focused on the formation of rills and gully from the road surface runoff. Therefore, due to the availability of adequate literature on fluvial and hillslope geomorphology in particular, the primary aim of this review is to summarise the findings from sporadic publications on the impact of TIs on the altering processes of hillslope and fluvial geomorphology. Based on the existing research gaps, the study has also highlighted the future research prospects on the transport infrastructure induced alteration in fluvio-geomorphological forms and processes.
In terms of the quantitative assessment of TIs inducing changes in fluvial and hillslope geomorphology, different secondary metadata has been processed and mapped including different quantitative findings from the reviewed literature. To established the fact of hillslope alteration by transport networks, proximality analysis has been done between available landslide inventory data points and major road networks at global-scale as well as at regional-scale for India. For global-scale, a total of 6788 data points of Global Landslide Inventory from NASA’s Open Data Portal and major roadways by Global Roads Open Access Data Set from NASA’s SEDAC have been used to get proximity statistics. While for India, the landslide inventory data points from Bhukosh of the Geological Society on India (bhukosh.gsi.gov.in) and up to level three (i.e. State Highways) roads from Open Street Mapping (OSM) have been used for proximity mapping over the Western Himalaya (WH) (Seismic Level IV–V) and the Western Ghat (WG) (Seismic Level II–III). The comparative evaluation between WH and WG helps here to neutralise the problem of co-seismic landslides from anthropogenic (i.e. TIs) landslides, as Tanyaş et al. (2022) have also done a comparative assessment to show the similar effect of new roads on mass movement in a non-seismic zone (in the Arhavi area, Turkey) as like mass movement by the earthquake in a seismic zone (in the Gansu Province, China). To signify the interaction between fluvial processes and TIs, 60 year’s data on flood induced damages of public utilities (mostly comes from TIs losses) over the lower Gangetic belt of India has been analysed along with previous field-based research data on the potential interaction between transport networks and fluvial connectivity over Eastern India (Roy, 2022; Roy and Sahu, 2017, 2018). In particular, a detailed geomorphic survey followed by the standard protocols and flow modelling in Hydrologic Engineering Centre’s River Analysis System (HEC-RAS) have been carried out at selected crossing sites from the plan land of eastern India to estimate crossing structure induced changes in channel morphology and hydraulics. In addition, to understand the importance of standard river engineering on protecting channel morphology, equations to estimate required dimension of crossing structure have been used on the survey sites in response to the upstream channel characteristic (see Roy and Sahu, 2018).
Major transportation infrastructures (TIs) and their mode of interactions with hillslope and fluvial geomorphology
TIs generally contain fixed installation of essential structures, which are varying with the changing mode of transportation for example, air, land (road and railway), water, pipeline and space, in addition to different auxiliary arrangements like terminals, ports, warehouses, stations for rails and buses, etc. All such developments are directly or indirectly interacting with fluvial and hillslope geomorphology (Figure 4).
Schematic model to represent the effect of human actions related to the TIs development on the altering fluvial and hillslope geomorphology and possible outcomes.
Rural roads and footpath or trails
Rural roads are always the backbone of the rural economy for every country, which are normally carrying a low volume of traffic, mostly unpaved, and constructed by limited budgets (Kocher et al., 2007). Rural roads are commonly located in the forested and rangeland setting, which are serving as residential, recreational and resource management uses by connecting intra and inter villages or habitats or localities. India holds the largest and densest rural road network in the world with a total of about 3.3 million km, most of which are not supporting its serviceability in all weather conditions (World Bank, 2011). A footpath or trail is an unpaved road constructed by substantially altered vegetation and topsoil structure caused by the human movement on the grasslands, forests, scrublands, hillslopes primarily for recreational purposes for example, outdoor sports like hiking, off-road biking and cycling, horseback riding, etc. and popularity of which was increasing very fast since the 1970s (Callahan, 2008; Salesa and Cerdà, 2020). To extend the spatial connectivity in forested, mountains and inhabitant regions of the world, thousands of miles of new rural roads, footpaths and trails to be constructed in near future (Sidle and Ziegler, 2012). The poorly located, designed and maintained are to be the primary cause of water quality degradation and alteration of surface hydrology and correspondingly make changes to hillslope and fluvial geomorphology (California Department of Forestry and Fire Protection (CDFFP), 2003; Kocher et al., 2007; Salesa and Cerdà, 2020).
The fluvio-geomorphological impact of such transport networks comes through the alteration of landscape components like increasing soil compaction, reducing soil moisture, loss of organic content and soil strength (Brevik and Fenton, 2012), removing the vegetation cover and surface biomass (Gyasi-Agyei et al., 2001), increasing runoff by a higher concentration of overland flow and raindrop splash (Wallin and Harden, 1996), by exacerbating landscape connectivity along the compacted wheel track transfer larger sediment to the nearest streams (Basher and Ross, 2001; Sidle and Ziegler, 2012) and enhancing the topsoil loss, which is making the earth surface more prone to weathering, mass movement and erosion (Salesa and Cerdà, 2020) (Figure 5a). The direct impact of such alteration will face by the adjacent river system and valley slopes with a higher amount of water and sediment input and increasing the soil erodibility by altering infiltration capacity and structural stability of the soil, respectively. The experimental research by Voorhees et al. (1979) shows that due to the wheel activity the process of soil compaction may happen up to a depth of ~300 mm, which significantly reduced the porosity of soil and corresponding reduce precipitation storage capacity and increase runoff and related soil erosion. However, Wang et al. (2016) have noticed the differentiation in production of sediment from the different shape of unpaved road surface. The concave road generates highest amount of soil loss in compare with the convex, lateral and flat with higher flow velocity and erosivity. The worldwide estimated soil loss from trails ranges from 6.1 to 2090 Mg/ha/year (Salesa and Cerdà, 2020), although there is limited work on this field. In the Drakensberg Mountains of South Africa, trail erosion is one of the serious problems of land degradation, which is very difficult to control (Garland, 1990). Dunne and Dietrich (1982) observed that only 2% of aerial coverage by rural roads and footpaths in a densely populated region of Kenya has aggravated 25%–50% of the total soil erosion. López-Vicente and Navas (2009) also noticed in the Spanish Pre-Pyrenees basin only 2% surface coverage of mountain trails is responsible for up to 14% of the total erosion in the catchment.
(a) Typical example of trail erosion and formation of rills on trails around Folsom Lake, 2021 (P.C.: Kevin Knauss, 2021; insuremekevin.com), (b) an example of road surface runoff on NH-94 (India), which is choked by landslide debris during heavy rain on 21st September 2021, (c) landslide of Eastern Himalaya has disconnected the NH 110 and Darjeeling Himalayan Railway during the rainy season of 2021, (d) an example of land reclamation from South-China Sea for the artificial Hong-Kong International Airport, (e) debris sliding along the cut slope for a road, and (f) deep scouring at outlet and downstream bank failure for a box culvert.
Sunken Lanes
Sunken Lanes (SLs) are the typical form of the incised roads or tracks on the ground with an obvious ancient legacy and commonly form by combining actions of the passage of animals, humans, vehicles and gravitational flow of water and wind (Boardman, 2013; Zgłobicki et al., 2021). The unconsolidated and soft geology is the key condition to form SLs upon which the erosional actions are taking place profoundly, therefore a major appearance of it is showing in the European loess belt (De Geeter et al., 2020). Geometrically, these routes are widely varying in length (30–2300 m), top width (2–36 m), and depth (0.6–12.5 m) (Zgłobicki et al., 2021). Boardman (2013) has highlighted the importance of SLs on geomorphic connectivity by enhancing the flux of sediment and water from hillslope to valley bottom. SLs are also called ‘road gully’ (De Geeter et al., 2020; Zgłobicki et al., 2021) and ‘muddy flooding’ (Boardman et al., 2019) for their explicit role in soil erosion and sediment generation from their unprotected bank sides and beds, therefore it becomes an active erosional feature on the fluvial landscape with more than 70% source of sediment from the gullies formed on the steep sides of the lanes (Boardman, 2013). During the rainy season, SLs are acting as runoff channels and pathways for transferring enormous amounts of water and sediment to the river system (Zgłobicki et al., 2021). The movements of humans and vehicles are also reducing the infiltration capacity and generating a larger volume of runoff and also becoming hotspots for soil erosion by initiating large gullies around the lanes and correspondingly changing the channel morphology by the process of intensive sedimentation (Boardman et al., 2019).
Roadways and railways
Roads are a track or route on land connecting two or more places with a stabilised base and open to public traffic supported by different other structures like culverts, bridges, crossings, tunnels, junctions etc. (Organization for Economic Co-operation and Development (OECD), 2002). In the case of rail transport, wheeled vehicles (trains) are running on the permanent rail tracks usually constructed by a pair of steel and installed sleepers. Construction of roads and railways includes major engineering works like removing of a huge amount of earth and rock materials through digging and blasting, filling of depression, construction of embankment to uplift the tracks from the surrounding ground, construction of tunnels, bridges, culverts, ditches alongside the roads and removal of vegetation, which are directly interacting with fluvial systems and hillslope geomorphology. The impervious surface of road also acts as a dominant pathway for rainwater and enhances the water volume of the nearest stream (Figure 5b). Montgomery (1994) noticed that the establishment of artificial drains/ditches beside(s) the roadways increase the catchment area and also helps to concentrate large amounts of water to a particular outlet. As a result, in the immediate downslope of the outlet initiation of gullies is a common problem for hillslope. The cut-and-fill works also significantly altered the hillsides, causing instability in hillslope debris flow, and inducing landslide hazard (Harabinová, 2017; Siddique and Khan, 2019; Figure 5e). Such problems arise more often in the hilly regions of Himalayan (Agarwal and Dixit, 1986) and Oregon’s Cascade Range (Jones et al., 2000) during the rainy season. Petley (2021) has also noticed more than 70 landslides along the Oxley Highway of New South Wales (Australia) after heavy rainfall. Along with landslides sometimes the portion of rail and roadways are also slippery and discontinue the land transport (Figure 5c), which requires huge money and time for restoration. Konagai et al. (2005) have also reported that an earthquake (M6.8) was triggering a number of landslides in the Higashiyama Mountain district (Japan) and resulted in the closure of 233 route segments and isolated 61 localities.
In the context of fluvial geomorphology or geomorphic connectivity in particular (Wohl et al., 2019), embankments of road and rail networks are acting as artificial barriers on the earth surface and having a significant effect on the three-dimensional (dis)connectivity (e.g. lateral, longitudinal and vertical) and/or fragmentation of fluvial landscape by obstructing the free movement of water, material/sediments and energy, especially in river system (Blanton and Marcus, 2009, 2014; Roy, 2022; Roy and Sahu, 2017, 2018). Constructions of the lateral road are also destroying forests nearby the transport routes, because the zones close to roads are relatively easy to access (Wang et al., 2014) and makes a fragmented forest more than a clear-cut (Reed et al., 1996) and create routes of soil erosion and sediment supply to the local rivers. For example, in the Amazon about 95% deforestation has occurred within a 5.5 km buffer of road networks (Barber et al., 2014). The poorly located and functioning forest road (public or private roads exclusively for forestry – conservation, management and logging) can significantly alter water quality, wildlife, and causes significant soil loss and generates huge sediment from the road and its surroundings by changing slope stability, drainage pattern and overland flow (Misir et al., 2007; Nunamaker et al., 2007).
Tunnels
Tunnelling is a most advanced civil engineering structure to cross the obstacles on terrain (e.g. hills, snow cover, water bodies) by an underground passage in the way of transport networks. The Atal Tunnel of India (2020) is the longest highway tunnel (9.02 km) in the world, a profound example of an anthropogenic act on the geomorphological alteration in a high mountain pass of Himalaya (Pir Panjal Range). Excavation of soil and rocks is a mandatory task for tunnelling and producing large amounts of earthworks, which are stored nearby places and become a major source of loose sediments to the river system (Healy, 2016; Robbins et al., 2021). Tunnelling below the mountain and hilly region is also reviving the early landslide by triggering the stress redistribution on the unstable hillslopes (Jiao et al., 2013; Li et al., 2019; Wang et al., 2020). Such underground engineering work also acts as a barrier to groundwater flow and generates several co-actions on the surface (Liu et al., 2015; Pujades et al., 2015).
Waterways
Engineering adjustment is essential for the inland waterways like rivers and canals (e.g. White Sea-Baltic Canal, Suez, Panama etc.) to sustain enough depth and width, removing obstacles (e.g. sharp meandering bend, rapids, waterfalls, sediments, shoal etc.), maintaining water volume and current to accommodate the vessels and round the year navigability. In the process of removing obstacles for navigability, dredging of the river bed, straightening and channelisation are major anthropogenic alterations of fluvial morphology in most direct form (Gore and Petts, 1989; Gregory, 2001; Tockner et al., 2010), which also promotes change in the bathymetry, planform, longitudinal profile, cross-section, pool riffle sequence and flow patterns of rivers (Boon et al., 2010) and also generates possible threat to the aquatic ecosystem (Brooker, 1985; Hohensinner et al., 2018; Hondorp et al., 2017). The vessels with propellers have significantly increased the turbulence and velocity close to the river bed, causing scouring, increasing sediment load by elevated the fine sediments from the bed (Hey, 2006). The activities like boat wash, vessel navigation during low to medium water and passing close to the bank have also accelerated the erosion of banks and beaches (Garrad and Hey, 1987, 1988; Ten Brinke et al., 2004). Nevertheless, construction of infrastructures and dredging for marine navigation (e.g. ports, harbours, jetty, breakwaters, groyne, sea wall etc.) are also altering the geomorphology of coast like changing shoreline (Guerrera et al., 2021; Kundale, 2010; Mohanty et al., 2015), erosion and accretion of beaches (de Boer et al., 2019; Tsoukala et al., 2015), closure of river/creek/lagoons mouth (Kankara et al., 2018) by altering condition of the local wave, currents and sediment transport processes (Kundale, 2010).
Airports
To ensure the continuous serviceability in the aviation industry with the exponentially increasing global demand of mobility for passengers and freights (http://www.iata.org/), construction of new bigger airports or expansion of existing ones is now an urgent requirement as transport infrastructure development for every country. In this context, the alteration of natural topography through ground engineering is a primary anthropogeomorphic change for the essentially flat and stable ground by removing major to minor topographic highs, filling the depressions, and also channelisation and/or diverting the closest rivers for the project (Pijet-Migon and Migon, 2018). According to the Airports Council International (ACI), globally around 17,678 commercial airports are currently active, which are receiving airliners, cargo and business aircraft, whereas, the figure increases to 48,556 if all airports (active or inactive mode of operation) are included like aerodromes and airfields, both civilian and military throughout the world (Favargiotti, 2018), where ~10,000 airports have been increased only in past 10 years (World Airport Traffic Forecast (WATR), 2019).
Airports also encounter unique erosion issues. The large impervious surfaces formed by runways, taxiways and terminals create considerable stormwater runoff, which helps to a rapid increase in adjacent stream flow and subsequently increases stream erosion. Alteration of fluvial and hillslope geomorphology by airport construction has been scarcely studied except the ubiquitous interest in earthworks until the end of the 20th century. Douglas and Lawson (2003) have initially started the drive on the geomorphological dimension of airport construction. With the help of several case studies (Hong Kong, Osaka Kansai, Singapore, Incheon-Seoul), Douglas and Lawson (2003) have been able to highlight the effect of airport construction on changes in landforms and channel morphology. However, there is negligible or no attention on this aspect even in the recently published anthropogeomorphic books like Goudie and Viles (2016), Das et al. (2021). The processes of land reclamation (new Hong Kong Airport) and artificial island generation for airports (Kansai Airport, Japan) are now trending anthropogenic activity to modify regional geomorphology (Dávid et al., 2010; Pijet-Migon and Migon, 2018; Figure 5d). The multidimensional effects of such engineering work could be summarised as major, minor and second-order changes in fluvial and hillslope geomorphology (Table 1). The increasing number of obsolete and abandoned airports worldwide is a matter of concern for area of soil erosion by initiating rills and gullies around the abandoned sites (Figure 11c and d).
Causeways
Causeway is a frequently used term to denote the point of intersection between any river and road or railway. Identifying any river without a single causeway will be a challenging job. There must be a presence of any form of causeway either ‘low-water crossing’, a structure implemented on a road with average daily traffic of fewer than 25 vehicles, vented or unvented (<6 inches of flow depth) and water run over the roadway at the high flow of stream (Gautam and Bhattarai, 2018; Ring, 1987) or a ‘road-stream crossing’, which is any bridge or culvert passing over a creek, river, stream or formed channel (Melbourne Water, 2011). Bridges are defined legally as structures with a centreline span of 6 m or more, whereas, the culvert is a crossing structure with 6 m or less in centreline span width between extreme ends of openings in different shapes (Federal Highway Administration (FHWA), 2012; Indian Roads Congress (IRC), 2004; National Bridge Inspection Standards (NBIS), 2006). The construction of roads to facilitate transportation in the low-lying tidal marshes is also known as a causeway and has a significant effect on the process of sedimentation in the coastal region (Knowlton et al., 2017).
Apart from the significant effect on river ecology (Boulton et al., 1998; Bouska et al., 2010; Bravard et al., 1986; Resh, 2005; Wellman et al., 2000), such crossing structures are also effectively altered the geomorphology of river system (Roy and Sahu, 2018) (Figure 5f). Federal Highway Administration (FHWA) (1990) has reported the significant general and local effects of highway and bridge construction on the geomorphology and hydraulics of river systems and categorised them into two types: immediate and delay (Table 2).
Short-term and long-term effects of in-stream highway, bridge and/or culverts construction on river geomorphology and hydraulics.
(Source: Federal Highway Administration (FHWA), 1990).
Effects on major geomorphological processes on hillslope and fluvial landscapes
Movement of earth materials from hills and rivers
Due to the construction work, a massive volume of earth materials is moved or filled in the on-site or off-site situation, this is popularly known as earthworks (EW). Wolman and Schick (1967) have noticed that the construction work induced soil erosion in a single year is higher than the decades of erosion from natural and agricultural activities. The quantity of EW varies with the type of transport infrastructures as well as the configuration of the existing topography or terrain, which is calculated by multiplying the sectional area with the length. The cutting and filling method for the construction of any road is generally carried out with respect to the formation line (FL) after considering the longitudinal and cross-sectional profile of the ground. The FL is an imaginary line drawn to level the ground, which depends on the regional hydro-geomorphological setting in particular the regional flood level, gradient, height of the bank, soil type etc. The required volume of EW is estimated through the longitudinal profiles of the original terrain and required FL or proposed road track, although may have some uncertainties in the real world (Mohamad Karimi et al., 2007). Higher profile difference between actual ground level and required FL indicates the need for extra EW cutting and/or filling and visa-versa. Sthapit and Mori (1994) have explained the importance of the hill-factor (practical cutting depth) and slope-factor (country cross slope) on the estimation of earthwork in the hilly region. An experimental study by them on a 10 km road with a formation width of 5 m under the Sandhikhark Project (Nepal) has estimated that as per the modelled equation the total required earthwork volume is 377,511m3, which is 345,644m3 on the actual ground. Therefore, the validity of this model of earthwork estimation is acceptable with only 9.2% error which is quite low than the desired accuracy set by the respective authority of Nepal. Based on an assumption that with the density of 2 t/m3 about 10m3 of earth material is moved for per metre of road, in the USA only around 3.8 billion tons of earth materials have been excavated and moved per annum for road and related infrastructures as direct anthropogenic interaction with the earth surface (Hooke, 1994). The world’s longest railway tunnel project with two tubes in Switzerland (i.e. Gotthard Base Tunnel) has also excavated about 28.2 million tons of earth material from its total 152 km of the tunnel bore (Simoni, 2014). Road construction has a profound effect on sediment production also (Goudie and Viles, 2016), for example, road cutting in Georgia has been increased the sediment yields up to 20,000–50,000 t/km2/year (Wolman and Schick, 1967). Such activities are sometimes directly impacted the hillslope and fluvial geomorphology, for example, Figure 6 shows the ongoing massive EW and related alteration of river beds and hillslopes for a new road project at the foothill of Darjeeling Himalaya (India). The road work on the hilly section has been significantly disturbed the hillslope by removing the vegetation canopy and top soils (Figure 6a), which might be the cause of new landslides and severe soil erosion. On contrary, in the plain terrain of the foothill section, the filling process is predominant to prepare the road embankment for which an enormous amount of earth materials has been excavated from the nearest river bed (Figure 6c) including the cutting materials from the hills.
Earthworks on hillslope: (a) and nearest river bed of Ghees, (b) for the construction of a new road and filling the embankment as well as approach road and (c) at the foothill of Darjeeling Himalaya.
Fluvial (dis)connectivity and altering channel morphology
The fluvial connectivity, an inherent part of geomorphology, helps to sustain the process of material and energy transfer and/or storage within the fluvial landscape (Brierley et al., 2006; Wohl, 2017). TIs are significantly disturbing the connectivity of landscape, hydrology, and sediment in all three dimensions like lateral, longitudinal and vertical through creating obstacles between different components of the riverine landscape (Blanton and Marcus, 2009). While SLs have generally increased the fluvial connectivity by transporting large amounts of sediment and runoff from hillslope to the valley bottom after generating within the lanes and/or collecting from surroundings (Boardman, 2013). Measuring disconnectivity in the process of sediment flux (Ondráčková and Máčka, 2019; Poeppl et al., 2020), hydrological and geochemical transfers (Fuller and Death, 2018; Lemm et al., 2021), multidimensional coupling on hillslopes (Ayoubi et al., 2021) is an emerging field of research in the anthropo-geomorphological investigation. However, most of these studies have overlooked the role of TIs on the alteration of such geomorphic processes, although roads and railways are ubiquitous features in all landscapes. The study by Blanton and Marcus (2009) has revealed the factor of topographic relief on the level of interaction between the transport network and river system. In the region of plain and wide alluvial valleys, the rate of disconnectivity is higher than the rivers with narrow valleys in the hilly regions. The floodplain is an essential component of riverine lateral connectivity, which is occupying only ~1.4% of the global surface but containing about 25% of the terrestrial ecosystem (Tockner and Stanford, 2002). The presence of transport networks on the floodplain is disturbing it by lateral disconnectivity between channel and floodplain. Blanton and Marcus (2014) and Snyder et al. (2002) have obtained that in the USA about 60% of the Holocene floodplain area is laterally disconnected by roads and railways, which is a considerable limitation for stream and floodplain restoration. The study from the alluvial plain of eastern India has also noticed that about 21% floodplain area is laterally disconnected by linear TIs and the proximity between roads and channels are reducing swiftly over the time of infrastructural development (Roy, 2022; Roy and Sahu, 2017). Figure 7 could also enlighten such problems across India, where in the past 60 years the trend of flood-affected areas has not been significantly changed and although having a negative trend, whereas, the normalised value (based on Wholesale Price Index) of public unitality loss (mostly comes from TIs damage) has been increased significantly (MK 0.399; p < 0.0001). This result helps to conclude that although the flood-affected area is nearly same, however, the construction of public utilities mainly TIs have been increased within the floodplain region of Indian rivers.
Scatter plot shows the converse trend of floods affected area and damages to the public utilities in India and its two most flood prone states West Bengal and Bihar.
Apart from the profound effect of dam on longitudinal disconnectivity by changing flow regimes, sediment fluxes, and altering channel morphology (Petts and Gurnell, 2005), the TIs like bridges, culverts, causeways are also significantly influencing all physical as well as ecological components of the river system. Natural rivers at the site of road-stream crossings (RSC) are negatively affected by direct sediment input, channel deepening and widening, inter riffle distance, scouring, and input of solid wastes (Bouska et al., 2010; Douglas, 1985; Merril and Gregory, 2007; Roy and Sahu, 2018; Thomaz and Peretto, 2016; Thomaz et al., 2014). Bouska et al. (2010) have noticed the crossing type-level effect on stream geomorphology and the significant alteration in channel morphology between upstream and downstream of the crossing sites. In particular, the mean riffle spacing has been nearly doubled in the downstream (8.65 m) in comparison with the upstream section (4.4 m). According to Douglas (1985), channel enlargement takes place at the immediate downstream of bridges only because of the constricted channels under the bridge and associated raise water level with the higher flow velocity. Thomaz and Peretto (2016) have estimated that suspended sediment and discharge are increasing about 413% and 50% respectively at the crossing point of the headwater streams. The effect of longitudinal disconnectivity on the floodplain ecosystem through the approach roads of numerous bridges has not been also studied yet (Roy, 2022).
The result from field investigation of selected crossing sites shows the parameters of channel morphology has been significantly changed towards downstream in comparison with upstream (Table 3 and Figure 8a). In particular, width of the channel (w) has been significantly increased (~23%) at the immediate downstream only not for the entire reach, whereas, mean (d) and maximum (D) depth of channel, cross-section area (a) stream power (ω), stream velocity (v), Froude Number (Fr) are significant increase towards the downstream up to 50 m (Table 3). The impact of pipe culverts is more prominent than box culverts and small bridges. The shape of the channel is also influenced by crossing structures, the asymmetry index of channel (Knighton, 1981) shows that close to the crossing structure shape of the channels are becoming symmetric and it becomes asymmetric with increasing distance. The principal of fluvial geomorphology learns natural streams are typically exhibiting asymmetrical cross-sectional form over much of their length (Leopold et al., 1964; Schumm et al., 1987), whereas streams form is highly symmetric when modified by human activity (Rayburg and Neave, 2008). The values of sinuosity are also significantly decreased in the downstream and channels turn straight in comparison with the upstream sinuous planform (Figure 8c). It is also estimated that the dimension of installed crossing structures is significantly lower than the required which causes profound alteration of channel geomorphology (Figure 8d). Another study by Suvendu (2013) shows that the formation rate of in-stream bars has increased and particularly accumulated around the crossing structure after its construction, and the braiding index has increased to 1.27 from 0.34 in-between 2003 to 2011. The inter-pool distance has also been nearly doubled in the downstream in comparison to the upstream on the same site.
Crossing type wise differences in the mean values of channel parameters between upstream and downstream reaches.
(a) Significant alteration of channel morphological parameters from upstream (U) to downstream (D) at selected crossing sites, (b) changing channel shape with the proximal relationship between crossing structure and channel, (c) alteration of channel planform in the downstream of crossings structure, and (d) structural and hydraulic inefficiency in the construction of different type of river crossing structures [P-Pipe Culvert, B- Box Culvert, BI- Small Bridge].
The vertical connectivity defines the surface-subsurface interaction of water, sediment, and nutrients, which is popularly known as the ‘Hyporheic Zone’ (HZ; Brierley et al., 2006). HZ is a layer composed of the shallow, saturated sediment below and to the sides of the stream bottom (Schindler and Krabbenhoft, 1998). The sustainability of HZ depends on the water movement, permeability, substrate particle size, resident biota and physicochemical features (Boulton et al., 1998; Olsen and Townsend, 2003), which are now significantly controlled by different in-stream TIs like river crossings, bridge piers etc. A study by Merril and Gregory (2007) finds that at all types of crossing structure (arch, box, bridge, pipe) depths of HZ has been significantly decreased in the downstream direction due to the scouring by higher velocity of restricted flow from the crossing structure to the downstream channel. The ecohydrologists are primarily focussing on this aspect, whereas, no systematic geomorphological investigation has been done on the problem of vertical disconnectivity.
Instability of hillslope
Slope instability is triggering the mass movement ranging from small soil creep to large landslides (Lee and Ho, 2009). Landslides are frequently reported from different hilly regions of the world (Mondini et al., 2021). The major factors behind the instability of mountain slopes are steepness, cut slope geometry, alignments, strength and permeability of underlying rocks, composition and depth of soil, and major triggering factors are like the intensity of rainfall, groundwater condition, land use practice, toe erosion, seismic acceleration and faulty engineering practices (Hearn et al., 2021). Anthropogenic imprints such as urbanisation, deforestation, dam, and expansion of transport networks are the primary extrinsic causes of slope instability and corresponding landslides (Kellerer-Pirklbauer, 2002; Skilodimou et al., 2018). Discontinuities of rock with underlying geology and geotechnical parameters by the excavation of natural hillslopes during the cut and filled method of road construction is the primary cause of slope instability (Sangra et al., 2017; Siddique et al., 2020). The process of road cut makes the natural dip very steep and prone to landslide (Hack, 2016). Construction of road in any flysch section of the hill where slope value is more than 15% and presence of 3–10 m of thick soil layer is considered as a high-risk zone of landslide (Arbanas and Dugonjić, 2010). As roads can relevantly alter overland flow (Montgomery, 1994), the torrential rainfall in Vietnam during October 2020 promotes multiple landslides along the National Highway 12A and the government has to spend about $15 million of repair cost for the highway (Petley, 2020). Similarly, the Himalayan region of India, a world-famous tourist destination, is also facing the problem of landslides every year mainly during the rainy season along the major roads and railways (Batar and Watanabe, 2021; Siddique et al., 2020; Figure 5c and e). An ongoing railway project (North Bengal-Sikkim railway link, 52.7 km) of the Northeast Frontier Railway (NFR) of India, passing through the steep terrain of Darjeeling Himalaya with 14 tunnels (32 km) and 17 bridges, has triggered and reactivated multiple landslides in the surrounding region (Singh, 2019). Tanyaş et al. (2022) have pointed that out of 557 events of mass movement in Arhavi, Turkey ~90.1% happened only during the period of road construction with closely associated with the roads, in particular ~88% of them occurred within a 100 m road-buffer zone.
The world level proximity analysis shows a typical allocation between the landslide inventory points (6788) and major road networks across the globe (Table 4 and Figure 9a), about 40% of global landslides occurred within the 500 m distance from the nearest roadways, and half of the landslides happened within a kilometre only. While at the regional scale, the proximity statistic is more relevant to establish the fact of road induced slope instability with more than 60% landslide events within the 500 m of any level-three road networks over the mountain ranges of Western Himalaya and Western Ghat in India (Table 4 and Figure 9b and c). Though situated into different levels of seismic zones no significant differences have been observed regarding the occurrence of landslides within the specified buffer zones of road networks (Figure 9d).
Proximity analysis between the location of landslide points and major road networks in different spatial scale and in different seismic zone.
Source: Global Landslide Inventory form NASA’s Open Data Portal; Global Roads Open Access Data Set from NASA’s SEDAC; Landslide Inventory of Geological Society of India for Western Himalaya and Western Ghat.
Distribution of the landslide inventory and alignments of major roads at global-scale (a) as well as at regional-scale in India (b and c) with a comparative assessment between proximity statistics among the three maps (d).
Riverbank failure around the crossing structures is also a commonly observed incident related to TI induced slope instability (Figure 5f). The undersized culverts and bridges are creating backwater and clogging effects on the upstream of crossing structure, which led to reduce the capacity of channel by sedimentation and also widening channel by bank erosion (Roy and Sahu, 2018). The higher flow height and volume of water at crossing structure are generating stationary waves with higher velocity than downstream mean flow velocity (Huggett, 2007). Such stationary waves are widening channels by acceleration shear stress and roughness in channel wall and making symmetric channel shape with steep bank slope (Leopold et al., 1964; Roy and Sahu, 2018). Such steep slopes are a having higher probability of bank failure and causing the collapse of crossing structure by bridge pier loss and crumble of culvert abutment.
Runoff generation and surface flow dynamics
Road surface works as a hydrological pathway and significantly influence in changing time, volume and spatial distribution of the runoff (Croke and Mockler, 2001; Jones and Grant, 1996; Montgomery, 1994; Wemple et al., 1996). Tarolli et al. (2013) have recognised the significant effect of roads and trails on changing routes of surface flow using high-resolution topographic data like Terrestrial Laser Scanner (TLS), Airborne Laser Swath Mapping technology (ALSM), LiDAR (Light Detection and Ranging) in mountain areas. Hydrologic response of road surface is an important issue for hilly as well as erosion-prone regions where road runoff plays a critical role in accelerating soil erosion by reducing hydraulic conductivity and threshold limit (Montgomery, 1994). Intercepting the infiltration of rainwater road surface generates higher runoff than agriculture, forest or any other natural land cover (Ziegler and Giambelluca, 1997; Ziegler et al., 2000). An event-based experimental study shows that the surface runoff coefficient for paved and unpaved roads are >90% and ~80%, respectively, which is only 0%–20% in the agricultural fields, therefore the lower depth of rainfall is adequate to produce overland flow and subsequently increase soil erosion (Hou et al., 2020; Ziegler et al., 2000). The insufficient penetration of rainwater by impervious road surface also affects the process of sub-surface flow and groundwater recharge (Wemple and Jones, 2003), and the consequent result shows by severe water shortage, flooding, and waterlogging conditions in the urbanised area with a higher density of paved road (Gu et al., 2018; Kramer, 2013; Xiong, 2016). The cut-and-filled roads are constructed generally parallel to the hillslope contours, which act as sub-catchment and collecting surface and sub-surface flow from the above segments of the hillslope (Wemple and Jones, 2003). In such processes, the roadside ditches are transferring the water very first to the nearest streamlines via culverts and altering the storm hydrograph (Jones and Grant, 1996; Wemple and Jones, 2003). The impervious area of any airport also generates enormous direct runoff and works as a major source of wastewater, which is significantly change the hydro-geomorphology of surrounding rivers through improved sewer systems, channelisation and re-routeing of channels. Different operations of airports are also contaminating the river water by ejecting pollutant-mixed wastewater (Sulej et al., 2011).
Sediment production
About 75% of the earth’s land surface is facing traceable anthropogenic pressure (Venter et al., 2016), which significantly increase the production of sediment with a higher rate of soil erosion (Poesen, 2018; Walling, 2006). The increasing sediment load creates a multi-dimensional effect in the river systems (Hatono and Yoshimura, 2020). The engineering works related to TIs like roadcut and fills generate maximum sediment from hillslopes by increasing gradient (Figure 10a), removing the surface vegetation, intensifying the effect of raindrops on the bare surface, reducing cohesiveness and shear strength of soil (Jankauskas et al., 2008; Seutloali and Beckedahl, 2015; Sheridan et al., 2008; Sidle and Ziegler, 2012; Sidle et al., 2011). The construction of unpaved roads within the flood-prone area also acts as a source point of sediment to the rivers (Figure 10b). Roadway also makes disturbances in the process of coupling between hillslope and channel system and forced to deposit huge sediments in the headwater basin (Katz et al., 2014; Sidle and Ziegler, 2012), which might introduce significant environmental consequences for the rest of the basin. The higher slope value around the road is transforming the rainfall into maximum surface runoff and triggering the rate of soil erosion (Arnáez et al., 2004; Cerdà, 2007). The effects of slope factor (gradient, length and aspect), rainfall erosivity and ground cover density on roadcut induce soil erosion has been thoroughly evaluated by Megahan et al. (2001) in Idaho, USA and a 6-year long investigation shows due to the construction of low standard logging road in this zone the downstream sediment yields have been increasing by 45-times (from 8.8 to 396 metric tons/km2/year) (Megahan and Kidd, 1973). Arnáez et al. (2004) has recorded about 16 times higher rate of soil erosion on roadcut embankment surface than a filled slope in the Iberian Range, Spain. Jordán and Martínez-Zavala (2008) have also estimated a total soil loss of 106 g/m2 from roadcut and 17 g/m2 from side fills in southern Spain. The excavated materials from the tunnels in the hilly region are often stored in the surrounding hillside and acting as a major point of sediment source to the allied river valley as observed in the Darjeeling Himalaya of Eastern India (Figure 10c–e). Tanyaş et al. (2022) have also revealed that the production of sediment through the mass movement for road work is nearly equal to the possible effect of a medium magnitude earthquake (Ms > 6.0).
Input of sediment from the riverside road cutting (a) and by flooding on the riverside unpaved road (b); (c–e) An ongoing Indian project for railway tunnel construction in the Darjeeling Himalaya generates huge debris through the excavation which has been dumped within the floodplain and channel of Teesta River and turned into a major source of sediment to the corresponding river system.
Excessive sedimentation in the forest rivers of the western United States has been observed after the development of numerous roads for logging, natural resource management and recreation (Luce and Black, 1999; Madej, 2001). Logging roads in forests are mainly unpaved and generate huge amounts of mobile fine sediments from loose surface materials, which increase the suspended sediment load of nearby streams or rivers (Anderson and Macdonald, 1998; Thomaz et al., 2014). From the surface of unpaved roads, large amounts of sand and gravel are also transported to the river system via rills and gullies (Katz et al., 2014). In this process of sediment generation, stream crossings are high-risk sites for increasing sediment delivery (Harris et al., 2008). For example, Thomaz et al. (2014) have estimated that the value of suspended sediment concentration was 3.5 and 10 times higher in the downstream of stream crossing than upstream, although, the effect is scale-dependent as the result is insignificant in the higher-order streams. A study in California by Madej (2001) shows the erosion control treatment (excavating culvert and road fill, reshaping stream bank) to the forest roads may reduce only <2% of sediment delivery to the rivers in comparison with an untreated road. The rate of sediment production from an untreated logging road is about 1500–4700 m3/km.
Initiation of gully
Formation of rills (depth < 0.25 m) and gullies (depth > 0.25 m) (Poesen, 2018) by the concentrated runoff (after crossing the limit of soil specific critical threshold) is a profound geomorphological process for the production of sediments and landscape lowering and/or land degradation worldwide (Valentin et al., 2005). The concentration of flow is a function of the wide variety of natural and anthropogenic characteristics (Dube et al., 2020). Among the anthropogenic factors, TIs like paved and unpaved road surface, RSC, road-side artificial drains or ditches and their outlets induced soil erosion is a trending concern for road engineers and land managers from the perspective of environmental impact assessment (EIA) (Jungerius et al., 2002; Nyssen et al., 2002; Standish and van Zyl, 2007). Research on the expansion of rills and gullies around the TIs is an emerging field of investigation worldwide, whereas, the problem becomes severe for the developing countries due to lack of maintenance and provision for safe outlets for excess runoff (Adams and Watson, 2003). Montgomery (1994) initially recognised that runoff from the road surface is reducing the required drainage area to cross the threshold limit for channel initiation and destabilised the slope (Figure 11a). In addition, Figure 11b shows that the artificial drains alongside the roads and their outlet (mostly by pipe) have also provoked gully development in the downslope by increasing the drainage area and associated additional waters to the water outlet (Montgomery, 1994). Mati (1984) has estimated that about 50% of gullies in the Kiambu District, Kenya was formed by road drainage. Wemple et al. (1996) monitored the formation of gullies below the stream crossing at the steep slope (>40%) region with long ditches around the roads. Nyssen et al. (2002) noticed that road building has disturbed the surface equilibrium and lowered the topographical threshold values to form gully in the Ethiopian Highland region. Katz et al. (2014) also estimated that road surface affiliated watersheds require lower critical slope value (Scr = 0.21 A−0.45) to form gully than watersheds without roadways over the Pike-San Isabel National Forest region (USA). The position of roads on the landscape plays a crucial role in the possibility of gully formation because when the road crosses along the hillslope and pediments developed more gullies than roads along the ridge top and toe slopes (Nyssen et al., 2002). Sometimes the abandoned airfield pavement also becomes a cause of gully formation in the adjacent landscape by concentrating rainwater on the impervious surface (Figure 11c and d).
(a) Formation of the gully at the roadside by the concentrated flow of road-surface runoff, (b) development of gully at the outlet of a road side ditch, and (c and d) impervious surface of an abandoned airfield/runway generates large amount of runoff and concentrated flow of such water has developed rills and gully at the Surichua Air Base of India.
Ways for sustainable development of TI
Sustainable geoengineering practices are essential to reduce negative interaction between transportation systems and hillslope geomorphology. For sustainable slope management in the road corridors, the Department for International Development (DFID) of the UK has been experimented with low-cost engineering plans on several sites (Hunt et al., 2008). The combination of bioengineering techniques such as linear grass-planting, direct seeding, brush layering, fascines, palisades, truncheon cuttings, live-check dams, tree-planting, wattle fences and small-scale low-cost engineering structures like slope-trimming, masonry walls, surface drains, trench drains, dry stone wall, revetment are performed effectively in stabilising the hillslope as illustrated by different case studies in Hearn et al. (2021). The effect of stormwater events on roadside soil erosion and slope failure can also resist by applying effective rainwater harvesting methods on road surfaces like ‘low-impact development’ (LID) (Prince George’s County, 1999) and ‘Green Road’ (Lin et al., 2018). The prescribed methods in LID help to use rainwater for irrigation on roadside vegetation and reducing runoff up to 43.5%–54.5% annually by enhancing infiltration facilities to recharge the aquifer and mitigate the risk of flooding (Lin et al., 2018).
Avoiding the alignments of transport networks within the ‘river corridor’ to sustain lateral connectivity between channels and floodplains might be a sustainable practice of TI development. The Vermont Agency of Natural Resources (VANR) (2008) presents a best process for river corridor delineation instead of simple buffering of river lines. Such practice may also help to avoid infrastructural losses due to flooding and bank failure. For example, $46 trillion of global exposure of TIs to the river and coastal flooding could be turned into $158 trillion by 2050, due to regular and unplanned socio-economic development within the flood-prone areas (Jongman et al., 2012). The installed culverts are often undersized and create problems in the river’s longitudinal connectivity (Roy and Sahu, 2018). Therefore, the standard protocols, like consideration of flood flow at 100-year return period, should be followed in culvert and bridge designs even in headwater streams as provided by Vermont Department of Transportation (2001), Indian Roads Congress (IRC) (2004), National Bridge Inspection Standards (NBIS) (2006) and Federal Highway Administration (FHWA) (2012). In this context, ‘deculverting’ or ‘daylighting’ (Wild et al., 2011) is now a successful concept in river restoration programmes worldwide. It allows rivers for a healthy ecosystem, reducing flood risk and giving recreational benefits.
Future research prospect on the interaction of TI and geomorphology
Collapsing roads by frequent landslides around the world, especially during the rainy season of hills, becoming a serious threat to human life and infrastructures settled over the mountains. The team of NASA on Global Climate Change has predicted that the changing climate could trigger more landslides in the mountain regions of Asia (Merzdoft, 2020). Quantitative findings are claimed that climate change induces to happen more catastrophic rainfall events frequently over such high mountain regions, which are the prime cause of such hazards and other cascading events like dam failure and devastative downstream floods (Liu et al., 2021; Merzdoft, 2020). Consequently, effective research plans incorporating the effect of climate change are essential to building new sustainable road networks around the hilly region of the world with special attention on the drainage system and altering surface runoff on the unstable hillslopes. Changing rainfall pattern on the plainland also creates problems for existing designs of the water crossings and other TIs by altering the hydrology and associated floodplain geomorphology (Kalantari et al., 2014, 2017; Michielsen et al., 2016). Currently, the seasonal storms are making the transportation systems more vulnerable than ever to floods and erosion across the globe. Such events are a wake-up call for the scientific community and planners to investigate on increasing resilience of TIs to the upcoming threat of weather extremes since the crossing structure should design to last at least more than 50 years. In this regard, region-specific digital inventory of bridge and other TIs are required to prepare for a proper exploration of causes behind the collapse of any type of TIs, as similar database generated by the Department of Homeland Security of the USA (National Bridge Inventory (NBI) Bridges - CKAN (data.gov)). For example, using such database Cook and Barr (2017) finds that USA has experienced about 60% of bridge failure due to natural disasters, mainly through flood events (52%), which is helpful for upcoming plans of bridge construction.
The quantitative assessment of longitudinal disconnectivity by TIs has been studied well by Gregory and Brookes (1983), Douglas (1985), Castro (2003), Merrill (2005), Harris et al. (2008), Thomaz et al. (2014), Roy and Sahu (2018), however, the effect on the lateral disconnectivity, especially the alteration of sediment, water, nutrient fluxes, is still required more intensive studies for sustainable floodplain management. While the existing studies on lateral disconnectivity are only dealing with quantification of the amount of floodplain disconnected and spatial variation on the potential interaction between streams and transport networks (Blanton and Marcus, 2014, 2009; Kumar et al., 2014; Roy, 2022). Since a significant amount of the world’s transport networks are highly vulnerable to geohazards (Koks et al., 2019), assessment for suitable plans to reduce such vulnerability are also essential under the threat of climatic instability.
Region-specific research could also help to understand the critical interaction between TS and earth surface processes as it varies with morphogenetic regions. For example, hilly regions are more prone to slope instability due to road construction in comparison with plains. Similarly, the problem of lateral (dis)connectivity is more prominent in the plainland rivers than in highland rivers. Nowadays facilities of transport infrastructure in the arid and semi-arid regions are changing very first with the growing economy, which might have a significant effect on arid geomorphology. The construction works for TIs in particular on the dry surface are also generating a large amount of fugitive dust and deteriorating the air quality. For example, as per Lehr and Lehr (2000) in 1995 about 28% and 15% of fugitive dust were originated from unpaved and paved roads, respectively. From an environmental perspective, this is an emerging field of investigation.
Considerable researches have been also focused on the effect of ‘forest road’ on the hydro-geomorphological alteration, however, in those studies, the effect of deforestation and/or clear-cut logging has been merged with the impact of road construction and related infrastructure. Therefore, specific research on the effect of TIs within the forest geomorphology could be more informative for land managers. The application of advanced remote sensing for example, UAV, LiDAR, InSAR and ArcSAR Interferometry, Gigapixel imaging and Acoustic sensing including the platform of machine learning and deep learning are showing their enormous potentiality in landscape study and monitoring the processes on earth surface very precisely (Romeo et al., 2021). Such technology might be effective in the present context of research also.
Conclusion
Although having a history of about 2500 years of modern road construction, however, the initial recognition of the impact of TI construction on geomorphology comes from the engineers during river engineering in the 1970s. Since 2000, technological advancement has intensified this field of research, where researchers have started to monitor the impact of trail to huge tunnel development on the alteration of earth surface processes and landforms with a leading investigation on the hillslope and fluvial geomorphology. The multidimensional effects of TIs on hillslope and fluvial geomorphology have been encountered by different studies across the world. Therefore, ‘transport geomorphology’ could be a component of anthropogeomorphology, which can feed the argument for the Anthropocene epoch. Primarily the alteration of surface hydrology by increasing runoff from reduced infiltration has exaggerated the problem of soil erosion by rills and gully formation after reducing the critical threshold limit and slope stability. As a result, the development of TIs over the mountain topography is a significant cause of different geomorphic problems like frequent landslide and mass movement around the cut-and-fill slope, gully formation at ditch outlets. Therefore, within a kilometre of any road network about 55% of global landslides were happened as per landslide inventory map, while after considering the roads up to SH level at regional-scale which is turned into ~75% in India. The construction of TIs and corresponding earthwork are also a major source of sediment input to the world river systems and altering the channel geomorphology and ecology as well. Landscapes especially the river systems of plainland regions are facing problems of lateral and longitudinal disconnectivity due to the rapid expansion of transport networks, as road and railway embankments and undersized causeways are playing the role of artificial barriers for free movement of water, sediment and nutrients. Although having the needs of TIs development, the sustainable construction practices like bio-engineering and low-cost engineering for roadside slope management, deculverting, consideration of 100-year flood return for causeway construction, mapping of river corridors, road water harvesting are now essential to follow for less effect on geomorphology and regional sustainability.
Supplemental Material
sj-xlsx-1-anr-10.1177_20530196221128371 – Supplemental material for Role of transportation infrastructures on the alteration of hillslope and fluvial geomorphology
Supplemental material, sj-xlsx-1-anr-10.1177_20530196221128371 for Role of transportation infrastructures on the alteration of hillslope and fluvial geomorphology by Suvendu Roy in The Anthropocene Review
Footnotes
Acknowledgements
I would like to thank Dr. Costanza, Editor, The Anthropocene Review, and two anonymous reviewers for detailed review and construction comments to improve the manuscripts.
Data availability statement
The data use for present work is freely available in the public domain and sources are also properly mentioned in the text. I have also enclosed the detail review work in supplementary file.
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
