flood analysis

GIS for River Floods

By: S M Ramasamy, The author is a Distinguished Professor, Dept. of Geology, Alagappa University, Karaikudi
Flood is a complex outcome of an interface between geological, physiographical, hydrological, meteorological and anthropogenic dynamics. It mandates high computational techniques for the analysis of a large volume of hierarchical pictorial data. GIS technology, inseparably interleaved with remote sensing, helps to panoramically beam complex issues involving river floods and the feasibility of detecting and mitigating them.
Magazine Articles Planning n Mitigation

Floods have now started occurring everywhere including hills, plains and river valleys, urban agglomerations and even the deserts of Rajasthan. Barely could the subcontinent overcome the shock and grief of the Uttarakhand floods (2013) when a massive flood inundated a large part of the Chennai metropolis in 2015 killing 470 people and affecting the lives of more than 3.3 million families (PTI, 2016). The latest disaster occurred in Kerala during August 2018 and ravaged the State with a loss of reportedly 3 billion USD (LiveMint, 2018). It goes without saying that floods have emerged as a major concern in developing countries. Flood is a consequence of several interlinked and  multivariate systems—geological, hydrological, physiographical, meteorological and anthropogenic—which need to be understood in order to predict and mitigate it. Flood analysis demands high-powered technologies and scientific acumen.

In this context, GIS technology, comprising of ground based and airborne/satellite borne remote sensing data provides major inputs. The digital computing of map data (points, lines and polygons) on multiple themes and huge volumes has emerged as an unparalleled tool in the mapping, monitoring, modelling and mitigation of natural disasters. The article takes a panoramic view of river floods and the advantages of GIS in tackling them. Once planners and managers see value in it, plans on optimal flood disaster protection and preventive development can be formulated and executed. This has been enumerated in this article
taking certain type cases from India and the adjoining countries.

Flood Analysis | Life History of Rivers

Like humans, rivers too have a life history with a youthful stage in the high order mountains that form the catchment, a mature stage in the plains  and an old stage in the coastal regions as seen in Figure 1 (Thornbury, 1985). The well-defined life history of rivers is dominantly controlled by the isostatic and eustatic phenomena of the Earth. While isostasy deals with the elevation of the path of the rivers, the eustatic phenomenon is related to the mean sea level (MSL) or the base level of erosion. When rivers are born in the mountains—much above the MSL, they have a lot of energy. They incise the mountains vertically and reach the foothills as soon as possible as their aim is to attain MSL. Since they have high energy like youngsters, the stage is designated as such (1, Fig. 1). Having reached the plains, the rivers ply only a few hundred meters above MSL—with comparatively less energy, moving in sinuous paths to reach the coasts. This is designated as the mature stage or second stage as shown in Figure 1. Once the rivers reach the coasts, they attain the base level of erosion/MSL and do not have any energy to travel further, nor carry the sediments they brought all along. The rivers then surrender to the waves and tides of the ocean, developing deltas—designating it into an old stage (3, Fig.1). An understanding of a river’s life history thus forms the basis of geotectonic and anthropogenic aggressions that add to the flood phenomena.

Flood in Catchment

Owing to the youthful stage of rivers, hilly regions are inherently the kitchens for floods. But, wherever there are topographic features with semicircular escarpments (4, Fig. 2), streams witness massive headward erosion due to the steep slopes causing heavy floods (5, Fig. 2) in the foothills and the adjacent plains. The GIS based analysis of the IRS LISS III FCC imagery of the Kozhikode-Palakkad region of Kerala shows that the massive floods in Kerala in August 2018 were due to this phenomenon, besides others.

Kosi Floods

The recurring flood in the Kosi River has also been worrisome. The study of Kosi region shows branches of the old course of the Kosi in its eastern part (6, Fig. 3) whereas currently the river (7, Fig. 3) is flowing with westerly convexity. This indicates that the Kosi has gradually shifted towards a westerly and southwesterly direction, dumping huge quantities of sediments brought down from the Himalayas at the foothills, forming the ‘Kosi Cone’. Multi dated topographic sheets, satellite data and integrated GIS analyses show that the Kosi cone is spreading, indicating the phenomena of sliding. The post collision tectonic model (Ramasamy, 2006) shows that the northerly-directed compressive force which had originally shifted the Indian Plate towards the north, is still active, but since the Himalaya is obstructing from the north, the compressive force is partly uplifting the Himalayan mountains and partly deforming the Indian Plate. Because of such an uplift and the relative subsidence of the southern plains, the Kosi cone is gradually sliding. This is causing westerly shift of the river (6 to 7, Fig. 3). Due to such a migration, the old courses, (6, Fig. 3) the present course (7, Fig. 3) and the western parts are often flooded. As the Kosi cone is in a continuous process of sliding, the shifting of the river is also continuous, flooding newer areas in the western part of the present course (7, Fig. 3). In fact, the Kosi cone is a typical example of all Himalayan rivers, as at the point of debouch, most rivers behave in a similar way. Thus all outlet points of the Himalayan rivers demand detailed studies of remote sensing and GIS.

Tectonic Subsidence and Floods in the Central Parts of Kerala

Kerala is prone to seismicity, especially in the central parts of the State (Rajendran et al.,2009). During the Thohuku earthquake, 2011, off the Pacific coast of northern Japan, major land subsidence occurred (Imkiire and Koarai, 2012). Similar observations were made in other parts of the world as well. Kerala too exhibits a similar possibility where clusters of maximum seismicity may have caused the central part to subside. Further, while the studies of Ramasamy (2006) indicate tectonic activity along the lineaments of Kerala, observations by Singh and Raghavan (1989) and Raj et al. (2001) show more episodes of seismicity along north and northwest (NNW), south and southeast (SSE) faults. Since the central part of Kerala is subject to land subsidence as the GIS based spatial analysis carried out between isoseismal maxima and flood inundation shows, it would attract more floodwaters as revealed in the August 2018 floods (9, Fig. 4).

Flood Analysis | Tectonic Disturbance to rivers and floods in Trichy

Selvakumar and Ramasamy (2014) have inferred that tectonic features significantly control floods in different parts of Tamil Nadu. A GIS study of urban Trichy and adjacent areas of the central Tamil Nadu shows the influence of tectonic subsidence as well as the triggered aberration between tributaries and the main river. There are two major sub parallel faults in the Trichy region, which extends from Puducherry in the northeast to Kambam valley in the southwest (10, Fig. 5), along which continued land subsidence occurs (Ramasamy and Karthikeyan, 1998). It is because of this that river Cauvery flows sluggishly within the fault-bounded Trichy region. Therefore the Arayar tributary system in the north (11, Fig. 5) and the Koraiyar tributary in the southwest (12, Fig. 5) are not able to deliver their floodwater load into the Cauvery, especially when the river is in full spate. In other words, Cauvery finds it difficult to carry its own floodwaters within the fault-bounded Trichy block,rendering it impossible to accommodate the floodwaters of its tributaries, which consequently leads to the inundation of urban Trichy and adjacent regions (13,  Fig. 5).

Floods in Brahmaputra River

The Brahmaputra is infamous for its annual floods, where a huge quantum of water recurrently debouches onto the plains through a narrow passage in the Himalayas. A GIS based integrated study of flood plains, old courses and the present drainages show that there are older flood plains (14, Fig. 6), younger flood plains marked with the noodles of old courses (15, Fig. 6) and the present river. The study reveals that the river is migrating towards the north. It also shows that under the present conditions of its shifting towards north, the northern bank is more prone to flooding. However, in the event that the northward movement is restricted, the river may flood the older courses. This mandates a holistic remote sensing and GIS based study of the tectonics, migratory history, meteorology, flood dynamics and geochronology of the old courses of the Brahmaputra.

Buried River Courses and the Chennai Flood of 2015

The GIS study of the region shows there is a major river system buried in the west, northwest and north of Chennai, (16, Fig. 7) with a large number of old rivers and streamlets embedded within. Ramasamy et al. (1992) infer that possibly the Cauvery flowed through the Chennai region about 3,000 years ago and left behind traces. The recent GIS study carried out involving a hierarchy of geo-anthropogenic parameters shows that the buried riverlets found within the major buried river system (16, Fig. 7) have acted as conduits for the flood 2015 in the Chennai region (Ramasamy et al., 2018). Once this buried river system was flooded, the inundation reached the other parts of the Chennai city. Such palaeo and present river dynamics must be studied in the context of anthropogenic aggressions, where the use of GIS can present exhaustive vistas.

Compressed Meanders of the Cooum River and the Chennai Flood of 2015

Drainage morphometry, in other words the architecture of the drainages, plays a vital role in river floods. For example, rivers that flow along straight courses will not cause any floods unless otherwise obstructed. Annularly flowing rivers will cause floods along their convex banks whereas when the excess water flows along the compressed meanders that have more sinuosity, it will flood either sides of the rivers. This is because compressed meanders may not be able to hold the high quantum and velocity of floodwaters. The GIS analysis of the compressed meandering pattern derived from IRS LISS III FCC data of the Cooum River in the Chennai region and its integration with 2015 flood data shows that the river has flooded the Egmore region extensively due to such compressed meanders (17, Fig. 8). Settlements thus should not be developed at the convex banks of the annular drainages and in the compressive meander zones. In addition, the existing settlements can be saved by laying straight canals linking the compressed noses of these river segments.

Eyed Drainages and the Cauvery Flood: Mysore

At times, drainages that flow as a single stream may split up into two or three, especially at their intersection points where faults occur. After crossing the faults they will again rejoin to flow as single stream. This has been called as ‘drainage anastomosis’ by Smith et al. (1997). The same has been mapped as eyed drainages in different parts of south India by Ramasamy et al. (2011) who inferred that the eyed drainages indicate tectonic subsidence at the intersection of faults and drainages. In this context the drainages, lineaments and flood plains interpreted from LANDSAT FCC and also IRS LISS III FCC data were integrated using ArcGIS for the Mysore region. It shows that between NNE-SSW faults (18, Fig. 9) and the eyed drainages (19,  Fig. 9), the land is subsiding due to which the Cauvery recurrently floods the region as marked by the colour red (20, Fig. 9) in the satellite data.

Main Channel-Tributary Dynamics and Bharatpur Floods

In mapping floods the interface dynamics between the main river channel and the tributaries is essential. A study of the Bharatpur-Agra region shows that the Yamuna has bundles of old courses (21, Fig. 10) to the west of its present course (22, Fig. 10). Similarly, the river Banganga has old courses (23, Fig. 10) to the north of its present course (24,  Fig. 10) in the south. The integration of the data over a common GIS overlay reveals that the Yamuna migrated about 150-200 km towards the east and its tributary the Banganaga migrated about 40-50 km towards the south. This shows that when the main river, the Yamuna, was flowing in the western region in the area north of Bharatpur, the Banganga used to meet it there. But when the Yamuna shifted eastwards, its tributary the Banganga tried to catch it by taking a number of courses (between 23 and 24, Fig. 10) but could not achieve it and ultimately buried itself near Bharatpur. This is the reason for the waterlogging and marshiness in the Bharatpur region.

Flood Analysis | Sea Level and Floods

As discussed earlier, when rivers reach coastal zones, they attain MSL, depositing a delta to mark the end of their journey. At times, however, land may show elevations lower than the MSL where rivers stagnate. Such topographic conditions prevail in the Dongting region, China, where vast areas of several thousand square km has elevations lower than MSL. So the Xiang, the Zi and the Yi Yang rivers stagnate on reaching the Dongting region, causing huge floods and marshiness in the area (25, Fig. 11). Now, China has made artificial pathways to drain the flood out. In river flood studies, analysing topographic elevations along the path and profile of the rivers are also essential. A GIS analysis based on DEM profiling forms an excellent tool.

Floods in the Indus Delta: Pakistan

When rivers reach MSL they get a landward thrust from waves, tides and creeks which may cause drainage congestion and flooding. The LANDSAT data of Indus delta of Karachi region is one such example (26, Fig. 12). In coastal areas flood dynamics need to be studied involving oceanographic parameters such as waves, tides, creeks, storm surge; riverine parameters and tectonic parameters and more, for which GIS is a credible tool.


The river flood is a complex issue involving a hierarchy of geo, hydro, meteo and anthropogenic parameters. Only technologies that involve multi-spatial parametrics such as GIS and remote sensing, can provide credible solutions to flood related issues.


Raj K. G., M. A. Paul , V. S. Hegde and R. Nijagunappa, 2001. Lineaments and Seismicity of Kerala: A Remote Sensing Based Analysis, Journal of the Indian Society of Remote Sensing 29 (4):203-211.

LiveMint, 2018. Bodies found as floods recede in Kerala, death toll rises to 400, LiveMint, August 20.

Press Trust of India (PTI), 2016. Northeast monsoon claimed 470 lives in Tamil Nadu: Jayalalithaa, The Hindu BusinessLine, January 4.

Rajendran C. P., B.John, K. Sreekumari, KRajendran, 2009. Reassessing the Earthquake Hazard in Kerala Based on the Historical and Current Seismicity, Journal of the Geological Society of India, 73 (6): 785-802.

Ramasamy S. M., A. Vijay and S. Dhinesh, 2018. Geo-anthropogenic aberrations and Chennai floods:2015, India,Natural Hazards: Journal of the International Society for the Prevention and Mitigation of Natural Hazards 92 (1): 443-477.

Ramasamy S. M., C. J. Kumanan, R. Selvakumar and J. Saravanavel, 2011. Remote Sensing revealed drainage anomalies and related tectonics of South India, Tectonophysics 501 (1): 41–51.

Ramasamy S. M., 2006. Remote Sensing and active tectonics of South India, International Journal of Remote Sensing, 27 (20):4397–4431.

Ramasamy S. M.and N. Karthikeyan, 1998.Pleistocene/Holocene graben along Pondicherry–Cumbum Valley, Tamil Nadu, India, Geocarto International 13 (3): 83–90.

RamasamyS. M., V. Venkatasubramanian, S. R. Abdullah and S. Balaji, 1992.The phenomenon of river migration in Northern Tamil Nadu – evidence from satellite data, archaeology and Tamil literature, Man and Environment 17 (1): 13–25.

Selvakumar R. and S. M. Ramasamy,2014. Evaluating influence of active tectonics on spatial distribution pattern of floods along eastern Tamil Nadu, India, Geomorphology 226: 25-34.

Singh H.N. and V. Raghavan, 1989. A note on earth tremor of September 2, 1988 in Trivandrum district, Kerala, Journal of the Geological Society of India, 34 (4): 323-325.

Smith N.D., T. S. McCarthy, W. N. Ellery, C. L. Merry, H. Rüther, 1997.  Avulsion and anastomosis in the panhandle region of the Okavango Fan, Geomorphology  20 (1–2): 49–65.

Imakiire, T. and K. Mamoru, 2012. Wide-area land subsidence caused by “the 2011, Off the Pacific Coast of Tohoku Earthquake,”Soils and Foundations 52, (5): 842-855.

Leave a Reply

Your email address will not be published. Required fields are marked *