Hungry water

Civilisation exists by geological consent, subject to change without notice.  – Will Durant

Rivers are the most commonly accessible of natural water resources. But rivers do not flow everywhere, nor do all rivers carry equal amounts of water. Each river is an individual entity – not living, but certainly dynamic. A river, like any other natural feature, has shape, size, context, as well as a history and process. The combination of these elements is distinctive, making each river unique. It is also dynamic, continuously evolving, and no river's behaviour can be reflected in simple deterministic models. Likewise, any operation on the river – removing water, impeding the flow – will produce effects that do not follow simple arithmetical rules. These changes induce change in the river's process, and therefore have cascading consequences. In other words, humans need to take the river itself into consideration as a participant while intervening in its process.

Let us begin with some potentially oft-overlooked basics. A river begins when water drops on high ground, and the natural level of rest for all water on the earth's surface is the sea. Water arrives at high ground through precipitation in the form of rain, or as snow, which melts, and this transportation process is powered by solar energy. At the high point of its trajectory, this water has a large amount of potential energy, which is proportional to its height above sea level, and it is this energy that ultimately drives the water downhill until it reaches the sea. This entire flow mechanism is essentially a process in which the potential energy is expended.

Rivers typically have three stages. The first takes place in the higher reaches, where the slopes are steep and 'fresh' rocks are eroded by the river. Second, as the water flow through the plains, some silt is deposited and some is picked up, the net effect of which is sediment transport. Third, and finally, as a river approaches its lowest point, where energy drops to near zero, this silt is again deposited. In the case of Himalayan rivers, these transitions are particularly pronounced, due to the extremely high levels of relief that are part of this process. Most Himalayan streams begin from glaciers or glacial lakes at very high altitudes, up to 6000 metres above sea level. From there, they follow tortuous courses, along and across the long grain of structural alignments, until they emerge onto the plains.

The plains area is dominated by 'trunk' streams – those large flows into which tributaries empty their water and sediment, mainly the Sindhu (Indus), Ganga and Brahmaputra (Jamuna). The basin drained by the Sindhu system emerges from the mountains and flows directly to the sea. The Ganga and Brahmaputra both flow for considerable distances parallel to the mountain, in troughs between the Himalaya and the uplands of the Indian peninsula. The Ganga in particular is closely bounded by rocky high ground to the north and south, and its geometry within these limits is very interesting, as we will see.

The constant drift
Any river's characteristics are ultimately a consequence of its tectonic frame, which determines the behaviour of both the trunk and the tributaries, as well as determining erosion patterns. The great peninsular wedge of the Subcontinent, pushing northward under the Tibetan landmass, forces the uplift of the Himalayan ranges. It is this segment that is exposed as the central Indian plateau, which also passes under the alluvial plains of the Ganga. Tectonically, this basin of the Ganga is a 'foredeep', or sediment-filled, depression at the front of the rising Himalaya. The 'Ganga Maidaan' we see today is made entirely of sediment brought down by the tributary rivers. These plains are composed of a complex mosaic of floodplains, recording the history of the Ganga as it shifted and migrated.

At any point on its course, water naturally flows down the line of maximum gradient. The total energy content is thus measured as the momentum plus the residual potential energy. Momentum is the product of mass and velocity, which gives some very interesting consequences. Running water erodes its bed, and picks up grains of sand and silt, which in turn increases its mass and thus reduces its speed. As the stream slows down, it cannot continue to bear the load it had been carrying, and some of this drops down as sediment. The ability of a river to carry sediment depends on both the velocity (a function of the gradient) and the volume of its water flow.

Two measures are commonly used to measure velocity: competence, which reflects the largest size of particle that can be moved; and capacity, which indicates the total quantity that can be carried. One significant aspect in this regard is the fact that as a river moves towards its lower reaches, its water picks up more and more sediment; but at the same time, its velocity is reduced as the slope down which it is running becomes gentler. Although the slope reduces towards the downstream, this does not necessarily happen smoothly, as abrupt variations come about due to bedrock characteristics and structure. Sudden reductions in gradient result in the sediment load being dumped at one spot, where the ebbing current will subsequently remove the finer particles. In the rivers of the Himalaya, this process is seen in the formation of 'fans', where the rivers come out of the hilly tracts. Since the deposits in these areas are dominated by coarser sand and gravel, water tends to seep in, and thus much of the flow is underground – a characteristic feature of the so-called Bhabar zone, the plains immediately south of the lower Himalaya.

Throughout this process, there is a constantly changing balance of velocity and suspended sediment load. This scenario is complicated by additional influences, including gravity, friction, and the addition or subtraction of water. The process of balancing of energy, discharge, sediment load and gradient operates all along the watercourse, the effects of which are erosion and deposition in a complex pattern. As noted previously, in the lower reaches the velocities are slower, the suspended silt is finer in grain, and the river follows a convoluted, meandering course. This course, one needs to remember, is an ever-shifting line, but the degree to which this line can change is more extreme in the downstream reaches. Together, these processes, while integral and inseparable components of river behaviour, cause many of our problems – eroding banks, flooded swamps and the deposition of silt in the plains. Conversely, however, they have also produced the picturesque valleys of the Himalaya, the great fertile alluvial plains and the awe-inspiring deltas, among other things. Floods created the plains, and the floods maintain them.

Unpredictable avulsion
The principal factor that causes this flooding is tectonic uplift – a factor that is, of course, continuing. In fact, new research suggests that the high erosion rates caused by heavy monsoon rains remove weight, thus accelerating this uplift. Such processes do not occur as one single block, and the old structural breaks and faults lead to uneven rise, irregular changes of slope and thus unpredictable behaviour of the stream courses. The implications of this should frighten those who dream of 'taming' any river.

Indeed, the effects of measures already taken in this regard have exacerbated the problem. Such approaches were followed in flagrant contempt of traditional knowledge, which respected the power of the river and adjusted to it. In 2008, the annual flood led to avulsion, the sudden changing in the course of a stream, as the Kosi riverbed had been raised to an unsustainable height (see Himal Oct-Nov 2008, "A river at disequilibrium"). More such measures will provoke more such catastrophes. For instance, the potential construction of a high dam upstream in Nepal, as is currently being discussed between India and Nepal, would impound the silt, leaving only clear water to flow. But there is a saying that holds that "Clean water is hungry water," underlining the fact that any river needs silt to balance its available energy. Since the bed is armoured by coarser grains, the downstream energy of a dammed river would be spent in eroding laterally, and woe betide those who settle in such areas in the assurance that government officials have truly tamed the river.

Modifying the behaviour or the courses of rivers is essentially geomorphic engineering. This naturally involves all of the factors of geomorphology, which comprises the ensemble of processes and objects (for want of a better word) in a condition of dynamic equilibrium, a state that is constantly changing. Methodological tools for proper analysis and modelling of river morphology are not available yet, and the limits of many of the new mathematical branches are being extended every day by these challenges. The system takes its own time to settle into a new, also dynamic, equilibrium. This time is on a scale much longer than that of executive decisions and plans. Therefore the potential of the changing river system settling into its new flow has been neglected or conveniently ignored as not significant.

Even with our incomplete knowledge, it is possible to foresee the trend of changes. Suppose a dam is built. It will impound water, which will flow as it is released, or when the reservoir is full. It will also impound the sediment, which will not move further. The water downstream will be clear, and the energy of the gradient will make it pick up sediment. The coarse sand and pebbles of seasonal flow will not be moved, and the stream will have to lengthen its course to use the energy. (A longer course for the same drop in height means a lower gradient.) Since the surrounding areas are higher or made of coarser sediment, another abrupt change of course is inevitable, with all the attendant misery for the hapless populace in the new path.

The situation in the tract where the Kosi emerges from the hill ranges is almost like the head of a delta. The trunk stream had so much less energy that very little of the sediment could be removed, and there is a large, wedge-shaped tract of land that carries the scars of past courses. Each of these courses operated as long as it fit the energy-water combination. When it could erode no more, or deposit no more, a neighbouring low tract was adopted. Rivers have no freedom of will; they do not flow or change by caprice. It necessarily follows that the course just abandoned is out of balance with the present stream. So, when we hear that the Kosi has been 'restored to its original course', we have to wonder: Which is the original, and why did the river leave that course at all? Since there are no ready answers, it is now only a case of wait and hope.