Pakistan benefits from an extraordinary water supply, sourced mainly from swift-flowing glacial melt from the Himalayas in late spring, and monsoon activity between June and October. To take advantage of this tremendous resource, the country has been highly engineered in hydrological terms: irrigated areas represent 82% of all farmland and 43% of the 170 million-strong population is directly dependent on farming activities. However, irrigated areas are exposed to flooding hazards, and consequently the largest sector of the economy and the majority of Pakistanis are vulnerable.
Additionally, many villages are situated on river terraces, or in low-lands, and urban migrants tend to informally settle in low-lying high risk areas. As the great flood of 2010 has illustrated in vivid detail, floods are typical in the five major rivers of the Indus River Basin. Twenty major floods, and many more minor floods, occurred in the 50 years from independence in 1947 to 1997. Thus Pakistan is exposed, susceptible and sensitive to regularly occurring flooding events which at times are exacerbated by the river engineering necessary to maintain the irrigation infrastructure.
The development of Pakistan’s flood-management system can be characterised by two dominating approaches and two corresponding periods: 1947 to 1973, a period of risk acceptance and limited risk management; and 1973 to the present, a period of comprehensive physical risk management. Although flood-irrigation techniques – where water is distributed across the soil by gravity – had dominated farming along the Indus River since pre-historic times, the original canal network, upon which the current system is based, was conceived and executed under British colonial rule, beginning with the Upper Bari Doab Canal in 1859. Throughout the colonial era, the system was maintained and expanded, such that, on the eve of independence, there were 150 major canals extending thousands of kilometres through the country.
The colonial approach to flood management depended on a network of “bunds” (linear levees along rivers and ring levees around cities), which the army could strategically breach when waters approached flood stage. During periods of high water, barrages and cities with bunds were protected, but massive flooding would occur in breach areas and regions without protection. The general public had little influence on flood management, though public opinion in affected areas fell decidedly against risk acceptance. The bund system of flood management was carried forward after independence.
In 1960, the Indus Basin Development Programme (IBDP), a colossal engineering project signed into existence with the Indus Waters Treaty between India and Pakistan, further fashioned much of Pakistan’s countryside into an extensive network of canals and reservoirs. The focus of flood planning – shaped through the lens of the Indus Waters Treaty – was on drainage procedures to avoid damage to recently constructed critical infrastructure.
The IDBP was part of a wider trend in modern flood management, born out of the experiences of inundations that beset the Tennessee Valley and the Great Plains of the United States early in the twentieth century. Armed with the vanity of modern engineering techniques and the doctrine of economic growth, international financial institutions and donor countries began to promote and incentivise mega-projects, like the IDBP in Pakistan and the Helmand-Arghandab Valley Project in neighbouring Afghanistan, offering enormous loans to developing countries. This international one-size-fits-all engineering approach to hydrological mega-project spread to developing countries around the globe, in spite of important regional peculiarities.
These water projects, while credited for transforming developing countries into the world’s producers and exporters of commodities like wheat and cotton, are also widely criticised for their environmental impacts. Biodiversity plummets in the face of habitat destruction, soil erosion increases, grazing land disappears and water-borne disease proliferates. In addition, the changing nature of river aggradation and erosion processes can result in accentuated flood events. Some of these consequences in the case of the Indus were even recognised under the British Colonial administration – but were generally considered to be the price of development.
Questions also arise about the relevance of large-scale projects to goals of poverty reduction. Engineering projects can exclude and marginalise the vulnerable poor, whose livelihoods are already sensitive to shocks. So much of rural, subsistence agriculture in developing countries is based on flood recession irrigation. Famous examples from Africa, both the Kainji Dam in Nigeria and the dams on the Lower Omo River in Ethiopia, have resulted in massive disruptions to flood recession agriculture livelihoods, on which hundreds of thousands of vulnerable poor depend.
Moreover, developing countries like Pakistan, whose rural livelihood systems, infrastructure and economies are utterly transformed by these projects, suddenly become vulnerable not only to flooding events but also to fluctuations and shocks in international commodity markets. Market-led growth in the absence of social programmes has another consequence: growing disparity between the haves and have-nots, who incidentally became the most vulnerable to river flooding.
Upon completion of IBDP in 1970, Pakistan’s agricultural production expanded substantially. However, shortly thereafter, in 1973, when massive flooding generally overwhelmed the canal network, the risk-management paradigm shifted. Vulnerability of the system was revealed, as well as the resource and experiential constraints of regional flood managers in dealing with newly engineered canals and reservoirs.
In 1978, the Federal Flood Commission was established to implement a comprehensive risk-management strategy, the National Flood Protection Plan. The tool kit of the new strategy included greater resources for reservoir operations, including procedures, inspections and training; schedules for bund maintenance and reinforcement and bund breaching plans; expansion and modernisation of data-collection techniques, including satellite monitoring, run-off modelling and flood forecasting; as well as the implementation of a flood-warning system. In spite of these improvements to the flood-management system, weaknesses remained evident and flooding events disastrously re-occurred, most notably in 1988 and in 1992.
Scholars have noted several institutional limitations to adequately addressing the fundamental issue of flooding. First, a failure to adapt the system to natural processes like aggradation and erosion was causing a mismatch between river flow measures and actual hazards. Most water entering the system is withdrawn for irrigation purposes, leaving little water in the system to flush the channels and carry the highest silt loads in the world to be flushed out to the sea. This long term reduction in channel capacity to carry floods was one of the key reasons for exacerbating the effects of the exceptionally high floods in 2010.
Secondly, monitoring stations were, in some instances, unable to take measurements and report them in a timely fashion due to their own physical location relative to flooding. Even when measurements were taken and alerts were issued, public warning, evacuation and safety measures, in some cases, were ineffective and haphazard. On the flood-management side, canal and reservoir operators were not empowered to make important split-second decisions about flow adjustments that would ease flood hazards, and in some cases reservoir managers, for lack of system coordination, released waters exacerbating deadly down-stream flows.
Besides the systemic weaknesses at the macro scale, the negative consequences of flood hazard at the local scale are often disproportionately experienced by the poor and most powerless segments of the population. Because of hierarchical canal policies practiced by the British colonial administration and then the post-independence Pakistani government, the small farmers were often disadvantaged by virtue of being at the tail end of canal commands.
The canal administration system also has a strong colonial ethos in its legislation and bureaucratic practices, which discriminate against smaller farmers in terms of redressing complaints, water delivery and important levee-breaching decisions. All the infrastructure on the Indus basin rivers has a safe design capacity, which has been exceeded quite often in the past. To protect this infrastructure, upstream levees are often breached to relieve pressure. The operation of the breaching section is a decision taken by the local canal officer who is often influenced by local large-scale farmers. In such situations it becomes a question of which farmer has the most influence to either prevent a levee breach or to affect the breaching of an alternative levee. There are accusations in the Pakistani press that in fact, some of the levees were breached to protect the lands of specific influential interests. The veracity of the media claims is under judicial investigation but, suffice it to say, political influence in levee breaching decisions is a routine occurrence in Pakistan.
This historical perspective of Pakistan flood policy shows that, by ignoring the river’s natural systems and marginalising the poor, engineers and water managers have exacerbated the country’s physical and social vulnerability to floods. The relationship between anthropogenic environmental degradation and catastrophic flooding in Asia, Latin America, Europe and other regions is well documented. Conversely, we know there is an established link between healthy watersheds with flow capacity – wetlands, marshes, estuaries and mangroves – and flood mitigation.
Since disasters have been shown to be costly to long-term development goals, questions need to be answered about the need to invest in risk reduction. And, with the rising challenges of climate change, we must ask ourselves: can our engineered systems keep pace with climatic trends?
An academic version of this article was published in Water Alternatives. It is reproduced here with permission.
Daanish Mustafa is a Reader in Human geography and David Wrathall a PhD student at King’s College, London.