The “science” component of “climate science” has seemed for some time to have been largely eclipsed by rhetoric, sound-bites, grinding axes, and the shrill voices of the blogosphere. Attempts at lowering the emotional level of the “debate” semantically - “global climate disruption” instead of global warming, climate “cynics” (avoiding the unpalatable connotations of “deniers”) – make little difference; the very human nature of cognitive biases will always win out. Although it is not my intent, for the moment at least, to enter the fray of post-climategate and post IPCC review turbulence, I would suggest that one thing is clear: the loss of trust in not only climate science but science as a whole is both tragic and potentially hugely damaging.
In this sad state of affairs, it seems to be a fact of life that, in order to find primary sources and reliable voices of reason, a complex and often highly irritating navigation of the blogosphere is necessary – much of the discussion (again, often a euphemism) takes place there, and much of it is vituperative, tribal, and composed of anonymous ad hominem commentary. It has therefore been with some pleasure that I have recently been following a new climate science blog, Climate etc., the initial welcome and purpose for which can be found here. To quote:
This blog provides a forum for the exchange of ideas about climate science and the science-policy interface. Climate Etc. is envisioned as a gathering place for climate researchers, academics and technical experts from other fields, citizen scientists, and the interested public. Open minds and critical thinking are required, and stimulating discussion and group learning is expected…The climate blogosphere is a vibrant environment but the signal is often hidden by the noise. Towards separating the signal from the noise, Climate Etc. will experiment with different formats…The idea is to have a focused and snark-free environment in which academics and other technical experts with little or no blogging experience will feel comfortable engaging with the broader community.
The blog has been set up by Judith Curry, who is a hurricane science and atmospheric modelling specialist and has been the Chair of the School of Earth and Atmospheric Sciences at the Georgia Institute of Technology since 2002. I met Judith at the Scifoo event a couple of months ago; I talked to her initially because the sandfish robotics work that had fascinated me was done at Georgia Tech but our conversations moved on to the issues faced by climate science and how to better deal with describing uncertainty (Judith kindly joined me in facilitating a discussion of this at the event). Despite her reputation as an academic scientist, she had recently been tarred and feathered in certain corners of the blogosphere for simply recommending that scientists should read Andrew Montford’s book, The Hockey Stick Illusion (see, for example, this review). I have now read it, and found it as compelling as a well-written detective story – I recommend it, but the details are for another time. Judith has started Climate etc. with the aspiration of providing a rational place for discussion of climate science issues, free, as far as possible, of “noise.” In my view, she has succeeded: her own posts are interesting, balanced, and provocative in the right way, and the commentary is largely free of rhetoric, filled with useful links, and, astonishingly, largely written by people who use their real names.
But I’m writing here about Judith’s blog not simply because I wanted to recommend it, but to follow up on her post from a few days ago, Pakistan on my mind. In it, she describes the fact that, in the aftermath of the devastating floods, “Most of the response of the climate research community to this catastrophe has focused on the attribution of the floods, i.e. whether greenhouse warming played any role in causing the floods,” citing, for example, Christiana Figueres, head of the U.N.’s Framework Convention on Climate Change, who stated that “All these events are constant reminders to governments that they need to deal in a consummate manner together to address climate change.” Judith asks two compelling questions: with what confidence can we ascribe these events to climate change and, regardless, is this the right approach to attempting to understand, manage, and ameliorate the impact of these events? The post and the (extensive) comments are informative and well worth reading. But I was particularly interested in the following section:
...the Climate Himalaya Initiative argues that engineering structures and human error may have played a major role in the catastrophe. There are a substantial number of barrages (dams) on the Indus River that support irrigation and hydropower. The flood occurred when the rising river bed (owing to the huge silt deposition in the upstream areas) was trapped by the Taunsa barrage, obstructing the water flow. These heavy silt loads were then transported through western tributaries of the Indus River. Construction of protective levees and dykes has also contributed to raising the riverbed and the sedimentation of upstream areas; moreover, the rising riverbed levels have rendered protective levees ineffective.
So, I thought that I would follow up a little (and my apologies – this follow-up has evolved into something that is certainly not “little”). The Indus system is huge, fascinating, and complex. A few facts:
- The Indus itself is over 3000 kilometres long, rises on the Tibetan Plateau, and drains an area of more than a million square kilometres (see the map and satellite image at the head of the post – and note the areas marked “modified landscape”).
- The vast network of tributaries flow into the Indus from Afghanistan, China, and India as well as from within Pakistan. The south-eastern tributaries are controlled by India, under the terms of the Indus Waters Treaty of 1960.
- The Indus is, by nature, one the largest sediment-producing rivers in the world, transporting debris from a huge area of the Himalayas where erosion rates can be up to 4 mm per year.
- The sediment load of the river (the suspended stuff – see my discussion of what we don’t know about bed load) varies enormously during the course of a year – during the monsoon and the summer snow melt the river will carry up to a thousand times as much sediment as it does at its low point.
- There are ten major dams and countless smaller dams and barrages along the Indus and its tributaries: it has been effectively engineered into the largest integrated irrigation system in the world.
As it emerges from the Himalayas, the Indus itself carries around 300 million tons of suspended sediment per year – let’s say around 15 million good-sized-dump-trucks-full. Prior to 1947, much of this sediment settled on the plains, but still a substantial volume reached the delta which was typically growing out into the ocean at a rate of 30 metres per year. But today, as it emerges from the mountains, the river reaches the huge reservoir of the Tarbela dam:
Downstream of the dam, the suspended sediment load is a mere 2 million dump trucks per year, the rest trapped behind the dam. Well, not right behind the dam, but rather dumped in the reservoir. When the dam was completed, in 1976, the estimate was that it would have a lifespan lasting only until 2030, by which time the reservoir would be completely filled with sediment. This proved to be pessimistic, but the fact is that 6 billion tons of sand, silt and mud had accumulated in the reservoir in the first twenty-five years of the project’s life. And all that sediment is building up in a huge delta below the surface of the water, slowly growing downstream towards the dam; not only would an earthquake potential set off a devastating underwater landslide, threatening the dam, but, even without such an event, the delta is fast approaching the intakes and tunnels of the dam. So, what can be done about this?
Well, unfortunately, the answer requires a slight diversion into the arcane but vital concept of a river’s base level. I made an attempt to describe the concept in the book:
Take a map of a river and run a piece of string along its winding course; then straighten out the string and draw a graph of the river’s elevation versus the distance from its source—the profile of the river. For any river at any given time, there are two points that are externally fixed: sea level and the elevation of the source of the river in the mountains. Between these two points, the river is free to behave in any number of ways. For the river, sea level is the primary base level, its ultimate terminus, below which no part of the river can drop (to avoid the impossible feat of flowing uphill). Base level is a concept universally agreed on and at the same time hotly debated, but the principle is clear: change a river’s base level by raising or dropping sea level, and the river must change its behaviour. If base level drops, the river must erode, cutting downward to reflect this. If base level rises, the river must build up its profile by depositing sediment. A change in base level changes the gradient of the river’s profile, requiring the river to make changes in order to smooth out that profile.
So sea level is the primary base level and sea level changes profoundly alter a river’s behaviour and architecture. But put a dam in the middle of a river and you have introduced an entirely new local base level, and, upstream of that dam, the river responds. You have raised the base level from the natural altitude of the river’s valley before the dam and therefore the river reacts by depositing its sediment load. If the level of the reservoir is maintained as constant then the system reaches a new stable state, but if the level varies, then the river continues to respond by erosion or deposition. In the case of the Tarbela delta problem, the reservoir level has been raised to induce more sediment to be deposited upstream, therefore depriving the delta of sustenance. But that upstream deposition continues to completely modify the geometry of the valley (that had anyway not been natural since the dam’s construction) – and the environment on which depends the livelihood of everyone who lives there. The floods upstream of the Tarbela dam were appalling.
So this is but a vivid example of how the entire Indus system has been modified. It is modified every time a dam is built and it is modified every time the level of water in a reservoir is changed. It is a totally and dynamically unnatural system, deprived of sediment and, today, virtually no sediment at all reaches the delta of the Indus. Floods on the scale of those recently experienced would always have been devastating, but their impact is exacerbated by the wholesale manmade modification of an entire river system.
In her conclusion, “Looking forward,” Judith Curry writes:
While climate researchers and policy makers ponder whether the flood can be attributed to global warming, it seems to us that this emphasis diverts attention from actually using climate, meteorological and hydrological knowledge and research in the application of pressing current needs in the developing world. Substantial benefits would accrue directly for developing countries, as well as indirectly for other countries concerned with global security, by applying advanced research and technologies to support:
- improved river routing models for major rivers
- advanced techniques for providing probabilistic flood forecasts on timescales exceeding two days and integration of these forecasts into early warning systems, preparedness and contingency plans, and rehabilitation measures
- application of advanced dynamical and statistical methods to assess future risk of floods and droughts, and integrate this information into engineering assessments for future water structures
- improved probabilistic rainfall forecasts on timescales of days to weeks to enable farmers to optimize agricultural decision making regarding planting and harvesting.
Based upon our work in Bangladesh along these lines (see here and here), we have demonstrated that such developments are feasible and can be successfully integrated into national meteorological and hydrological services. We have contacted the Pakistan Meteorological Department and have had discussions with relevant personnel in USAID and UNDP regarding how we might help. The current situation in Pakistan is mired in both national and international politics. However, we hope that the international climate, meteorological and hydrological communities can offer something to help Pakistan other than a debate over whether global warming contributed to the floods and a mistake in the IPCC AR4 regarding the melting of the Himalayan glaciers.
Other than possibly adding a few more geological – and specifically sedimentological – components to her list, I agree wholeheartedly and can only wish her and her colleagues every success.
For a further sense of the complexity of the Indus basin and the distribution of dams and barrages, plus data on catchment flows in July and August this year, here’s a map from the World Food Programme – the high resolution version can be found here.
[For an overview of the role of dams in the Indus basin, see this paper from the International Commission on Large Dams. For details on sedimentation in the Tarbela dam reservoir, see http://www.adb.org/water/topics/dams/pdf/cspkmain.pdf, and http://www.paralia.fr/cmcm/e01-28.pdf; for an interesting piece of research on the sands of the Indus, see “Petrology of Indus River sands: a key to interpret erosion history of the Western Himalayan Syntaxis”; for extensive documentation of the recent landslides in the upper reaches of the Indus system as well as of the flooding, see Dave’s Landslide Blog, written by Dave Petley, who is the Wilson Professor of Hazard and Risk in the Department of Geography at Durham University in the UK.]