Perhaps the best analysis of how the error evolved is in a comment by Dan following Tim Lambert's post about the controversy. It would appear that the source of the error was in a 1999 India Environmental Port article, which changed a 1996 Russian's rough estimate of how long it would take all glaciers to melt from 2350 to 2035.
New Scientist helped perpetrate the error in a 1999 article, and now claims (wrongly, it appears) that its story was the original source of the error. NS journalist Fred Pearce also says that Indian glaciologist Hasnain had used 2035 in an interview with him. (In New Scientist he says it was an email interview; yet in The Times he is reported as saying it was a telephone interview. If he has the email, I would certainly like to see it. Pearce says Hasnain now admits it was just a rough estimate.)
Pearce's 1999 story claims that the 2035 figure appears in Hasnain's ICSI report, but it's apparently not there at all.
I would like to see directly what Hasnain says about this now. Was he responsible for the error in the Indian Environmental Portal article too? Or is it possible he's "confessing" to something he said in a interview 11 years ago of which he does not have a transcript? (He complains today that he never used 2035 in his research papers, and was never consulted by the IPCC before it used that figure. He's not denying he quoted the figure to Pearce, but I still wonder.)
Anyhow - there is no doubt at all that this is a very, very bad look for the IPCC, especially given IPCC head Pachauri's decision to come out swinging on a clearly wrong figure.
But - that was not really the point of this post. I wanted to note how confusing the whole topic of glaciers (and in particular the effects of their loss) appears to be. In particular, : just how important to Indian rivers is water from glacier melt?
This 2005 Nature report of Barnett & Ors about the dangers to water supply from melting glaciers (and less snow) is an important one. In it, we find in a section talking about the Himalaya-Hindu Kush area:
The hydrological cycle of the region is complicated by the Asian monsoon, but there is little doubt that melting glaciers provide a key source of water for the region in the summer months: as much as 70% of the summer flow in the Ganges and 50–60% of the flow in other major rivers[40,41,42]. In China, 23% of the population lives in the western regions, where glacial melt provides the principal dry season water source[43].This figure in bold sounds very high, but is repeated in many other places, although I won't link to them now. The references supporting the claim are not available for free online, and the abstracts at least don't seem to repeat it.
On the other hand, Science has quoted a note by an American hydrologist Donald Alford, the purpose of which is:
... to present the results of a preliminary analysis of the hydrologic contribution of the 5000 -7000+ m altitudinal belt of the Nepal Himalaya to the annual streamflow volume of the major rivers of Nepal, and to assess the hydrologic role of the glaciers within this belt.His conclusion (although it appears to be a very tentative one, pretty much a "back of the envelope" calculation I reckon) is that glacier melt only accounts for 4% of total annual streamflow of the rivers of Nepal. (I think all Nepalese rivers end up in the Ganges.)
Big difference, it seems. Is the issue that:
The Indus and Ganges Rivers currently have little outflow to the sea during the dry seasonas stated in an interesting recent study that found one Himalayan glacier seems to have put on no "weight" since the 1950's, since there was no radioactive layer from the atom bomb tests at that time. (So, if the Ganges has little outflow at all in the dry season, might it be that a very small feed from glacier melt might still account for 70% of it?)
The point of this "nuclear glacier" paper is that loss of glacier volume may be occurring by "high elevation thinning", and this has not been taken into account when working out rates of glacier loss. But, then, at the same time I have to admit that the paper repeats the mistake that:
The surface area of glaciers across the TP is projected to decrease from 500,000 km2 measured in 1995 to 100,000 km2 in 2030when it should have been (see Dan's comment above) 2350.
Furthermore, someone in comments at Real Climate has linked to some background notes used at the recent AGU conference for a press presentation which has lots of relevant information. (Be warned, it is a very big .pdf file.)
This is actually well worth reading carefully. They point out that Himalayan glaciers are behaving differently in different zones, but overall they are losing mass. Their estimate of the average rate of Himalayan glacier loss (measured by area, not volume, I think) is anywhere from .05% to .01% per annum (see page 14). If the higher rate is true (although they seem to think it unlikely) that would be 20% loss in 40 years. (Total loss in about 200 or so years, then, I guess; which isn't so far off the 350 years that we earlier mentioned.)
But as to the effect on water supply of Indian generally, the conclusion (see page 42) is:
As we have calculated, melting glaciers (specifically, negative mass balance components of the melt) contribute an estimated 1.2% (perhaps factor of 2 uncertain) of total runoff of three of the most important drainages, the Indus, Ganges, and Brahmaputra combined. The seasonal flow regulation influences and the negative mass balance is more important in local drainages close to the glacier sources, where glaciers can dominate the hydrology in arid regions, but on the scale of the subcontinent, glaciers are secondary players in looming hydrologic problems, which stem more from population growth and inefficiency of water resource distribution and application.So, there are some mighty confusing figures being flung around as to how important or unimportant glacier melt is to Indian water supply.
Is it possible that the very high figure of 70% of the summer flow of the Ganges (as mentioned in the Barnett Nature paper) is actually including basin snow melt and not just glacier melt? That could explain a lot. If climate change reduces snow in those areas, it may well be much more important issue than glacier melt, at least further downstream. And the title of the Barnett article is, after all: Potential impacts of a warming climate on water availability in snow-dominated regions.
Besides which, even without worrying about snow and glaciers, at least one study (and I am sure there are more) suggests climate change:
could influence monsoon dynamics and cause less summer precipitation, a delay in the start of monsoon season and longer breaks between the rainy periods.The reliance of India on the monsoon is pretty remarkable:
The summer monsoons are responsible for approximately 75% of the total annual rainfall in major parts of the region and produce almost 90% of India's water supply, he said.Anyhow, despite all this reading, I still remain quite confused on the issue. Glaciologists and hydrologists seem to have done a pretty bad job at dealing with the issue without confusing themselves, as well as the public.