New images of the Attabad dam and Spillway

The Pamir Times is now back in action after the owners were trapped on the road by the landslides caused by the recent heavy rain.  Tonight they have posted a selection of images of the site taken yesterday or today - go and take a look, they are good.  I will reproduce two here as they help to understand the current situation.

First, this picture shows the spillway from the lake side:

Two things to note here.  First, the freeboard appears to be in the order of 3 m, suggesting that the NDMA figures are about right.  Second, it looks to me like the far side of the spillway failed in the heavy rainfall earlier this week, which appears to have dumped a lot of loose material in the spillway, and it has narrowed it considerably.  I may be wrong about this, but that is how it looks to me.  Needless to say this is not a good sign.


The second image shows the area downstream of the main landslide mass:

The water that is flowing here is the seepage - there is clearly a considerable amount.  Note just how deeply the channels have incised, which shows just how erosive the materials in this mudflow deposit actually are.  Again, this is not a good sign.

The lake is now 18 km long.

Volcanic landslide in the Democratic Republic of Congo

Whilst the Attabad landslide continues to be my major blogging preoccupation at the moment (I'm guessing you might have noticed...), of course other landslides continue to occur around the world.  On16th May a slip happened on Nyiragongo volcano in the Democratic Republic of Congo.  The reports suggest that 43 people are likely to have been killed, and a further 1,500 directly affected by this slide.


Nyiragongo volcano is active, so this is likely to be a lahar - i.e. a debris flow in volcanic deposits.  This Google Earth image, although of low resolution, suggests that this is not the first failure on the east side of this volvano:

The El Salvador landslides of last year provide a good indication of the nature of a lahar - see previous posts here, here and here.

NDMA freeboard information for Attabad

In the absence of any other information on the state of play at Attabad, I have plotted up the daily updates on freeboard from NDMA.  The reliability of these is somewhat unclear as there are often contradictions between the indicated water level rise and the daily difference between freeboard measurements.  There are also sometimes errors - e.g. the data for 19th May suggests that freeboard increased rather than declined overnight, but I have corrected that.  So the graph for the last 19 days or so looks like this (the last datapoint is for 19th May at 18:00 local time):

Over the last four days the water level has risen at an average of 1.1 m per day, which means that the remaining freeboard should be lost in 3 or 4 days from 19th May, even allowing for the apparent slight reduction in rate in the last couple of days.  On this basis, overtopping might be expected tomorrow (22nd May) or on Saturday 23rd May, but it may happen earlier.

Of course all of this is dependent upon the NDMA values being correct, but for the last two weeks the Focus and NDMA freeboard figures have effectively been the same.  This does also mean that some of the media reported values of lake rise due to the GLOF earlier in the week were wildly incorrect.

I have no idea what is happening at Attabad now!

The situation at Attabad is now completely confused, so apologies for the lack of posts and updates, including to the monitoring site.  For all the right reasons Focus have now stopped collecting their daily data because the site is too dangerous.  The last measurement was before the GLOF, and the actual impact of that event is very unclear.  Meanwhile various news agencies in Pakistan posted suggestions that the lake had risen 10 feet, or even more.

Today NDMA posted their latest update which suggested that the freeboard is still 5.32 metres and that the water level rose just 0.27 m yesterday (as opposed to a rate of 90 cm per day a week ago).  Very strange!  At the same time, the Gilgit-Baltistan Times is reporting that NDMA have said today that overtopping will occur on 22nd May.

Finally, the Pamir Times appears to have stopped posting stories.

So who knows what is going on?  I will post again when I get more information.

Attabad - a video of a new landslide at the dam site

Thanks to Ivan Montanari for pointing this one out to me.  There was a big new rockfall at the Attabad landslide site today, caught on camera and posted on Youtube.  You should be able to view it below:



The implications of this include:
1. There is a chance that a failure like this could block the spillway;
2. There is a chance that a landslide could send a wave over the top of the dam;
3. If the slides are entering the lake then the water level will increase more quickly;
4. These slides of course represent a real danger to anyone in the area.

The final spillway at the Attabad landslide

The GLOF into the lake behind the Attabad landslide yesterday appears to have raised the water level by something over two metres, in addition to the increase of about a metre from the normal inflow (please see the data on my monitoring site).  The upshot is that the freeboard is likely be 4 metres or less.  Inflow has increased substantially as well to about 93 cubic metres per second.  It has not been possible to measure seepage in the last 24 hours as the site is now too dangerous. 

All work on the spillway has ceased and the machinery has been withdrawn.  The final form of the spillway is shown in this Focus image:

It pains me to say this, but I simply cannot see how this will resist the peak flow in the Hunza River.  However, it has at least served the purpose of reducing the total volume of water involved in any flood that should occur.

The boat service has also now ceased as it would clearly be too dangerous should a breach start.

Water flow past the blockage may start through one of three key processes:
1. Overtopping and water flow through the spillway.  In Tangjiashan the initiation of erosion started some hours after initial flow; this may or may not be the case here.  Erosion may start on the downstream face or at the crest of the spillway, or there may be slope failures on the downstream face.  It is also possible that the sides of the channel may fail, creating a new blockage, stopping water flow for a while. 
2. A seepage induced failure.  This is less likely than 1., but remains a distinct possibility;
3. Overtopping initiated by a landslide into the lake.

None of these can be ruled out at this stage, but 1. is the most likely.  If so, the size of the resultant flood will depend upon the rate of erosion, if any. 

I hope above all hopes that erosion happens slowly.  My thoughts are with the people of Hunza.

Updated: Attabad - the lake level rises by about 3 m

Updated: some newspapers in Pakistan yesterday reported that overtopping at Attabad wss imminent.  The news reports suggest that a Glacial Lake Outburst Flood (GLOF) today raised the water level to 358 feet (109 metres), which would be within 1 metre of the spillway. 

The GLOF has also been reported by the Pamir Times.

It now appears that the rise was about 10 feet (3.3 metres).  I am awaiting data to confirm this, but if so this would leave about 4.4 metres of freeboard intact.