Pakistan flood update: the Kirthar Canal

The floods in Pakistan continue to extend to new areas, increasing the magnitude of the impact on a daily basis.  The PakMet hydrographs continue to show increasing flow levels.  At Guddu the discharge is still rising, and is now close to the peak level reached in the first flood wave (click on "read more" to reveal the rest of this post):

Meanwhile, as expected serious problems are developing in the area of (but so far not on) the Guddu Barrage, where the discharge is also increasing:

It appears that the second flood wave has not really reached Sukkur as yet.  However, Dawn this morning reports that: "...residents of Qubo Saeed Khan town and over 100 villages in Qambar-Shahdadkot district were asked to leave their homes on Monday after a powerful current of floodwaters gushing from Garhi Khairo and overtopping the Khirthar canal smashed five gates of the Garang regulator and washed away the entire structure." 


The importance of Sukkur Barrage and the canal network that extends from it is described in this old Pakistan Observer article "Federal Minister for Labour and Manpowewr Syed Khursheed Ahmed Shah has said that Sukkur Barrage is the lifeline for Sindh’s agricultural economy".  Agriculture in this area is heavily dependent upon the water delivered by this canal network; damage to these structures will represent a major problem in the future.  Note also that it is likely that the water will have carried subtantial volumes of silt, which may be deposited in the canal network.  This may need to be removed once the floods subside.  Meanwhile, the water will have destroyed a substantial part of the cotton crop, which should be harvested next month, which in turn will damage the textile industry.  Farmers would normally plant winter wheat in the autumn; the viability of this crop must also be in doubt.
 
Upstream, the impact of the floods and landslides in the mountains is being well-described by the Pamir Times.  The Pakistan High Commissioner in London has stated that rehabilitation will require about $15 billion of investment and a "Marshall Plan" style of approach.  It is interesting to see the UK Deputy Prime Minister describing the international response to the floods as "lamentable".  To date the UK has committed £31 million to the Pakistan flood disaster, which to be fair is reportedly a quarter of all of the assistance pledged.  However, in comparison with the cost of the Afghanistan and Iraq wars (£20 billion over the last ten years for the UK alone), this amount is tiny.  Investing in aid to Pakistan, and being seen to be a force for good in a troubled area, would seem to be a good strategy.  Building on the Marshall Plan analogy, isn't it time for a Berlin Airlift style aid effort?

Update on China and Pakistan - 13th August 2010

1. The Gansu landslide in China
The rescue operations in Gansu have now clearly transitioned into a recovery and rehabilitation phase, greatly hampered by continuing very heavy rainfall in the area.  Landslides are continuing to occur in the area, with resultant casualties.

Tom Dijkstra of Loughborough University visited the site as part of a trip to look at collaborative landslide research in the area back in November.  He has kindly sent two images of the town to me and has agreed that I can put them online.  This image shows the area affected by the landslide from the other side of the river.  The steep, deforested mountains in the background are clearly the source of the flow:

The second image shows the catchment source of the flow:

It is worth comparing the above image with this AP picture that I posted a few days ago of the site after the flow:

The reasons for the very high loss of life, now estimated to be 1,144 people, with a further 600 still missing, are clear given the density of buildings in the affected area.

2. Pakistan
The two flood waves in Pakistan continue to cause extreme levels of suffering.  According to the FFD hydrographs the water level at Sukkur is now falling slowly after the passage of the first flood wave:

What has been particularly interesting though is that the flood level is not really increasing substantially at Kotri, the large gauging station downstream:

This presumably means one of two things.  First, it could be that the water is finding another route - i.e. that it is bypassing the gauging station.  Alternatively, the water is in effect trapped between the two sites, which might explain the very slow falling limb of the hydrograph.  The Google Earth satellite image below shows Kotri:

It is possible that the water has flooded the adjacent land, but the news reports indicate that this is not the case, with the suggestion that the bridges downstream of Sukkur are slowing the flow down.  This is dangerous in the context of the second flood wave, which at the moment remains smaller than the first.  This is the hydrograph for Taunsa, which is just below the "Extremely High" flood level:

At Guddu the discharge is still falling, but only very slowly.  Indeed the discharge remains well above the "Exceptionally High" level:

The danger must be that the second flood wave starts to catch up with, and build upon, the stalled first wave.  This would create the potential for an extremely damaging second phase of floods.  It took six days for the first wave to pass from Taunsa to Guddu, and a further day to Sukkur.  The hope must be that the water level starts to fall quickly at these two sites before the second wave arrives.

Unfortunately, it is clear that this slow motion disaster has several more weeks to go, even if there is no further heavy rain.

Update on the Pakistan floods: 8th August 2010

The Indus flood wave is continuing to flow steadily southwards across the country.  The annotated WHO flood map below, from my earlier post, is for reference in terms of the locations of the hydrograph stations:

The Pakmet hydrograph data shows that the water level at Taunsa is now declining markedly, although it remains in the defined "high" flood category:

Downstream at Guddu, the discharge has been rising, although it should now be close to the peak level:

Downstream at Sukkur, which is the area that is attracting most of the media interest at present, the discharge has been increasing very rapidly:

The barrage here apparently has a design capacity of 900,000 cubic feet per second, which is clearly a concern, although it is likely that it will survive the flood.  Other agencies have suggested that the design flood is as high as 1.5 million cubic feet per second. It is however reported that a 25 m wide breach occurred in a levee at Kandhkot, about 70 km upstream of Sukkur.  Other levees are showing signs of distress, and one levee (at Ghouspur) has been intentionally breached to try to reduce the size of the flood wave.

The final station downstream is at Kotri.  Here water levels have yet to start to rise rapidly.  The peak of the  flood wave is expected to arrive in two or three days from now:

Meanwhile, the official weather forecast states "widespread rain-thundershower, heavy at times, expected in Punjab, Sindh, Eastern Balochistan, Khyber-Pakhtunkhwa and Kashmir." 

The impact of these floods on the long term welfare of the people of Pakistan should not be under-estimated.  The combination of loss of life, loss of personal assets, loss of dwellings, loss of crop, damage to infrastructure, loss of sanitation and loss of clean drinking water across such a large area cannot be underestimated.  That aid agencies are describing this as being worse than the 2005 earthquake, which is thought to have killed 100,000 people, gives an idea of the scale of the catastrophe.  I would also note that, given the troubled history of this area, there is a very real opportunity for Europe and the US to be seen to be a force for good by providing a proper and effective response to the disaster.  This would surely be a good investment for the future of both Pakistan and the west.

Update on the Pakistan floods: 6th August 2010

The flood wave in Pakistan continues to travel down the Indus River and is now approaching Sukkur.  The authorities are responding with a large-scale evacuation of low-lying communities - The Nation reports that 500,000 people are being relocated but that 1.5 million people are likely to be affected.

The latest FFD hydrographs depict the situation very clearly.  The locations of the gauging stations are shown on this annotated 2007 OCHA map of national flood potential for Pakistan:

At Taunsa the FFD hydrograph is now on the falling limb, but the discharge is declining slowly:

Meanwhile, as expected the water level at Guddu is continuing to rise towards the million cubic feet per second level, and is now substantially above the extremely high flood level:

The low-lying telief around Guddu is clear in this Google Earth image of the area:

The barrage evident in the image above is deigned to direct flow into the channels on either side of the river, which then provide irrigation to about 12,000 square kilometers of land.  Clearly, damage to this structure would have a substantial impact, but the reported capacity is 1.25 million cubic feet per second.

The next key location downstream is Sukkur, where the hydrograph is now showing rapid increases in discharge:

The Google Earth image of Sukkur shows that it is a much larger settlement:

Sukkur also has a barrage, as shown on this Panoramio image:

This barrage provides irrigation to 20,000 square kilometres of land.  Damage to the structure would be serious.  The reported capacity is 900,000 cubic feet per second, so it should survive.   

Finally, downstream is Kotri, where the discharge is also now starting to increase:

It will be a few more days before the main flood reaches this point.

Meanwhile, the Pamir Times has published another set of images of the state of the Karakoram Highway, of which this is the best:

Two things to note here.  The obvious one is the huge amount of material deposited on the road by this debris flow.  The other is the huge queue of trucks in the background.  These trucks are trying to travel southwards out of the mountains.

A further update on the Pakistan floods

The media have now cottoned onto the fact that the flood wave is causing destruction downstream from the northwest of Pakistan, an issue that I have been highlighting for some days.  The flood waters are certainly abating now in Taunsa, although note that three days after the peak the discharge is still above the "very high" level:

Meanwhile, at Guddu the water level is still rising, and is now close to the "extremely high" discharge level.  It is expected to peak in the next 24 hours. 

The discharge at Sukkur is now very much on the rise as well as the flood wave starts to approach:

It is frightening to think that this flood has at least five days to run before it discharges into the ocean.  There is a huge amount of damage yet to occur.

Meanwhile the Pamir Times has published a further set of images of the Karakoram Highway, amongst suggestions that it will take six weeks to reopen it.  The latest images are well worth a look - the damage is very serious indeed:

As I noted during my drive up this road in February, a huge amount of investment has been made in the last couple of years in this road, with lots of work on new, inadequate, culverts.  It is tragic that so much of this investment will have been wasted.  Meanwhile, the hardship suffered by those upstream of the Attabad barrier will have been amplified by the lack of communication along the remainder of the road.  One genuinely has to fear for the welfare of those living in Hunza now.

Unfortunately, there is now no data about the state of play at Attabad itself.  The continued threat posed by the dam could be forgotten amongst the flood tragedy.  This must not be allowed to happen.

Latest update on the flood wave in Pakistan

The flood and landslide disaster in Pakistan continues to play out in slow motion.  The recovery operation in the north of the country continues to be hampered by the level of destruction inflicted upon the roads.  The Pamir Times yesterday posted some images of the Karakoram Highway north of Gilgit, which serve to show all too well just how difficult the recovery operations are going to be:

The areas affected by the flood, and those that will be affected over the next few days, are shown on this new map from OCHA:

A good sense of how bad things are in the northwest of the country can be gained from this Save the Children rapid assessment report for the Swat Valley.  They found that, for example: 

• 21% of housing structures have been damaged; 
• 3% of housing structures have been effectively destroyed;
• Only 28% of household have access to a functional latrine;
• In the 80 surveyed communities, there were 75 reported cases of separated children, 6 cases of missing children, and 16 children who have lost one or both of their parents. 
• 26% of school buildings have been damaged
• 54% of communities reported that most families in their communities have no food stock available at home
• Only 1.5% of communities reported to have received food aid
• For 45% of surveyed communities, the main sources of drinking water are rivers, ponds, or lakes
• 55% of communities reported that injured and sick community members are not receiving any medical treatment.

Meanwhile the flood wave continues to move downstream, causing misery as it proceeds.  Judging by the PakMet hydrographs. It is now clear that the flood has peaked at Taunsa (for locations see my earlier post; they are also marked on the OCHA map above):

Whilst the river level is now rising rapidly at Guddu:

And has started to rise at Sukkur:

Meanwhile further rain is forecast for Sindh today, which could serve to exacerbate the situation.

Pakistan floods - the progression of the flood wave down the Indus River

The floods in Pakistan appear to be starting to generate substantial interest in the UK, although this may in part be a result of the perhaps surprising decision of the Pakistan president to travel to London today.  Whilst the crisis continues to develop in the north of the country, of equal concern now must be the progression of the flood wave down the Indus River.  Of course Pakistan is a country built around the Indus - the green swathe that cuts north-south across the deserts of central and southern Pakistan (see image below) is essentially defined by the Indus and its tributaries.  The Google Earth image shows Pakistan with the various gauging stations and other larger settlements on the Indus:

The passage of the flood wave is rather well illustrated by this series of hydrographs, from the Flood Forecasting Division website.  The units are cubic feet per second - there are 35.3 cubic feet in a cubic metre.

The hydrograph for Besham, in the north looks like this.  The peak occurred around about midnight on 29th July.

A little further downstream lies Tarbela, whose hydrograph looks like this.  The peak occurred early on 30th July.  Note that Tarbela is a (very) large dam, so will probably have modified the flow somewhat.
 

 Further downstream lies Chsashma (see the map above), whose hydrograph looks like this.  Note that here the peak occurred late on 1st August:

And downstream again is Taunsa (see map above), where the hydrograph looks like this.  In this case the peak flood has yet to arrive.  However, already the flood is defined as "extremely high".

Taunsa is 800 km from the mouth of the river as the crow flies (so over 1000 km along the length of the river), so it will take some days for the flood wave to reach the sea.  These lower reaches are mostly surrounded low lying land - the fabulous image below, by Najamuddin Bhatti, shows Sukkur on the banks of the Indus for example:

The reasons for concern about the potential impact of this flood wave downstream are clear.

New images of the level of destruction in the Cusco area of Peru

The Spanish language blog El Caminerito has been covering the magnitude of the rainfall, flood and landslide disaster in Cusco, Peru. They have put together a very helpful map showing the locations of serious damage:


Ver Cusco en Emergencia en un mapa más grande

The scale of the disaster, which is still being ignored by the western media in favour of coverage of the tourists at Machu Picchu, is well-illustrated by these images of the Huacarpay region:


Meanwhile, via the Typeboard site, the Spanish language site Peru.com reports that the village of Zurite was severely damaged by a landslide, which sounds to be a debris flow (Google translation):

" the landslide of mud and stones covered the Plaza de Armas, the town’s Church, the main streets of the city and damaged 500 houses. The incident occurred about 3 pm on Thursday after they noticed a crack on a hill and began to take appropriate action, emergency services were able to evacuate the entire population."

This is the Google Earth imagery of Zurite:


There is no shortage of landslide scars on the hillside above the town, plus the quarries, one of which appears to have excavated out the toe of one of the scars.

Here is an image of the central square via the Panoramio site:

Debris flow damage from Typhoon Morakot in Taiwan

In addition to visiting the Shiaolin landslide in Taiwan at the weekend, we also managed to visit a couple of other places that had suffered damage during typhoon Morakot. Damage from debris flows and river floods occurred extensively throughout the upland areas of southern Taiwan, as these images show:


The government has set aside about US$5 billion for reconstruction. Unfortunately this area is threatened by another typhoon, called Parma:

Special session on Typhoon Morakot in Taiwan

This afternoon at the Chi-Chi earthquake conference in Taiwan the organiser laid on a special session on the impact of Typhoon Morakot in August. This is of great interest to me, given the impact of the landslides, so I thought I'd give summary of the key points. Apologies for the note form - I have done this Twitter-style!

1. The magnitude of the typhoon
For Taiwan this was an extraordinary event. It appears that in terms of river discharge the floods were the largest since records began - over 200 years ago. For example, Prof. Tsai of the National Cheng Kung University showed that for the Gaoping River the peak discharge was 29,100 cubic metres per second. That is the equivalent of the daily water needs of 150,000 people - each second in a single river! These huge floods were driven by extraordinary rainfall. At Chiayi the statistics are as follows:
Total storm rainfall: 3005 mm
Maximum hourly rainfall: 136 mm
Max 3 hour: 325 mm
Max 6 hour: 548 mm
Max daily: 1623 mm

This is close to but not actually quite, the world record rainfall.

2. Why was it so intense?
Prof. Jou of the Taiwan Meteorological Agency suggested that there were two key reasons why the typhoon was so intense. First, the storm slowed down as it crossed the island. Before it made landform it was moving at 20 km per hour. When it came ashore it moved only about 50 km in 24 hours. This led to very high rainfall accumulations. Second it appears that as it formed the typhoon interacted with a substantive monsoon trough, which drew in moisture from the inter-tropical convergence zone to the southwest. This meant that the typhoon generated huge rainfalls on the south edge of the storm which is where the maximum damage occurred. Interestingly, he admitted that one of their dynamic models run the day before the typhoon struck forecast 1900 mm of rainfall. However they didn't trust the model.

3. Landslide impacts
Very little was presented on the impact of the landslide at Siaolin, but Meei-Lin Ling stated that the death toll was 491 people. She stated that at the moment they have records of 1349 landslides, 46 debris flows and 298 road slope failures. I suspect that that many of the landslides may be classified as debris flows? These landslides cover an area of over 50,000 hectares. Prof. Jenn-Chuan Chern the Deputy CEO of Morakot Post-Disaster Reconstruction Council, suggested that the volume of sediment is 56 million tonnes (I wonder if this is rather low though?). Tainan County had the largest number of failures (515) the Kaohsiung (288) and Chiayi (216). The landslide distribution closely reflects the rainfall distribution. In Sinkai village 32 people were killed by a debris flow. There was also very extensive damage to the highway network - there is some doubt as to whether the road to Alishan (a major tourist area) can be repaired.

4. Other damage and impacts
Coastal flooding was a major impact. The peak of the storm coincided with a high Spring tide, causing major inundation. These flooded areas and those affected by river floods, have had major problems from silt accumulation. In some cases over a metre of slit has had to be removed from houses. This is quite straightforward, but clearing the water and sewage pipe network has been a major headache. Over 500,000 tonnes of wood has been deposited and is having to be removed. This is a substantive task.

5. Human costs
Human impacts are 701 fatalities with another 58 missing. 120,000 houses were flooded, 310,000 houses damaged. 4489 people are being housed in army barracks even now; reconstruction is going to be complex. At present 55 aboriginal tribal villages have been displaced. Government evaluations suggest that 31 of these are permanently unsafe. Therefore resettlement is not simple - the aboriginal communities want to reconstruct their villages in order to maintain their culture (e.g. through targeted schooling), but given the dangers of the sites this is not simple. Government land is being earmarked and NGOs are helping to liaise to find an appropriate solution.

6. The elephant in the room - climate change
None of the speakers wanted to ascribe this event to climate change - there was a real sense of caution about saying anything unwise. This is sensible. The point was made that in recent years the number of landfalling typhoons has been high compared with historic record and also that a dramatic increase in precipitation intensity has been noted. The floods met and even the calculated probable maximum flood of the rivers. Thus, it appears that the rainfall that is now happening is different from the historic record. Is this climate change? They key factor seems to be the interaction of the typhoon and the monsoon. A discussant implied that this could be because the monsoon front is moving. Whether it is climate change or not it does seem sensible to plan for more intense rainfall events. This is going to be very challenging in a place like Taiwan.

7. Meteorology
The final interesting comment was a note that classifying typhoon intensity by wind, which is the convention, is meaningless when most of the damage is done by rainfall. They suggested that a new classification is needed that combines both wind and rainfall, allowing better forecasting of landslide and flood impacts. Frustratingly I submitted a grant application to DFID ten years ago to develop a scheme to do exactly this, tied to a terrain classification scheme. They didn't fund it.

All-in-all a fascinating set of presentations that really helped in the understanding of this extraordinary event.

Your questions and thoughts are welcome!

A video of a river bank failure

The BBC News website has a short video showing the effects of heavy rainfall in Southern China in mid August. Most notable is a pretty remarkable piece of footage of buildings sliding down a river bank and into the torrents below. You can view it here.

Updated: Tangjiashan (Beichuan) dam - a summary of what we know

In the next 24-48 hours the water level at Tangjiashan should reach the spillway and flow should start. I thought therefore that it would be useful to summarise what we know about the site:

• The landslide dam is 124 metres high;
• At the time that the over topping begins the lake will have a volume of about 205 million cubic metres of water (UPDATED - the volume is now estimates as 211.6 million cubic metres as at 2 pm (loca) Wednesday. The lake level had 1.37 m to go before reaching the spillway)
  
• The channel that has been cut appears to be about 5 (now known to be 7 m) wide at the bed (though note that the film with the journalist in the bed suggested less than this) and about 15 m at the top. It is 12 m deep The bed is unlined and unprotected. It is formed in silty materials with a few clasts (stones);
• Failure can occur through any of three mechanisms:
• The water can reach the spillway and start to flow over the dam;
• The two known seepage points (total reported flow = 10 cubic metres per second) in the dam could cause collapse;
• A landslide into the lake could trigger a wave that would overtop the dam. A dangerous slope has been identified at Xuanping township.
  
• Downstream there are three further (smaller) landslide dams in the Beichuan area;
• The Chinese government has relocated about 250,000 people. In total, about 1.3 million would be threatened by total collapse.

Reuters image of the top of the dam showing the lake and the spillway under construction

These are the best and worst case scenarios for the dam as I see them:
Best case: water flows over the spillway and erodes it slowly. The lake level reduces gradually over a few weeks;
Worst case: a large landslide into the lake triggers a wave that overtops the dam, causing complete collapse. The entire volume of the lake is released very rapidly. The resultant flood pools behind and then breaks through the downstream dams, creating a devastating debris flow.

The most likely scenario at this stage is that water will flow over the spillway. A ball park flow rate appears to be about 100 cubic metres per second when there is no rainfall. This channel will not be able to cope with this as far as I can see, leading to rapid erosion and therefore collapse of the dam. I suspect that most of the lake volume will be released in a few days at most. The water is likely to pool briefly behind the landslide dams downstream, which will probably increase the size of the flood wave once released.

Where will the water go? I have tried to produce a quick Google Earth summary image below of the anticipated flow path (click on the image for a better view):-


The flood is likely to be large, fast and very damaging in the mountains, but should spread at least a little and slow on the plains. We can only hope for the best.

I am travelling to Nepal this afternoon, but will do what I can to keep this updated from there.