Updated summer 2017
In 2012, following weeks of ostensibly effortless liberations of north Malian towns while jihadists repeatedly slipped away into the desert to fight another day, in early March the French-led Operation Serval concluded a decisive but protracted battle in the far north. At the time the Chadian army announced the death of the notorious Abou Zeid (‘ABZ’, left) in the valley of Amtetai (or Amettetaï) somewhere in the Adrar des Ifoghas hills. It was confirmed by DNA tests a few weeks later.
The strategic significance of this obscure valley became clear when the assault began in mid-February. A huge effort was made to defend it during which time scores of militants as well as many Chadian soldiers were killed and huge caches of arms were found. A few days after Ametetai had been ostensibly cleared there was perhaps hasty talk by the French of drawing down their troops in April, while the French defence minister deemed it safe enough to make a surprise visit to the battle site a few days later.
Detailed description of the battle and aftermath
Video from Chad TV
Report from France24 after the battle
Report from Le Monde
Just a day after ABZ’s death, Moktar Belmoktar (‘MBM’), supposedly ABZ’s rival and with him one of the three key jihadist ’emirs’ in north Mali at the time, was also (but incorrectly) reported killed by the Chadians. In fact the Pimpernel-like MBM had been ‘killed’ many times in the last decade or two (mostly recently in 2015 and 2016). In 2013 shot to Bin Laden-like prominence as the instigator of the suicidal attack on the Algerian gas plant at In Amenas.
It’s more likely that at the time MBM (right) may have been directing AQIM operations further south, possibly in the region of Tin Keraten. Old maps label this as a well and a oued, some 100km northeast of Gao and 220km south of Kidal (see map above left). It doesn’t seem to match the terrain profile of Ametetai, but a battle was reported here a day or two later and where a fourth French soldier was killed. It’s also just as likely that before it’s all over MBM may yet again slip away into the neighbouring countries or that he was never there at all. The latest news is that following an ineffective US air strike in 2015, in November 2016 a French drone strike in southwest Libya injured MBM gravely. He has been removed from the American RFJ programme and has since lost his position in the Al-Mourabitoun group he led.
Real name Mohamed Ghadir, a small-time smuggler from Deb Deb in eastern Algeria close to the Libyan border (a book including his background reviewed here), Abou Zeid named his katiba or brigade ‘Tarek Ibn Ziyad’ after the 8th-century Moorish general who conquered Spain (‘Gibraltar’ is an anglicised version of Jebel Tarik). Following spells in prison, Abou Zeid rose up the ranks of the GSPC and became noted as a hardliner – ruling his group with ‘an iron fist’ as one former French hostage explained (her husband remains in captivity in north Mali). ABZ was said to be responsible for the execution of British-born hostage Edwin Dyer in 2009 (see link below) as well as more recent mutilations in Timbuktu in the name of Sharia laws. More on ABZ here.
Many news reports parroted the ‘Ametetai valley’ as the location of this decisive battle which led to ABZ’s death, though none mapped it any more precisely than this BBC image. Seeing as it’s rare to get a specific location of what you presume might have been a key jihadist stronghold where some of the ten or so hostages may be kept and where ABZ’s brigade made their last stand, I tracked Ametetai down. (Some reports also mentioned air raids on ‘Etagho’ or ‘Oued? Entouwike’, as the ‘key to unlocking the bases in Adrar Tighaghar’ massif, but again, no accurate location was given and I couldn’t track them down on the usual maps.)
A well with the name ‘Ametetai’ was easily found on the IGN ‘Kidal’ NE-31 map (right), located on the northern edge of the Adrar Tighaghar hills and about 60km northeast of Aguelhok. Aguelhok is a village on the trans-Saharan Tanezrouft track which runs from Reggane, Algeria down to the Niger river and Gao (Route B1 in Sahara Overland).
Bing vs Google imagery of Ametetai
Zoom out and around this Google maps link centred close to the site of the well as shown on the IGN map and you’ll see this ~17km2 sat image tile or quadrant (dated Jan 2006 at time of writing) is higher res than the surrounding ones. The next nearest in high res are Aguelhok and Tessalit (which we visited in 2006 laying plans for our big crossing of north Mali) so a random hi-res tile in the middle of nowhere always makes you wonder. But the Ametetai tile is also oddly spotted with unseasonal clouds for January, few of which have corresponding ground shadows which seems even more odd. Is this obstruction of ground detail significant?
For the first time comparing Google with the same area on Bing ‘aerial’ maps, it comes up as much brighter and cloud free – clearly from another occasion and time of day. On Bing there’s much more cultivation evident (see above right) as well as better developed and unusual concentrations of tracks criss-crossing the hills as below right. This suggests Bing imagery is more recent except that there are more trees which implies it’s older. In my experience trees get cut down all over the Sahara quicker than they can grow back, even by a oued.
Continuing the comparison, some walled enclosures on Google are not present on Bing (above left), while a few more on Bing have been enlarged as above right and on the left (location upstream). Perhaps the ready-made stone in old walls was taken down to expand compounds elsewhere? Finally, are the shadowless Google clouds covering up something ‘tactical’ that’s exposed on Bing Aerial? Not as far as I can tell.
You can play this ‘spot the difference’ game for a while to try and fit what is now well known to all: the ‘narco-traffiquants‘ as Chad reports derisively describe AQIM/MUHAO/Ansar Dine, had a major stronghold in the Ifoghas as has been speculated on for years. And the Ametetai valley was probably it. Here they dug caches and laid traps while successfully camouflaging their actual presence from overhead drones and presumably preparing for the assault which finally came in the form of Operation Serval. The seeming expansion of cultivation and some enclosures in a place so far from other know villages with little evidence of actual dwellings might be seen as odd.
A couple of weeks later it was reported that fighting was continuing in the ‘Hades valley… 60km from Tessalit’. This could be the next, even narrower valley to the south of Ametetai in the Tighaghar massif called ‘In Tegant’ on some maps. An unusually dense mass of tracks joins the two valleys across the stony hilltops, including the image above right.
Note: Aerial images in the region may have been updated since this article was written
What is the ‘Adrar des Ifoghas’?
‘Adrar’ usually means a region of low hills, escarpments or plateaux, but it can mean a single peak or jebel, like the Adrar Asref in the IGN map above. The Ifoghas (pronounced ‘Iforas’) is a region of low and dispersed, rubble hills spread across northeastern Mali and is also called the ‘Adagh’ by some. Separated by broad sandy plains and dissected by slender oueds, the actual ‘massifs’ barely rise a hundred metres above the surrounding 500m plain and can never be said to resemble ‘mountains’ far less a ‘near-impenetrable range’ as some reports describe. There are no distinctive peaks to speak of, nor a contiguous massif like the Hoggar in Algeria or the Tibesti in Chad, although you could say the Adrar Tighaghar is a small massif or eroded plateau within the Ifoghas.
As with most low-lying Sahara hills, the composition is mostly sandstone or granite (as left), broken down by eons of weathering. In places this makes travel with anything with wheels very difficult or impossible, be it the sandy oueds which at least make natural byways, or more so on the crumbled plateaux to either side, as the Le Monde image left shows.
In my experience in the Sahara, it only takes a few inches variation in the ground’s surface to reduce car speeds to jogging pace, be it sand ripples or rubble and rock ledges. Bikes manage a little better, but even on foot the effort on the sort of terrain shown above is hugely amplified. Nevertheless, for in the hills either side of the Oued Ametetai the only way in was on foot with helicopter support. (As the reports above explained, the French and Chadians did so from three sides on the final assault.) This sort of terrain is so gnarly it’s also frequently but incorrectly ascribed as ‘volcanic’ in origin, but whatever you call it, trying to across it in 40°-heat with a 40-kilo backpack while trying to dodge bullets would be hard work indeed.
However, as this blog points out, comparisons with Afghanistan’s much more rugged and allegedly betunneled Tora Bora where OBL narrowly evaded capture in 2001 are inaccurate (as is calling all the north Malian jihadists ‘Al Qaeda’). Having said that the Ifoghas’ location close to the borders of Niger and Algeria does match the Tora Bora along the Pakistani border. This must have been a factor or a welcome coincidence when the Algerian-based GSPC (later AQIM) first established itself here in the late 90s. Hostages were frequently brought down into the Ifoghas from the north or east to a place which has always been a remote outlier in Mali, with Bamako some 1200km away. You get the feeling the indigenous Kel Ifoghas Tuareg were pretty much left to themselves or neglected, depending on your point of view, while from my observations a minority of entrepreneurial Berabish Arabs (Moors) controlled a low-value smuggling commerce from places like Al Khalil (In Khalil) hard against the Algerian frontier. (That all escalated once light, compact and high-value cocaine entered the picture a few years ago).
More than the sporadically rough terrain, the main thing that hampers occupation in the Ifoghas is the need for water. There are no river or lakes of course, but in this sub-tropical south side of the Sahara there are few perennial waterholes as can be relied on in southern Algeria (left), for example and which are easily spotted on sat imagery.
One benefit of the bare rocky massifs is they drive what late summer run off there is straight down into the many oueds where vegetation gets concentrated and trees can sink their roots, and where wells can more easily reach the water table. Because of this the Ifoghas hills are more suited to nomadic occupation than say, the sandy void (or ténéré in Tamashek) which fills the northwest corner of Mali. We crossed this ‘empty quarter‘ in 2006; we saw our last tree near Ouadane in Mauritania and the next as we neared Ikhalil, north of Tessalit, some 2000km later. Nor did we see another soul apart from our rescuers who met us midway. (As far as I know the remains of my Hilux – right – are still in Al Khalil).
One Chadian soldier in this Aljazeera video filmed after the Ametetai battle (see also Chad TV link top of the page) described the area as having ‘heavy tree cover’ which seems an exaggeration. Sure, there are small clusters of trees along the oueds, but more likely the sandy creek beds, huge granite boulders and low cliffs are what made hunting down the well entrenched jihadists so difficult.
East of the Adrar Tighaghar Google depicts a well-used track (see map below) which doesn’t appear on any old maps and which leads up towards Timiaouine over the Algerian border and continues south to Kidal, about 120km southeast of Aguelhok. When I last travelled down the full length of the Tanezrouft piste in the late 80s, Kidal was the site of a political prison and off limits, but in recent years it opened up to a few intrepid tourists, visiting the Essouk music festival, nearby. The Tuareg group, Tinariwen (left) are originally from Kidal.
This track would have developed in the more recent smuggling and kidnapping era as the Kidal region became busier. Doubtless many other tracks developed in the area as AQIM and later groups established themselves in this region over the last decade. With the long overdue Operation Serval, you’d hope that era may be coming to an end. So far in 2016, it doesn’t look like it.