Automation tracks Islamic State's IED use in Iraq
Automated processes that ensure the quality of intelligence inputs and improve the value of existing data have the potential to augment analytical processes-in this case, helping pinpoint changing trends in militant activity in Iraq.
A standard search of the IHS Country Risk database of intelligence events for improvised explosive device (IED) attacks in Iraq reveals a 70% increase in the number of incidents between January 2014 and August 2015, suggesting a significant change in the threat profile. IHS Country Risk maintains a database of more than 1 million intelligence events, which are metatagged and geocoded to provide structured analytics and geospatial intelligence to clients.
Further analysis of this dataset, using text-mining algorithms, disclosed that the number of events flagged as successful fell from an average of 80% of recorded events in 2014 to less than 50% of recorded events in 2015, as shown in the following map. Despite a marked increase in total IED attacks in Iraq, the growing proportion of unsuccessful attacks suggests at least some Iraqi security forces are improving their ability to intercept IEDs.
The success can be partially explained by coalition efforts to train and prepare Iraqi forces to better meet the threat represented by the Islamic State. In particular, Group Taji-part of the Building Partner Capacity (BPC) mission in Iraq, consisting of Australian and New Zealand forces-was deployed to assist in training Iraqi Army personnel in April 2015.
But since the collapse of Iraqi forces and the fall of Mosul in June 2014, the Shia militias of the al-Hashd al-Shaabi (Popular Mobilization Units: PMUs)-rather than Iraqi security forces-are positioning themselves as the only Iraqi force capable of countering the Islamic State. The coalition's BPC mission in Iraq does not train or support Shia militias currently operating on behalf of the Iraqi government.
Using automated processes to add geospatial metadata to the IED-related events revealed that, although the trend of increasing unsuccessful attacks is widespread, there are nonetheless substantial differences between provinces and cities, and particularly among predominantly Sunni provinces (Anbar, Salah ad Din, and Nineveh), mixed provinces (Diyalah, Kirkuk, and Baghdad), and predominantly Shia provinces in Southern Iraq.
Further querying showed the increase in the number of IED attacks flagged as unsuccessful between 2014 and 2015 in Iraq occurred largely in Sunni provinces. The percentage of attacks automatically flagged as unsuccessful in provinces identified as largely Sunni increased from 31.13% in 2014 to 49.06% in 2015. Combined with the significant increase in the number of attacks recorded, as shown in the following chart, this suggests an increase in unsuccessful or intercepted IED attacks in these areas during the period.
In the mixed provinces, the IED interception rate only marginally improved. This is unsurprising, given the nature of the conflict in the country and the deployment of PMUs, which have spearheaded offensives against the Islamic State in those areas. If this split had not been represented, it would have raised questions about the effectiveness of the automated process.
The majority of IED attacks in January-August 2015 took place in Salah ad Din province; 36.9% of total recorded IED attacks in Iraq occurred there, compared with 31.9% in Anbar province and 10.8% in Diyalah province. The Islamic State lost the majority of its territorial control in Salah ad Din province after being forced to retreat from Tikrit in March 2015.
Analysis of the automatically flagged events from January 2015 to early September 2015 showed that not only was Salah ad Din province the location of the highest number of IED attacks, it was also where the most IED attacks were flagged by the automated process as unsuccessful. In Salah ad Din, 67.6% of IED attacks were flagged as unsuccessful, compared with 31.67% in Anbar and 32.50% in Diyala.
Although this can be partially explained by the coalition's BPC mission in Iraq, the variation can be better understood when contextualized in terms of the Islamic State's own strategy to defend and expand its "caliphate" in Iraq, Syria, and elsewhere, and the differing IED tactics it uses in pursuit of different strategic goals.
In areas outside its direct control, such as Salah ad Din, the Islamic State relies on diversionary IED attack campaigns intended to demoralize the enemy and restrict military resources available for offensive operations against the militant group.
Videos published on social media of IED-clearing operations in Salah ad Din province and of interviews with personnel in explosive ordnance disposal (EOD) units indicate the Islamic State is relying on increasing the volume of deployed IEDs, as opposed to varying its IED manufacturing processes, in order to offset improving interception rates by Iraqi forces. For example, PMU media circulated a video on 1 September 2015 claiming EOD technicians dismantled 150 IEDs of the same type in a single day on a road connecting Baiji to Samarra in Salah ad Din province.
Dividing the data by city corroborates those findings, as shown in the following chart. In Ramadi, for example, the number of events recording IED attacks attributed to the Islamic State increased more than 400%, while maintaining a success rate close to 50%. On the other hand, in Samarra in Salah ad Din province, although the net amount of recorded IED attacks increased 800%, the success rate dropped from 100% in 2014 to less than 25% in 2015.
Julien Grossmann is analyst, while Meda Al Rowas is principal analyst, both with IHS Country Risk