Locating Landmines and UXO:
A Methodological Lesson from the Ethiopian Landmine Impact Survey
This issue may be outdated. Click here to view the most recent issue.
One of the most important parts about implementing effective Landmine Impact Surveys is first identifying which communities have a landmine/UXO problem. The author describes how Expert Opinion Collection (EOC) is used in Ethiopia to overcome the obstacles survey teams face when gathering their information.
by Michael L. Fleisher, Ph.D., Deputy Team Leader/Operations Manager, Ethiopian Landmine Impact Survey
In every country where a Landmine Impact Survey is being implemented,
those charged with conducting it are faced with the challenge of compiling
a reliable list of the communities that need surveying—because before a
community can be surveyed to assess the landmine/UXO impact, it must first
be accurately identified as a community having a landmine/UXO problem.
To make matters still more complicated, the country’s arduous terrain and poor-to-nonexistent roads make travel in the rural areas problematic in the best of times and, in innumerable areas during the long rainy season, all but impossible. Even in salubrious weather, many Ethiopian communities are accessible only by foot or by mule. The task of determining where the landmines/UXO are, so as to be able to formulate a complete, reliable list of landmine/UXO-affected communities for impact surveying, is not an easy one.
The technique developed for surmounting this problem, EOC entails gathering information from all available sources—civilian government administrators, military authorities, UN agencies, IGOs and NGOs, in-country scholars and so on—and using this information to compile a gazetteer of the affected communities to be surveyed. However, because time and resources are inevitably limited, survey teams are pressed to devise methodologies that will enable them to separate the landmine/UXO-affected communities from the non-affected communities as quickly and as efficiently as possible. To accomplish this, they must first devise a strategy that will enable them to determine where the most thorough, most reliable information regarding the locations of landmines/UXO may be obtained and how to acquire this intelligence as swiftly, systematically and cost-effectively as possible.
This article aims to facilitate this information-gathering process for all countries undergoing Landmine Impact Surveys by sharing some lessons learned by the Ethiopian Landmine Impact Survey (ELIS), which is currently being carried out by Norwegian Peoples Aid (NPA), under the auspices of the Survey Action Center (SAC), in Washington, D.C., and in close partnership with the Ethiopian Mine Action Office (EMAO), in Addis Ababa.
Ethiopian Survey Efforts
What is desirable in terms of both economy and efficiency, assuming it is possible, is to pinpoint an administrative level at which sufficiently thorough and accurate information regarding the presence or absence of landmines/UXO may be acquired. In the ideal world, we would hope to acquire all of the landmine/UXO data we needed at the national level without leaving the capital, but this ideal state of affairs does not exist in Ethiopia, nor has it existed in any of the other countries where Landmine Impact Surveys have been carried out. Experience has shown, in fact, that the highest, most inclusive administrative level at which generally reliable information may be found is the district.
In November/December 2001, an Advance Survey Mission (ASM) to Ethiopia, on behalf of the ELIS, identified approximately 5,000 of the country’s 26,000 kebeles as being possibly affected by landmines/UXO. Four months later, the ELIS survey team set about the task of establishing the number of affected kebeles more precisely so as to be able to quantify Ethiopia’s landmine/UXO problem as accurately as possible for the purpose of formulating a viable countrywide Landmine Impact Survey strategy.
In April/May 2002, ELIS field staff carried out a 17-day field assessment in Ethiopia’s Amhara region, one of the five highly suspect regions, with the following twin goals:
Requests and Responses
In response, the woreda administration provided the ELIS team with a kebele list. Here is the list provided for Wadla woreda, North Wollo zone (Figure 2).
Armed with such lists, the ELIS team conducted interviews with the residents of more than 50 kebeles, including those that woreda-level administrators had deemed to be landmine/UXO positive and others they had labeled as negative, in order to formulate as reliable an assessment as possible of the landmine/UXO threat in this area.
The RA strategy then employed was one of visiting and
conducting a group interview at every kebele that had been labeled either
as positive or suspect for landmines/UXO by its woreda
administration—except in those few cases where the distance and/or
relative inaccessibility of a kebele rendered an interview visit
impractical given the time constraints of this mission. The team also
visited as many non-suspect kebeles as possible, in an opportunistic
manner, in an effort to identify false negatives. Significantly, out of
the more than 50 kebeles interviewed, not a single false negative was
found. Although woreda administrators sometimes identified a kebele as
positive for landmines/UXO, only to be contradicted by kebele
interviewees, they never once identified a kebele as negative for
landmines/UXO that turned out to be positive. (It is false negatives that
pose a serious threat to the integrity of the ELIS, not false
positives—because a false negative, if left uncorrected, represents a
threat to the lives of people endangered by landmines/UXO in their
communities, whereas a false positive will invariably be surveyed and, as
a consequence, changed to a negative.)
The ELIS assessment also produced valuable evidence that information collected at the woreda level is of satisfactory reliability and specificity to ensure the ELIS’s requirements of thoroughness and accuracy. Although administrators at the woreda level occasionally erred on the side of pronouncing a kebele in their area of administration to be definitely or possibly landmine/UXO-affected and it turned out not to be (i.e., this kebele turned out to be a false positive for the woreda administration), they never erred on the side of pronouncing a kebele to be free of landmine/UXO contamination and it turned out not to be (i.e., the kebele administrations produced no false negatives). This is an extraordinarily positive sign.
Further assessment work still needs to be done, however, to determine whether the results derived for North Wollo also hold true for Amhara region’s other 10 zones—and also to gauge the landmine/UXO threat in Ethiopia’s four remaining highly suspect regions as well as in the six other regions that are so far not suspect.
This assessment has had (and will continue to have) profound implications for the ELIS. Had the worst-case estimates of the ASM proven accurate, some 5,000 kebeles, and an estimated 10,000 gotoch within those kebeles, would have had to be surveyed, a task of such magnitude that it would have exceeded the capacity of the ELIS team to carry out the survey within existing time, money and manpower constraints by a wide margin. Our hopes when we planned the North Wollo field assessment were the following: