Issue 7.2, August 2003

Multicriterial Analysis Application in Mine Action

The author proposes how multicriterial analysis should be used to select areas where demining will have the strongest impact on risk reduction, return of displaced persons, and revitalisation of economic and social life.

by Damir Goršeta, CROMAC


The scope of mine action has been expanded to comprise demining, mine risk education (MRE), mine victims assistance and advocacy to ban AP landmines. Mine action resources are scarce; thus, they should be carefully targeted. The experience of mine-affected countries shows that a major part of mine-suspected zones are not mine-affected. In most areas, only a few mines may be found. Thus, it is important to exclude a non-contaminated area from the mine-suspected area as soon as possible. This will contribute to higher productivity and efficiency of demining. Besides the need to exclude a non-contaminated area as soon as possible, there is a need to determine which areas should be demined first. Locations and sizes of these areas should depend on end-users’ requests.

Croatia, like many other countries around the world, is contaminated with mines, perfidious weapons that go on killing a long time after a war has ended. Mine victims are mainly civilians, injured or killed many decades after the weapons have become silent. In view of the fact that the number of victims, scale of destruction and costs of solving the problem brought about by mines could be compared to a large-scale ecological disaster, mines might be regarded as weapons that leave behind the most horrible consequences. In addition to human suffering, mines bring about disastrous humanitarian, social and economic losses. The areas contaminated with mines become unavailable and fertile land remains uncultivated and useless.

Problem Description

Croatia is one of the 10 most mine-contaminated countries in the world. The realistic estimation is that about one million mines have been laid in Croatia. The mine-suspected areas are estimated to cover 1,700 sq km. About 170 sq km are covered with minefields, while the rest of the area is contaminated with individual explosive ordnance. Mine-suspected areas and minefields are located in 14 of 21 counties. Problems are caused by the fact that not all minefields are marked or consistent, and a number of mines have unknown locations, are not registered, have been mined by non-professionals or have moved due to weather conditions and erosion. There is also a big problem with a large number of UXO.

In view of problems caused by mines, it is clear that the main goal is to demine the Croatian territory as soon as possible, thus eliminating all risks. In the meantime, it is necessary to evaluate areas contaminated with mines, asking which demining approach will have the biggest impact on lowering the risk for human life, preserving material assets, encouraging refugees’ return, and revitalising the economy and social life while incurring the least costs and the most safety. Because of evident conflicts of interest and local political pressure, the intention was to develop a methodology that would make the decision process of the optimal policies for risk reduction more objective and based on real parameters. In order to achieve this objective, a pilot project for the most endangered county was launched with the intention of establishing effective methodology and proving the possibility for application in other areas.

Pilot-Project for Sisacko-Moslavacka County

Regarding available data and the actuality of the humanitarian demining problem, it was decided that the pilot project take place in Sisacko-Moslavacka County. Communities of the county would be treated as homogenous zones (areas) that would be ranged according to the agreed criteria. According to the available parameters on the area of Sisacko-Moslavacka County, 640 minefields were registered. By terrain surveying, as well as by identification of suspicious areas, a digitalised database was created containing all mine-contaminated and suspicious areas with 72 polygons on 11 communities in total. Regarding the fact that all aforementioned polygons were not homogenous and it was impossible to make them homogenous by applying some simple procedure, it was decided that being part of the certain community should be a criterion for polygon joining (i.e., for forming of a set of actions/projects that would be ranked, i.e., analysed by applying multicriteria analysis in order to determine an optimal policy for risk reduction). Such an approach is reasonable because communities are the smallest territorial and political units that are involved in the evaluation of optimal policies for risk reduction.

According to the project demands, and in order to ensure all relevant data and enable straightforward generating of more generic data, a geographic information system (GIS) containing various thematic layers was created. ArcView and some other Environmental Systems Research Institute (ESRI) tools that enable more complex spatial data analysis were used. During the analysis, the following problem characteristics were evaluated:

Within the project the objectives were:
Figure 1: Position of Sisacko-Moslavacka County in Croatia and Europe.

As the solving methodology the following compromised steps are worked out:

Due to the fact that during evaluation of the optimal policies for risk reduction, several groups are involved in the decision process, the activities in the process of problem-solving were defined:

Figure 2: Layout of the methodology for optimal policies for risk reduction in mine-contaminated areas. ??Layout of the hierarchic approach in demining operations in Croatia.??

Figure 2 shows a schematic procedure, which contains GIS analysis as a first step and evaluation of relevant criteria presented as thematic layers. For the criteria that can be spatially presented, using GIS analysis, concrete numerical values as input for multicriteria analysis, are being evaluated. For the criteria that cannot be generated by GIS analysis, an expert team evaluation and mathematical estimation are being performed. For example, by using data from mine records from both parties involved in the war conflict, it is estimated that on the territory of this county, 30,506 mines were layed, of which 24,877 can be identified on the already known minefields in eight communities (Table 1). The location of 5,623 mines is unknown, so the most likely solution is that they are posed on the territory of 11 mine endangered communities, or less likely on the territories of other communities in the counties that, so far, are not contaminated with mines.

Continue -->