Camber Corporation Provides
ArcInfo Support to Bosnian Peace Talks


Presented by

Robert W. Aldridge

at

1996 Esri User Conference

Wyndham Hotel and Convention Center

Palm Springs, California

May 20-24, 1996

Introduction & Background

The Bosnian Peace Talks held at Wright-Patterson Air Force Base just outside Dayton, OH successfully ended in an agreement initialed by Franjo Tudjman of Croatia, Slobodan Milosevic of Serbia, and Alija Izetbegovic of Bosnia on Tuesday, November 21 after three weeks of negotiating. The agreement, signed at a formal ceremony held in December in Paris, marks the end of 43 months of civil war in the former Yugoslavia.

Although there has been conflict in the Balkans for hundreds of years, the current war�s origins can be directly traced to the declaration of independence of Slovenia and Croatia from the former Yugoslavia in June 1991. Since this event there has been an almost non-ceasing war. The warring parties are made up of three ethnic/religious groups: the Serbs, the Muslims, and the Croats. The Serbs are Orthodox Christians who envision a �Greater Serbia� that would include Serbia, Croatian Serb areas, and Bosnian Serb areas. The Muslims are Serb descendants who converted to Islam after being conquered by the Ottoman Turks in 1389. The Croats, who are of the Roman Catholic faith, played a crucial role in the military of the Austro-Hungarian Empire. Although conflict has arisen from time to time between the Croats and the Muslims, they are currently allied against the Serbs.

51%-49%

Several UN brokered cease-fire agreements have been enacted and subsequently broken down since January of 1992. Finally, after several NATO air strikes on Serbian targets in late August and September of 1995, a shaky cease-fire deal was signed on October 10. The peace talks in Ohio were formally opened by Secretary of State Warren Christopher on November 1, 1995 to begin what was referred to by President Clinton as "the last chance we have for a very long time" to end the war. The chief U.S. negotiator, Assistant Secretary of State Richard Holbrooke, had already made some progress. A 51%-49% territorial split between the Muslim-Croat federation and the Serbian entity had already been agreed upon at a September 8 meeting in Geneva and this served as the starting point for these talks. Issues that remained unresolved included a truce and separation of forces, exact maps of the 51%-49% territorial split and the status of Sarajevo.

The line that would define the territorial split, which finally came to be known as the Inter Entity Border (IEB), had to be finalized and agreed upon before any peace accord could be reached. There were several factors involved in the development of the initial coarse definition of the IEB. First, there was the October 10th Cease-Fire Line (CFL). Second, there was the Federation�s proposed IEB. Third, there was the Serbian�s proposed IEB. And, finally, there was the Contact Group�s Line (CGL). The contact group was made up of U.S., British, French, German, and Russian negotiators. The CGL basically split the difference between the Federation�s proposal and the Serbian�s proposal. In theory, beginning with these four lines, a final IEB on a 1:600,000 scale map of Bosnia would be established. The next step would be to simply transfer this line to the 1:50,000 scale topographic maps that would serve as the official map documents for the peace accord.

Notice the "In theory" above. The IEB began as a product of the aforementioned four lines with trade-offs from one side or the other here and there, but quickly progressed to a beast that had very little resemblance to any of its predecessors. In the thick of the negotiations, new proposals, either from the Serbs, the Federation, or the Contact Group would be drawn and revised on an hourly (or even more frequent) basis.

Camber Personnel Arrives

Camber personnel arrived on the scene on Friday, November 10. This is the same day that MG Nuber, Director, DMA arrived and expected to stay one day to see how things were going. (He ended up staying 2 weeks.) This is also the same day that the presidents of Bosnia and Croatia signed an agreement strengthening the Bosnian-Croat Federation.

The first order of business for the new recruits, after we had located where we would be sleeping and what hours we would be working, was to decipher the naming convention being used to store the vector files that represented the various lines (cease fire line, contact group line, etc.). The file names in use may have been practical in the beginning, but by the time we arrived, they had become very confusing. For instance, a file representing the CFL that was to be plotted on a 1:600,000 scale map may have the name PT600-43 (PT for peace talks, 600 for the output scale, and the last two digits increased in increments of one with each new file created). The next file created would have the name PT600-44. The problem is that PT600-44 may be a line representing the CGL rather than the CFL. There was nothing in the file name that let the user know what was included in the file. The first task for Camber personnel was to come up with a naming scheme that turned chaos into organized chaos. Included in this task was to set up a single directory where the latest APPROVED lines would be stored. Therefore, if someone needed the latest CFL, CGL, IEB, Zone of Separation (ZOS), etc., there would be one place to find all.

A primary function of ArcInfo software and the Camber personnel running it was the calculation of territorial percentages based on proposed lines. The calculation of these percentages was performed by digitizing the proposed line, merging it with the country border, editing the digitized work and building polygon topology, coding the polygons, and calculating the percentage of land occupied by each polygon. The initial process involved registering the map to the digitizer each time a new line was brought in. Since most of the new proposed lines were all drawn on and would be digitized from the same 1:600,000 scale map, a master tic coverage was created that facilitated hasty registration of the map to the digitizer. This cut the percentage calculation time well in half. This made negotiators happy.

Another function performed by Camber personnel was the generation of buffer zones around the CFL and IEB. Both buffers were four kilometers wide, except in specific areas where it was adjusted to avoid roads or other features. One interesting modification to the four-kilometer IEB buffer was at the entrance to the Gorazde (a.k.a. Clark) corridor. Here the buffer was to be two kilometers from the center line on the outside of the corridor and one kilometer on the inside. Slight modifications were made on the inside buffer line as to avoid a critical road that connected Gorazde to the rest of the Federation�s territory. Similar buffer zone modifications were made around the city of Sarajevo.

While all of the lines were being proposed, modified, digitized, percentages calculated, modified again, digitized again, percentages calculated again, etc., the map production team was getting tasked to put these new lines on maps. This involved making sure that the currently approved lines were in the correct directory, produce a 1:600,000 scale plot, compare the 1:600,000 scale plot with the 1:50,000 scale index, and producing the 1:50,000 scale maps that had changes. The plan was for the lines to be finalized and final maps to be produced before the initialing ceremony. The final maps to be produced would include: a copy of the 1:50,000 scale maps (51 sheets) with the IEB, the IEB Buffer (ZOS), the CFL and the CFL Buffer; a copy of the 1:50,000 scale maps (51 sheets) with only the IEB and ZOS; a copy of the 1:250,000 scale maps (9 sheets) with the IEB, CFL, and their respective buffer zones; and, a copy of the 1:600,000 scale map (1 sheet). That�s 103 individual map sheets that had to be produced before the initialing ceremony. This proposed task became more and more unrealistic with each tick of the clock.

Other miscellaneous tasks for Camber personnel included digitizing proposed trade-offs, roads, rivers, etc. The proposed trade-offs were areas that the Serbs would be willing to give up in return for other areas from the Federation, and vice versa. These were digitized and the individual percentages of the total area calculated. During one eight hour stretch, over thirty proposed trade-offs were digitized and calculated. Roads and proposed roads were digitized for a number of reasons. First, existing roads were digitized to make sure that the buffer zones avoided including them. Second, existing unimproved roads were digitized to analyze potential improvement costs. And, proposed roads were digitized to analyze the cost for construction. Requests were made for other miscellaneous lines over the course of the talks. For example, Kevin was asked to quickly find the approximate average distance from the shoreline of a river to the ZOS line. To do this he digitized the shoreline and road from a 1:50,000 scale map, then he digitized a number of evenly spaced parallel vectors connecting the road to the river, then he used the STATISTICS function in ArcInfo to find the average length of the parallel vectors. This was a quick technique for approximating the average distance between the river and the road.

The talks are over. We�re going home. (Or, are we?)

Rumors of the talks ending with no agreement were running rampant throughout the day Sunday (November 19). By midnight, we were all told that the talks had ended unsuccessfully because no agreement could be reached on what to do with the territory around the Posavina corridor near Brcko. All three Camber personnel were up by 6:00 a.m. Monday feeling great and ready for a sit-down breakfast and maybe enjoy some of the points of interest in Dayton, such as the museum of flight. After having breakfast, we returned to the Operations Center. Things were beginning to buzz, and at 8:00 a.m. there was a call for two ArcInfo experts. We were running full blast again.

The mapping team worked throughout the day and night Monday and into Tuesday morning. Finally, a final IEB on the 1:600,000 scale map was agreed upon at approximately 11:35 a.m. Tuesday. The percentages were very close to what they wanted and minor adjustments could be made later. Several hand drawn copies of the final overlay were made. Minor modifications were made to ensure that the line was precise in areas of conflict, i.e. around Sarajevo, Gorazde, etc.

An initialing ceremony was scheduled for 3:00 p.m. Tuesday afternoon. The digitized line was to be overlaid on a 1:600,000 scale map and plotted by the Remote Replication System (RRS). The RRS was an elaborate system set up for rapid reproduction of documents and maps. A map was plotted by the RRS and delivered to LG Clark. On his way to the airfield to meet one of the dignitaries flown in for the initialing ceremony (I think he was going to meet Sec. of Defense Perry?), LG Clark found an error on the plot. The final IEB had been digitized in UTM projection, zone 33. Apparently, when the final plot was made (in haste), an improper projection transformation was used, or the digital image of the 1:600,000 scale map was incorrect. Either way, the line was supposed to run just south of the small village of Odzak. However, on the plot from the RRS, the line fell just to the north of Odzak.

We had to go to the backup plan. Of the several hand drawn overlays (drawn by Camber personnel), one was chosen and copies of the 1:600,000 map with the overlay were produced for the initialing ceremony. After all was said and done, a 1:600,000 scale "cartoon" map with the hand drawn magic marker line was used for the initialing ceremony.

Final Tweaking of Line

The final line which had been developed on the 1:600,000 scale map now had to be transferred to the 1:50,000 scale maps to ensure that it fell exactly where it was intended. The line was ported to PowerScene. PowerScene is a software package developed by Cambridge Research. Using intelligence imagery, high resolution elevation data, and scanned maps of Bosnia, PowerScene allowed the user to "fly" through the country. Beginning in the northwest corner of the country, all of the 1:50,000 scale sheets were examined and the line moved in order to avoid villages and to include the highest elevation along the line. Including the high ground was especially important around key roads and sites where snipers have been a problem. After adjustments were made on a map sheet, the line was ported back to ArcInfo where percentage calculations were performed to make sure that it was still close to the agreed 51%-49% territorial split. This process was repeated for every 1:50,000-scale map sheet.

Once the line had been checked and adjusted on all fifty-one 1:50,000 scale sheets, the buffer had to be regenerated and modifications made to it. After the buffer was finalized, production began on the 1:600,000 scale map, the 1:250,000 scale maps, and the 1:50,000 scale maps.

Lessons Learned

  1. A two star general can make things happen.
  2. A three star general can make more things happen.
  3. The accuracy and speed of ones digitizing is inversely proportional to how long one has gone without sleep.
  4. When you�re in a room with no windows and the clock reads 3:00, it doesn�t matter if its a.m. or p.m.
  5. Never underestimate the value of experience.
  6. There actually are only 24 hours in a day.
  7. PowerScene is COOL.
  8. There is a need for worldwide geographic education.
  9. Warren Christopher jogs at 6:00 a.m.
  10. If these numbers are right, we can end this war tonight.

PowerPoint Presentation