SURVEYING WORLD HERITAGE ISLAMIC MONUMENTS IN NORTH AFRICA : EXPERIENCES WITH SIMPLE PHOTOGRAMMETRIC TOOLS AND NO PREVIOUS PLANNING

Different experiences of surveys of Islamic monuments from different sites of Tunisia, Algeria and Morocco are presented. They have been made with simple tools: one photographic camera and a laser meter, without a previous planning or prevision for the survey, profiting from visits organized during scientific meetings to which the author was invited. Some of these monuments belong to sites included in the World Heritage List, but no metric documents or only low quality information is available. Monumental Almohad gates from Rabat and Marrakech, the al-Badi palace of Marrakech, the minarets of Mansura and the Qala of Beni Hammad, the dome in front of the mihrab of the mosque of Tlemcen are some of the examples to be presented. The methodology applied is based on ideas and tools acquired in CIPA meetings proving the usefulness of these encounters but supporting the idea that “providers” should provide tools and methods and “users” should be responsible for documentation, never missing the opportunity of acquiring knowledge from the heritage during the survey process.


INTRODUCTION
The need for having good heritage documentation, especially in emerging or developing countries is being stressed during the political events that have occurred recently in the South and East Mediterranean areas, like the so call Arab Spring and the Syrian civil war.The destruction of cultural heritage, which sometimes are direct targets and other times collateral damage, deprives not only the involved societies but all of humanity from irreplaceable historical and cultural values.This is always an irreparable loss, which may only be minimally alleviated if at least there is a documentary record.When destruction is only partial, the documentation can make the restoration easier.
Unfortunately, in many of these countries the availability of proper documentation is very limited, especially metric, because of the difficulties involved in its elaboration.As a general rule, only technical drawings are made when restoration intervention is planned, and yet do not always have adequate quality.The need for valid and useful documentation is therefore evident and should constitute a priority action.
For years, the experience implemented at international level and shown, among other places, in the CIPA Symposia offers interesting solutions that allow us to glimpse the possibility of responding, at least partially, to the problem.We will show in this paper a number of experiences of surveys conducted on architectural heritage of Islamic origin in several countries of North Africa, mostly pertaining to property inscribed on the World Heritage List.

INSTRUMENTS AND METHODOLOGY
These case study surveys have a common set of circumstances that are of interest as experiences using a methodological approach for the urgent documentation of the heritage of countries as the above mentioned.First, they have been carried out with-out advance planning and therefore without a previous study of the building and its characteristics, which proves the versatility of the systems used and their easy adaptation to different circumstances.In all of these cases the purpose of the visit to the monument or site was not to perform documentation but simply a visit for a rapid architectural analysis of the buildings.Secondly, the instrumentation used in data collection has been extremely simple, in the majority of the cases a simple camera and an instrument for measuring, usually a Leica Disto.These are easy to transport and relatively inexpensive, not much more than what a typical tourist carries on.In some cases, the documentation was captured in such a discrete fashion that it was not evident that a survey was underway, which occasionally may be convenient to avoid suspicion.
Finally, the laboratory instrumentation is also relatively simple.The different case studies that are presented also show the evolution of the technology over time while the methodology can be considered unchanged.In the first example analytical plot-  This contribution has been peer-reviewed.The double-blind peer-review was conducted on the basis of the full paper.13 ters were used with very sophisticated software and complex management.Fortunately, for the most recent cases the instrumentation was not only less expensive but much simpler to use thanks to the improvement of software and the use of digital photos.
In The methodology followed is based on the use of Photogrammetry as ideal means to acquire large amounts of information quickly and easily.The use of digital photogrammetry allows greatly reduced instrument cost.Three restitution systems were used.Bundle adjustment of blocks of photographs, whose first principles are known from late Prof. Wester-Ebinghaus in CIPA (Kotowski et alli. 1988), allows the simplified acquisition of control data, even dispensing with the use of topographic instruments.ORIENT (Kager 1980) and MAAS-CR constituted the first software that was used, but Orthoware software now appears very versatile, reliable and user friendly, as it allows working on the measurement and the computation interactively, rather than deferred, as in older programs.The principles to operate with this procedure have been systematized and disseminated by Prof. Waldhäusl with the 3 x 3 rules (Waldhäusl & Ougleby 1994) that were basically followed, although not always with the thoroughness presented due to time limitations.Through a set of photographs and a few simple measurements and observations of the object (some distances between visible points and references of planes or lines that can be considered vertical or horizontal) a set of points was obtained with threedimensional coordinates in an established local coordinate system.
The orientation of the images block using a reduced control can raise doubts about the accuracy achieved, by the absence of check data.However, verification and calibration were tested in other projects.Recently the procedure was used to document and survey the main tower of the castle of Villena (Alicante, Spain), a building with square plan of 14 m side and 30 m of height.In addition to the photographs a number of points were measured with a total station with laser EDM (Electronic Distance Measure).In the absolute orientation of the block a first test was performed using only the distance between two points and the consideration that an impost that runs along the facades defines an horizontal plane.The result was fully valid in relationn to the size, shape and overall dimensions of the tower.The impost, however, is not perfectly horizontal, as on one side there is a difference in level of 5 cm which introduced minor errors of absolute position of the measured points resulting in maximum discrepancy of 6 cm.Using three control points measured with the total station for comparison, the maximum discrepancies between the measurements and those made with Orthoware were only 3 cm which proves the reliability of the system, demonstrating that work can be rapidly conducted without the use of surveying instruments with perfectly valid results for most purposes for which these surveys are used.
Final graphic documents, which essentially are vector drawings, were obtained in most cases through stereoscopic restitution.This procedure allows an observation and analysis of the object very close to that achieved with direct vision, enabling selection and synthesis during drawing for excellent condition results and with remarkable speed.We have always defended the importance that this process has for a better knowledge of the building or object that is being documented, which constitutes the ultimate goal of all survey (Almagro and Almagro-Vidal 2007).
Today the use of the analytical plotter SD2000 continues, although almost all the works shown here have been made with VSD of AGH (Jachimski 1995) and increased use of the software PoivilliersF.The latter, developed by Yves Egels, is a very simple and easy to use software, which works in two modes: with anaglyph glasses for simpler hardware (laptops) or 3D glasses like NVDIA 3D Vision, today easily accessible thanks to the development of 3D computer games.
When more information needs to be introduced such as textures and details photographic rectification software is used for all the elements that exist in a flat plane, very frequent case in architecture.The software, ASRix by Steve Nickerson, allows photos to be entered directly into AutoCad 3D drawings in a simple and accurate way.The instruments exist and are easily affordable, not only for a Government Office, but for an independent professional.It is only required to know how to use them and for this reason training is needed, and thus should be emphasized as a priority.

CASE STUDIES
The case studies range across North Africa from Tunisia to Morocco and Algeria.They were selected as they represent the full versatility of the methodology presented here and the evolution of the techniques available at that time.Tunis, Marrakech, Rabat and the Qala of Beni Hammad are inscribed in the World Heritage List.
Qoubba Bou Khrissane, Medina of Tunis, Tunisia.This funerary dome of the 11 th century, located in the Medina of Tunis was the subject of a practical case study in the Course of Tunis    This contribution has been peer-reviewed.The double-blind peer-review was conducted on the basis of the full paper.15 Dome in front of the mihrab in the mosque of Tlemcen, Algeria.This masterpiece of the Almoravid art from the end of the 11 th century lacked a metric survey, despite having been studied and published in numerous scientific books and publications.The documentation work was one of the most complexes, since when the photos were obtained the goal was not orienting a complete block, but only separate stereo-pairs to plot some building elements.After obtaining more advanced software, it was possible to make the full suvey and this has allowed the creation of a 3D model of the space beside the mihrab.In this case, the absolute orientation was made considering that the cornice where the dome is supported defines a horizontal plane.
Fifteen pictures were used for the block and 4 pairs for the restitution of the drawings.The photogrammetric survey was completed with manual measurements of areas which could not be recorded with photographs, such as the space between the dome and the roof.The façade of the prayer room into the courtyard of the mosque was also plotted and the next step will be to complete the survey of the minaret.(Golvin 1957: 185).This shows a significant magnitude of the error in the height of the tower from that drawing, of more than 1.5 m.This is the only documentation that previously existed for this important cultural property and demonstrates the lack of the appropriate accuracy.

CONCLUSIONS
The experiences presented here constitute good examples of the possibilities offered by Photogrammetry with the current state of the art of this technique; the instruments and software are fairly low cost and easy to use so they can be learned without too much difficulty by anyone who already possesses basic knowledge on forms and modes of representing historic architecture.It is more difficult to learn the nature, characteristics and importance of architectural elements than the photogrammetric documentation technology.To understand and thus properly document cultural heritage requires basic manual drawing skills, a capacity for synthesis and an understanding of history, values and artistic styles.Attempting to circumvent this knowledge by using automated systems can result in serious deficiencies in the quality of the surveys.
The limits for these systems are related to the size of the building and thus, with the quality (detail and resolution) of the pictures we can obtain, and finally, with the possibility of geting pictures according to the 3 x 3 rules.A characteristic case like a tower that can be photographed from the four sides and the four corners could be performed easily by people without much experience.More complex cases like the ones with interior and exterior areas and limited links between them should need more experience specially to solve problems when the orientation does not converge at the first try.Even if we do not have any experience yet with very large and complex buildings, we estimate that in such cases a basic control survey with topographic instruments will be needed to avoid too large errors.
In any case, the improvement and dissemination of these systems and methodologies should be a priority in order to provide managers and technicians involved in the conservation of cultural heritage appropriate instruments for documentation.This is especially true in those places where the social and political circumstances threaten the conservation of a heritage of incalculable value.

Figure 1 .
Figure 1.Set of photos used in the survey of the Qoubba Bou Khrissane in the Madina of Tunis

Figure 2 .
Figure 2. Elevation and 3D model of the Qoubba Bou Khrissane in Tunis

Figure 6 .
Figure 6.Elevations and 3D model of the western pavillion of al-Badi palace in Marrakech.

Figure 5 .
Figure 5. Elevation of the interior gate of Bab al-Kebir in Rabat

Figure 7 .
Figure 7. Block of photos, bundle adjustment and 3D model of the North hall of the palace al-Badi in Marrakech.

Figure 8 .
Figure 8. Cross section and 3D model of the dome beside the mihrab of the great mosque of Tlemcen

Figure 9 .
Figure 9. Elevation and 3D model of the minaret of Mansura