Evoking the risk, even in urban areas, brings us back almost mechanically to the risks of flooding, or even silting up in certain regions. However, there are many other risks in cities that are still little or no explored using GIS solutions. We will mention in particular the problems encountered in the outskirts of cities and so-called sensitive areas, but also those due to the very configuration of certain neighborhoods (concentration of buildings, absence of support infrastructure) which generate risk situations. , or rather danger. Of course, we cannot overlook the problem of old neighborhoods that are abandoned in certain regions, but where populations still live and are exposed to multiple dangers following the continuous dilapidation of buildings. Finally, it is useful, even necessary, to also mention the perception of risk, which is sometimes extremely vague, in the minds of populations who end up generating their own risk themselves.

In this article, we briefly examine the notion of risk, its perception in the collective mind and its consequences in everyday life. We then present, on a practical example, the contribution of GIS solutions and satellite imagery to the analysis and forecasting of risk in urban areas. The examples are taken from Algerian towns in the pre-Saharan zone. We will take the city of Laghouat as an example and we will develop the aspect of risk generation by the individual and society as a whole; we will also see how the local authority contributes to this situation of generation of risk or danger and, above all, how it can remedy it.

Keywords: urban risk, GIS, satellite image, grid analysis

Risk, danger, “being vulnerable”, provoking or undergoing danger are all relatively very vague terms in popular consciousness in many regions of developing countries and, in particular, in the Maghreb areas; this vagueness of assessment is even more marked in rural areas where the literacy rate is still quite low; however, it should also be noted in urban areas where populations with very varied literacy rates coexist. This situation is further accentuated by rapid economic development, by the introduction of lifestyles and consumption patterns which have completely upset the functioning (and therefore the perception of the environment and the other) of societies accustomed to living according to rules both rigid, very protective and particularly "simple" to apply and understand. As our objective here is to try to introduce a tool for measuring (or mapping) risk, it seems important to us to review, clarify and (even) complete the basic theoretical principles on these notions.

Rather than defining hazard, several authors define vulnerability as the measure of the effect of an exposure factor on an individual¹ while that resilience is the ability of this individual to recover from a disturbance to return to a state of stability or equilibrium with the exposure or risk factor.² It seems important to us to note here that this capacity of the individual (or of a group of individuals) to recover from an external disturbance strongly depends both on the "popular" consciousness of the group, on its cultures and traditions. (in terms of lifestyle) and also the rules established by “modern society” to integrate technological, economic and societal developments.

To this end, Moser³ introduces the relationship between vulnerability and economic changes and defines, in this relationship, the poverty of urban populations by taking into account several factors such as work, human capital, production goods, and social capital and intra-household relations. These different factors are assembled into three large groups of factors that make it possible to differentiate between urban and rural areas. These are the marketable nature of certain goods or “moneyability”, environmental risks and social fragmentation. From these three main groups, she defines vulnerability as "the insecurity and sensitivity in the well-being of individuals, households and communities with regard to a changing environment and, implicitly, their capacities to respond or to adapt ". But it should also be noted here that this significant difference between traditional urban and rural environments is increasingly blurred with the progress of the means of communication, both television (increasingly present) and the Internet or mobile telephony, which tend to homogenize, to smooth our ways of life, our gaze on the Other but also our revolt in view of the more visible and more fragrant inequalities. However, with social mixes appear other differences that are both more marked and more difficult to delimit: center-periphery differences, differences between social classes, generational differences.

It thus appears that vulnerability and resilience are determinants of health, in the broad sense, and of the quality of life of an individual or a population; they are assessed using quantifiable indicators (such as an exposure or the risk associated with it) and non-quantifiable indicators (such as a perception or practice related to an exposure). But their perception by “at risk” populations, even before their evaluation, is neither obvious nor simple; it calls for special attention if we consider the seriousness of the situation.

The analysis of the evolution of the urban landscape in a pre-Saharan environment has been studied by several authors; we will cite for the case of the urban area of ​​Laghouat the work of Abdellaoui et al.³ and those of Benblidia et al.⁴

The various surveys and investigations that we conducted during this project lead us to make the following preliminary remarks:

1. Since danger is present throughout daily life, it is trivialized and becomes a “normal” element in daily life and inseparable from the context and everyday life; it is no longer felt as an “anomaly”; thus, a trunk or a barrel on the road is quite naturally circumvented! if a person moves them, it becomes a singular point.
2. When you draw someone's attention to a danger, they are first "surprised" and then often leave it to fate: this wall has never fallen so far, why would it fall now? In Figure 1, a middle-aged man sits quietly in his chair to pass the time next to a wall that is in danger of collapsing at any moment. The danger is there, threatening; our character does not even notice it; if you "awaken" him from his torpor, he will give you a great lesson in rhetoric or theology or philosophy according to his convictions and his culture. In addition, his speech will be all the longer and more complex the less “cultured” he is. There is here a very strong part of affectivity, which prevails over “the reasonable and the reasoned” which makes any attempt at logical demonstration doomed to failure.
3. Most often, we go back to the temporal, to belief: destiny. In addition, this is not the prerogative of the less “literate”. The weight of tradition and social pressure helping this reaction will be present both in those who have never been to school and in the "intellectual". Figure 2 shows a woman accompanied by her daughter; on their way, on one side the danger of the road, on the other a pile of stones that constitutes a permanent danger. The woman is unaware of either danger; she may nevertheless be afraid, possibly, of cars; but what would she say? Is there another way to reach my goal?

Figure 1 Popular perception of danger.

Figure 2 The child and adult facing danger.

Figure 3 shows in (a) and (b) the absence of signs for obvious and long-standing dangers on the roadway (in (a) danger for both vehicles and pedestrians) and on a "pedestrian" lane » practically cut by rubble from the house which is in ruins and which will not fail to continue to fall, thus constituting a permanent danger neither taken care of nor even correctly reported. The surrounding population obviously continues to cross by bypassing and to live nearby.

Figure 3 Danger in public road.

Figure 4 highlights the illegal appropriation of space and the fabrication of permanent danger without the perpetrator being in the least worried; the passers-by themselves become accustomed to the danger they incorporate into their movement and behavior to such a point of trivialization that it virtually disappears, no longer shocking and existing only in theories of the management of the space.

Figure 4 Appropriation of space and generation of danger.

Many other examples can be given. Children playing in the middle of the road, in the middle of a bend, a little girl going to school on a path outside the inhabited area, the market which extends into an area where the habitat is in a state of disrepair advanced, a dangerous deterioration of the road does not attract the attention of anyone and even less of the Authority, the little boy at the tobacco kiosk (in Figure 5), the boys who throw stones unaware of the danger that they cause (Figure 6) … are so many dangers that no one pays attention to.

Figure 5 Ordinary tobacco: the most pernicious danger.

Figure 6 Child games.

The question then is how to remedy this situation. If the solution is almost obvious (information, awareness, training and education), it will be necessary to understand that the effects of this type of action are only very long term.

1. The variables

A commonly accepted definition of natural risk is the equation:

Natural risk = natural phenomenon causing damage × vulnerability (1)

In this definition, the natural phenomenon represents the threat and is expressed by a field of action, a magnitude, an intensity or a flow, a violence (impact) and a recurrence (frequency). Vulnerability is part of a system that encompasses bodily and moral harm to people and potential damage to exposed elements.

Vulnerability can be considered as a system, articulated around a large number of variables, natural and human, whose dynamics in time and space can generate more or less dangerous situations for an exposed society. It is thus part of a system that encompasses bodily and moral injury to people and potential damage to exposed elements (product assets, socio-economic activities and heritage); Thouret JC & Ercole R⁶ In urban areas, this system finds its roots in the very functioning of the city or outside of it.

It therefore seems difficult to distinguish vulnerability linked to natural phenomena from vulnerability resulting from specifically urban social phenomena. On the basis of this remark, we modify equation (1) by focusing more on risk by including both natural risk, "anthropogenic" risk in the classic sense of the term and "non-material" risk (or not materialized ) generated, almost unconsciously, by a population (or a group of individuals) by the very fact that it does not perceive this generated risk:

Risk= phenomenon generating damage × vulnerability × perception (2)

The phenomenon can then take one of the following three forms: natural, anthropogenic (structure, industrial unit, risk-generating project), non-material (person crossing an express road, for example).

2. Definition of variables

We are interested here, by way of illustration, in a natural risk (risk to public health due to sandstorms) and an anthropogenic risk materialized by accidents on the road and caused, in large part, by the perception of risk.

The damage-generating phenomenon will be assessed here by its intensity on a scale of 0 (minimum value, almost non-existent phenomenon) to Phmax (very strong phenomenon); this value includes both the four components of the phenomenon, namely magnitude, intensity, impact and recurrence; vulnerability (measurement of the effect of an exposure factor on an individual) will also be assessed on a scale from 0 (minimum value, zero vulnerability) to Vvmax (maximum vulnerability); perception will also be assessed on a scale from 1 (very high awareness of danger) to Prmax (no awareness of danger). The Phmax, Vmax and Prmax values ​​condition the precision of assessment of the various parameters and therefore the analysis precision; it is certain that the greater the precision requested, the more difficult the interpretation of the results will be (large number of output classes, greater probability of redundancy). To simplify the analysis and illustrate this work, we consider a phenomenon coded on a scale from 0 (non-existent) to 2 (strong); vulnerability will also be coded by three values: 0 (population not vulnerable), 1 (low vulnerability population) and 2 (high vulnerability population); finally the perception will be assessed only by two values ​​0 (population very aware of the danger and therefore possessing the means to protect themselves from it, therefore to significantly reduce the exposure situation) or 1 (total unawareness of the danger, even "voluntary" exposure but unaware of the danger). These various situations and their combination are illustrated by Tables 1& 2 obtained by two different processes. In Table 1, the combination of factors used to determine the risk is obtained by concatenating the values ​​of the variables. In Table 2, this combination is obtained by algebraic operation. We note that under the same working hypotheses (variable value scales) the “concatenation” method makes it possible to obtain better discrimination of situations; in Table 2, ambiguities of interpretation appear; the result is however relatively correct in a first approximation.

3. The division of geographic space

The spatial reference (or division of geographic space) is always chosen according to the problem of the study and the objectives to be achieved. Two ways are available to us: an “administrative” division as fine as possible according to the precision of analysis that we wish to achieve; we can thus choose as the finest element the iris or the island and we then work in vector mode; this division does not exist in the Maghreb (no standardized coding, no data available on this scale); the breakdown by municipality is completely inappropriate if we are interested in risk mapping in urban areas, especially for municipalities of around 100,000ha. We then adopt a mesh division, which seems to us the most suitable.

A mesh is the generating element of an isotropic division having the same characteristics (shape, size, surface) in all directions regardless of its location. Software currently allows the creation of square or hexagonal meshes. For the present work, we opt for the square mesh. Landscape analysis using the grid method has been studied by Abdellaoui et al.,⁴ Huzui et al.⁷

1. The study area

The study area is located in the southern foothills of the Saharan Atlas of Algeria and essentially includes the university and hospital district north northwest of the city of Laghouat. The study area is shown in Figure 7. The main orographic elements are represented by Djebel Ahmar (883m) SW, Dj. Oum Deloua (1023m) to the NW, representing a perched syncline, and a little more to the NW of the latter stands the great syncline of Milok, in the Atlas direction. The main Wadi that crosses the area is O. M'zi with a NW-SE orientation, then takes a global sub-latitudinal direction, from the SW end of the Dj. Oum Deloua. Its most important tributary is O Messaad, with a SW-NE flow. This area includes different shapes and formations sandpits whose layout depends on the orographic units, the detailed topography and the directions of wind transport.

Figure 7 Location of the study area.

The population of Laghouat, a pre-Saharan town, increased from approximately 6,000 inhabitants in 1830 to 42,800 inhabitants in 1977 and to 119,043 inhabitants in the year 2003, i.e. a variation of 178% (i.e. a multiplication by a factor of 2.8) between 1977 and 2003 (26 years). Indeed, the extension of buildings, under the pressure of growing housing needs, is currently reaching the natural limits, very close to areas of sand accumulation. Sand is thus a permanent and increasingly felt risk factor for both public health and quality of life.

Furthermore, this exceptional demographics, combined with the increase in the standard of living and the plan for bank loans at preferential rates, in particular for the acquisition of private vehicles, have led to a disproportionate increase in the flow of traffic on the roads and, as a result even the risk of an accident. This justifies our choice of these two risk-generating phenomena.

2. The grid approach

For the present analysis, we opt for a division of the landscape into square meshes of 150m; this mesh dimension represents a good compromise between the scale of representation of the urban space highlighting homogeneous sectors of built-up areas and the size overall grid. Figure 8 gives an overview of the divisions with grids of 500, 250, 150 and 100m. The assignment of the values ​​of the various parameters for each mile is carried out by drawing inspiration from knowledge of the field, surveys carried out from a sample and surveys carried out in Agadir as part of a doctoral thesis⁸ on the city of Agadir.

Figure 8 What size for the mesh?

3. Risk factors

To assign the values ​​to the various variables on each mesh, we use the hypotheses set out in paragraph a. For the "risk awareness" parameter, we have decided to consider here only two modalities coded 1 (good awareness of the risk; the amplitude of the generating phenomenon is not modified; on the other hand, this awareness of the risk is taken into account in the value of the vulnerability factor) or 2 (no awareness of the risk, the effects of which are then amplified). Figure 9 shows by way of illustration the assignment of values ​​for the “generating phenomenon” and “vulnerability” factors for the road accident risk.

Figure 9 Assignment of parameter values to cells: on the left, assessment of the phenomenon, on the right assessment of the vulnerability.

We are particularly interested here in the section of national road 1, particularly in terms of its proximity to the university which drains a population of nearly 30,000 people. Admittedly, this is a population with a relatively high level of education for which one should assume that awareness of the danger is high. However, the unconscious calculation of the shortest path, the weight of the "popular" conscience, habit mean that the footbridge overhanging the road is very rarely taken. These remarks lead us on the one hand to decide on the value 2 for the “awareness of the danger” factor (in other words, population without awareness of the danger) and to overestimate the vulnerability.

For the "sand wind" phenomenon generating the health risk, we have essentially concentrated in this "methodological" article on the dwellings built in the immediate vicinity of the areas of sand accumulation on the reliefs and which thus constitute the most exposed part.

4. Risk mapping

Figure 10 summarizes the map of the two risks: public health risk due to frequent sandstorms in March and April in this region.

Figure 10 Mapping of health risks related to sandstorms and traffic accidents.

It is obviously not surprising that the populations most at risk in terms of road accident risk are those who live (or work)

5. The responsibility of the local authority and the measures to be taken

Figures (3) to (6) are significant of the great responsibility of the local Authority in the permanent control for the respect of the rules and procedures. The photo on the left of figure (3) shows that the competent services should put up a warning sign of the danger, especially at night; for the photo on the right, we were told that the owners of the house tried many times to alert the competent services and had submitted a request for clearance which was never followed up.

Figure (5) is very significant of the trivialization of the danger and the inaction of the services of the local authority. On the left, we see a young boy in front of the cigarette kiosk; of course, the sale of cigarettes cannot be prohibited; but we must do more checks to prohibit it to minors and, if necessary, close the kiosk that recurs.

There are now drastic measures to be taken: the implementation of societal awareness campaigns. For this we should use all the channels available or to be created:

• The school in the first place, of course, to raise awareness among minors; one could imagine, for example, bringing in the classes at certain times in agreement with the education department, firefighters or personalities recognized for their commitment to education and societal awareness actions.
• The mosque in second place by asking imams to devote time to social awareness in their sermons; a type of speech could be prepared for this purpose.
• Finally, by inventing a tool for raising societal awareness of the concept of risk and danger in general and the production of danger in a particular way:

1. Display in bus stops, certain public places, motorway stations
2. Car with loudspeaker broadcasting societal awareness messages that should be prepared with great care!
3. Social awareness spots on local radio

This work has enabled us to emphasize the perception of danger by populations in developing countries; this perception, which is difficult to measure because it is essentially subjective, has a very strong influence on the assessment and mapping of risk and vulnerability in urban areas (pre-Saharan). However, we were able to introduce an assessment of the risk factor generated by the individual, by society or even by the absence of the local authority in the permanent control. Finally, we have proposed solutions for raising societal awareness of the concepts of danger and risk. We have also introduced mesh slicing for vulnerability and risk mapping; this approach is still little used in landscape analysis. However, it offers many possibilities for the apprehension of space. Its generalization still lies in the difficulty of automatic assignment of the values of the variables to the various cells.

None.

The author declares that there are no conflicts of interest.

None.

1. World Health Organization: WHO Preamble adopted by the International Conference on Health, New York. Actes officiels de l'Organisation mondiale de la Santé. 1948;2:100.
2. Abdellaoui A, Rougab A, et al. Characterization of buildings in the mountain-plain transition zone: case of the Blidene agglomeration (Algeria). Remote sensing of Urban and Peri-urban environments. 1997;75–83.
3. Moser CON. The asset vulnerability framework: reassessing urban poverty reduction strategies. World Development. 1998;26(1):1–19.
4. Abdellaoui A, Visan Livia, Patru Stupariu GH, et al. Étude de la viabilité du paysage par analyse de grille dans la région SousCarpatique de la Vallée de Prahova (Roumanie).  Revista de Géomorfologie. 2010;12:81–90.
5. Benblidia N, Abdellaoui A, Bensaid A, et al. Utilisation de la morphologie mathématique pour l’analyse de l’occupation de l’espace en zones urbaine et périurbaine présaharienne; cas de Laghouat (Algérie). Revue Télédétection. 2006;6(2):177–190.
6. Thouret JC, Dercole R. Vulnérabilité aux risques naturels en milieu urbain : effets, facteurs et réponses sociales. Cah Sci Hum. 1996;32 (2):407–422.
7. Huzui AH, Abdellaoui A, Patru Stupariu HG. Analyzing urban dynamics using multi-temporal satellite images in the case of a mountain area, Sinaia (Romania). International Journal of Digital Earth. 2011;1(17):1–29.
8. Atik M. Unsanitary housing in Greater Agadir (Morocco).thèse de doctorat; soutenue le 7 décembre 2009; Université Paris 12. 2009.