Accessibility Index of Aquatic Environments as an Indicator of Surface Water Vulnerability in Urban Areas: Case of the Okpara Basin (Benin)

Physiographic differences and conditions of use of water resources in anthropogenic basins explain the variability of risk to surface water. Based on the multi-criteria analysis of Saaty, the present work proposes an assessment of the vulnerability of surface water through a three-factor accessibility index: the slope, the subdivision rate and the drainage density of the basin. It is observed that the topographic (slope) and urbanistic (subdivision) conditions are the most important (weighting of 59%, 34%) in front of the hydrographic condition (drainage 6%) with an overall consistency of 2%. Thus, in the Okpara basin in the city of Parakou, the analysis of the vulnerability of surface waters by the accessibility index highlights two groups. The Ganré and Kokouro basins with about 30% of their areas in the low and moderate classes and the Wonka and Dama basins in the very strong class at more than 50% of their areas.

sources; especially for developing countries, considering the accelerating pace of demographic, urban and socio-economic growth recorded at the beginning of the 21st century [2]. According to [3], the renewal of surface water is influenced by rain runoff and controlled by ecohydrological specificities and as a result, uneven distribution on land. When referring to global statistics [1], the seriousness of the issue of access to water resources is better understood both globally and regionally [4]. In the context of rapid population growth coupled with the difficulties of access to land in developing countries, the peri-urbanization of ancient nuclei is very spontaneous and generally occurs in the absence of spatial planning [5] [6]. Years ago, several studies have attempted to evaluate the influence of uses, especially anthropogenic uses on the quality of water in the resource [7]- [12]. Within this framework, many methods have been used to assess groundwater vulnerability [13] [14] [15] [16] [17], but very few are interested in surface water [15] [18]. This apparent lack of interest is due, among other things, to the complexity of the surface water pollution mechanisms, on the one hand, and the active involvement of the geomorphometric specificities of the basins on the other and in addition to the rather well-known role of anthropic activities. The hydrographic system of the city of Parakou (West Africa) is divided into the basins of Upper Ouémé and that of Okpara. According to the administrative division of the city and the demographic statistics [19], nearly 70% of the urban slice of the city is found in the Okpara basin. The geographical boundaries of the city have moved quite rapidly over the last two decades with a gradual and monocentric occupation to the north-east and south-west of the Okpara Basin. It follows from these observations that the aquatic systems of this basin would be exposed to a greater risk of pollution due to the phenomenon of urban sprawl. The aim of the study is to evaluate, using Geographical Information Systems (GIS) tools and multicriteria analysis, an index of accessibility of hydrographic networks as an indicator of vulnerability of surface waters in the Okpara basin at Parakou.

Study Area
The work is done on the Okpara basin in the city of Parakou. The Okpara basin is about 235 km 2 in total, but consists of four sub-basins of small size in the town of Parakou located between the latitudes North 9˚15' and 9˚27', then longitudes East 2˚31' and 2˚45'. It presents a four-level fishbone hydrographic network according to Strahler's classification. The city is densely populated, strong in areas of ancient settlement (the urban core) and decreasing in size to peripheral areas.
The terrain is sloping North-South and North-East. The thalwegs of the four sub-basins (Ganré, Wonka, Kokouro and Dama) are positioned as the main transport channels of all urban pollution towards the Okpara River ( Figure 1). water catchments takes into account a factor related to the accessibility of the aquatic environment. The authors evaluated this factor with the only parameter of the geometric distance between the pollutant emission point and the aquatic environment. Several studies have established that surface water pollution is strongly linked to anthropogenic activities [21] [22] [23] [24] [25]. However, the transport of pollutants emitted to water plan depends on several other geomorphological parameters of watersheds. The probability of pollution entering the water system is a function of basin occupancy conditions, geomorphological features and the existence of the drainage network. The construction of an impact chain has allowed us to identify three parameters controlling the accessibility of aquatic systems ( Figure 2).

Data of the Study
We extracted the Digital Terrain Model (DTM) from the study area from the 30 m × 30 m (Shuttle Radar Topography Mission: SRTM 30, geoid WGS 84) square mesh raster, downloadable from the site (http://lta.cr.usgs.gov/srtm), then delimited the Okpara basin into four sub-basins designated by the names of their main talwegs: the sub-basins of Ganré, Wonka, Kokouro and the Dama sub-basin. This Digital Model was used to then generate the slope map and the drainage density map for each of the basins. The variation of these mapped indices has been neglected since the study area is geologically stable. The urban settlement map is obtained with the fragmentation polygons represented by their centroids throughout the basin. The situation for each basin was obtained by geoprocessing of the contours of the city with those of a centroids density map on the basins from updated files in dwg. format, subdivisions of 2004 and 2014 financed by the Millennium Challenge Account in Benin. Each of the mapped indices has been reclassified into five accessibility classes (Table 1). Open Journal of Modern Hydrology

Weighting Factors
Identified as the first condition of vulnerability of water resources in general, accessibility is for surface water one of the factors of expression of both the existence of pollution (activity of the source of pollution), and topographical conditions favorable to the contact between the aquatic environment and the polluting emission. However, in the light of BRGM's work, accessibility seems to be limited to the notion of proximity. It is then evaluated and mapped by the distance (in meters) separating the aquatic environment and the emission point from the pollution. In this work, we rather retain the principle of integrated effects of several factors on accessibility as an indicator of the vulnerability of surface water; thus this parameter is developped under a weighted combination of three factors. For time tracking purposes, the maps were produced for the dates 2004 and 2014. The multicriteria analysis method was used to weight the parameters identified in the chain of impacts ( Figure 2). The expression of calculation of the index of accessibility of the hydrographic network on a basin is characterized by the form.
This technique developped by Saaty towards the end of the 70's, allows to evaluate by a controlled decision process, the weights of each of the parameters from their comparison in pairs with criteria chosen to reach the goal. It has been used successfully for several works [26] [27]. Firstly, a comparison of the criteria was made by a square matrix of order 3. The scores are assigned according to Saaty's scorecard (Table 2) and respect the relation: ji a : Note translating the importance of the factor j compared to the factor i. 1 ij a  : i has more high importance compared to factor j in the decision. 1 ij a  : i has more weak importance compared to factor j in the decision.
We thus constituted a square matrix of order 3, also called original matrix. The original matrix is used in the computation of the eigenvectors with the expression: 11   The calculated weights are validated from the coherence or consistency ratio of the original matrix. Validation of the weights consists in eliminating or failing, minimizing possible errors of judgments made on the parameters during the notation. These errors are translated into the consistency ratio and depend on the number of parameters of the study. The coherence ratio is a function of a random index, the rational priority of the parameters and the number of parameters n. The priority vectors are obtained from the standardized matrix of as follows (Table 3):   Table 2. Grid for scoring parameters [28].

Importance on
Definition/Explanation 1 Both elements have the same importance 2 The line element is slightly larger.

3
The line element is a bit more important 4 The line element is moderately more important.

5
The line element is of high importance.

6
The line element is significantly more important than the columnar one.

7
The line element is very much more important than the columnar one. 8 The line element is very, very much more important than the columnar one.

9
The line element is absolutely more important than the columnar one.

Results and Discussion
The matrixes 1 A and 2 A are those used for the original and standardized matrices respectively in the calculation process (Table 4):

Conclusion
We observe through this study that the geomorphometric characteristics of watersheds significantly influence the vulnerability of surface water through the accessibility index. The topographic conditions and the human installation rate are two parameters of great importance and which are positioned as control and response factors respectively in relation to the drainage density of the basins. This interrelation highlighted justifies the interest of Saaty's multicriteria analysis method in assessing the vulnerability of surface water from several dependent factors.