The selected study area for this research is located in the Northern part of Jordan (Figure 1). It is on the border with Syria and has an area of 1235 km², which comprises approximately 1.4% of the total area of Jordan. The study area has several towns and villages (Figure 2) with an urban area of 31.74 km² which is approximately 2.57% of the study area. The agricultural area (Figure 2) within the study area is 27.76 km² which comprises approximately 2.25% of the study area. The agricultural activities in the area include growing vegetables, fruits and olives [10] . Most of farmers use excessive amounts of fertilizers and pesticides [11] .

The study area is mainly flat, where elevation varies between 642 m above sea level in the South to 1224 m above sea level near the Syrian border in the North (Figure 3). The average gradient for the study area is less 2%.

Surface water in the area is mainly from rainfall that occurs between November and March annually. The area receives approximately 250 million cubic meter of rainfall on annual basis. Most of the rainfall is lost due to evaporation (approximately 90%), while only 5% of rainfall generate runoff. The generated runoff flows through the Wadis (streams) that runs mainly from the North towards the South, South East and the South West (Figure 4). Surface water within the study area suffers from various sources of pollution. Among these sources are the followings [11] :

1. The use of fertilizers, pesticides, insecticides and herbicides by farmers in the area.

2. Runoff generated within urban areas which carryout garbage and other pollutant substances to the nearby Wadis.

3. Use of vehicles with oil spills, lead and corroded particles.

There are few mythologies developed in the USA to investigate surface water susceptibility to pollution. Among these methodologies are the followings:

Ÿ A methodology by [5] which defines surface water susceptibility to pollution based on 5 factors. These factors include river network, soil, urban distribution, land cover and slope.

Ÿ An overlay and index methodology developed by the USGS for rating the characteristics of a watershed [2] . It is based on using 5 factors including average annual precipitation, land surface slope, land cover, land use, and groundwater contribution.

Ÿ A methodology developed by [3] that allows the evaluation of the watershed susceptibility to surface water pollution based on major characteristics of a watershed and land uses. It is based on 7 factors including wastewater discharges, recreational land use impacts, agricultural land use impacts, size of watershed, transportation avenues, industrial land use impacts and the amount of vegetation Cover.

Ÿ A methodology developed by the Laboratory for Spatial Analysis and the Geosciences at the University of Minnesota-Duluth, USA ( [4] and [6] ). This methodology was initially developed in 2003 to investigate the surface water susceptibility to pollution from non-point sources. It is based on a fundamental principle that areas with more prone to runoff are capable of transporting suspended sediments to water bodies. Only four intrinsic factors are involved in the estimation of surface-water runoff potential for any given study area using GIS techniques. These factors are gradient slope, distance to water (streams and lakes), land cover and soil properties. Each factor has a weight to reflect its’ contribution to surface-water runoff and therefore a general indication of surface water pollution potential.

In this research, a modified methodology will be introduced by having the factors that might influence surface water susceptibility to pollution. The modified index (SWSi) has 6 factors; including gradient slope (%) (GS), distance to Wadis (streams) (DW), soil clay (%) (SC), distance to agricultural lands (DA), distance to urban areas (DU) and distance to roads (DR). In order to allocate the appropriate weight for each factor, 7 experts in the field of surface water quality from various Jordan Universities and organizations were invited to assign a weight for each factor. Experts were asked to give 6 for the factor that has the highest impact on surface water pollution and 1 for the lowest impact factor. A methodology adopted by [12] was used to come up with a single weight for each factor and overcome the variation between experts’ weights allocation. The methodology is based on using both the Mean and Median for experts responses and decides the appropriate weight for each factor. Table 1 summarizes factors weights suggested by the experts. Based on this Table, weights were given to these factors based on each factor contribution to the surface water susceptibility to pollution. The Median value for each factor was selected to represent the appropriate weight for the six factors used in this research. The Mean values also show that the selected weights are reasonable.

The highest weight (6) was given to the gradient slope (GS) (%), while the lowest weight (1) was given to distance to roads (DR). Table 2 summarizes the classes, weight and ratings for each factor. The justification for using these factors could be summarized as follow:

1. GS: Surface water runoff occurs whenever there is excess water on a slope that cannot be absorbed into the soil or is trapped on the surface. The steeper the slope of a field, the higher potential for runoff [13] . Surface water is more susceptible to pollution when runoff is high and infiltration is low [4] . In this research, GS was classified into 5 classes based on [4] (Table 2).

2. DS: It is an important factor in determining whether surface water is susceptible to pollution or not. In this research a modification to distances suggested by [4] was carried out through the use of meters instead of feet and classify the distance into 5 classes instead of 6 classes (Table 2).

3. SC: High clay contents’ soils have several properties that might lead to the movement of pollutants from agricultural lands. Also, surface structure of soils can become degraded in high clay contents’ soils. This will lead to the formation of crust which restricts infiltration and increases runoff. Runoff increases when clay soils are wet, due to soil compactness. The runoff may contain pollutants and could affect the surface water quality [14] . In this research, SC were classified into 5 classes as listed in Table 2.

4. DA: The agricultural non-point source (NPS) pollution is the leading source of water quality impacts on rivers and lakes [15] . DA were classified into five classes as listed in Table 2.

5. DU: Urban area is one of the most harmful factors affecting surface water health and a major challenge facing watershed managers. Urban runoff affects water chemistry by changing heavy metals and nutrients levels such as phosphorus and nitrogen [16] . DU were classified into five classes as listed in Table 2.

6. DR: Highway run-off could be identified as a major source of diffuse pollution that might contaminate surface water [17] . In this research, DR were classified into five classes as listed in Table 2.

The governing equation (Equation (1)) for the modified index (SWSi) is shown below:

where GS: Gradient Slope (%), DS: Distance to Surface Water, SC: Soil Clay (%), DA: Distance to Agricultural Lands, DU: Distance to Urban Areas, DR: Distance to Roads, w: Weight and r: Ratings.

The overall index calculation could be classified into 4 susceptibility classes as listed in Table 3 (Low, Moderate, High and Very High).

Dada required for this research were collected from several governmental agencies in Jordan and international agencies. Table 4 describes these dada and their sources.

Table 7 provides a summary statistics for the removal of one statistically significant factor on the SWSi values. This table indicates that the most important factors in the SWSi were DR, DU, DA and DW followed by GS and SC. The highest sensitivity value was associated with DR (47.9), while the lowest sensitivity value (35.8) was associated with SC.

Based on the map removal sensitivity analysis test developed by [29] and adopt- ed by [30] , which was used in this research to identify the sensitivity of each factor in the SWSy map. In this test, Equation (2) was used to calculate the sensitivity index (S) for the factors used in the vulnerability index.

where: S is the sensitivity index of the factor;

V is the intrinsic vulnerability index of the method;

N is the total number of factors used to calculate V;

V_xi represents the intrinsic vulnerability index obtained after removal of the factor X and

n: the number of factors after removing one factor.

Based on Table 7 that lists the determined partial indices and Equation (2), the sensitivity index was calculated for each factor of the SWSi. Table 8 indicates that SC, GS and DR factors have a strong influence on the SWSi map, while DW, DA and DU factors have a moderate influences on the SWSi map.