Mapping groundwater potential zone using groundwater potentiality index in Sulamu, Central Fatuleu and West Fatuleu, Kupang, East Nusa Tenggara

Water scarcity often occurs during the dry season in Kupang Regency, which is a semi-arid area. The drought limits surface water availability during the dry season, thus making groundwater one of the reliable solutions for fulfilling the need for water from springs and wells. Therefore, mapping the potential of groundwater in the research area becomes the aim of this study, which has never been conducted before. The method used in this research is the Groundwater Potentiality Index, which uses five parameters to identify groundwater potential areas: fracture, lithology, topography, drainage, and rainfall. Field observation and secondary data collection were conducted to provide those parameters. The results are groundwater potential index classified into five groundwater potential zones, resulting in very low (17.07% and 20.39% of total area), low (51.16% and 48.94% of total area), moderate (26.9% and 27.05% of total area), high (4.09% and 3.1% of total area), and very high (0.77% and 0.53% of total area). Zones of very high are found along major fault areas. Areas with high groundwater potential zones indicated favorable conditions for wells positioning at groundwater exploration. These results can be a basis for action related to groundwater resource development in the research area.


Introduction
The scarcity of water resources is a significant problem in areas with arid and semi-arid climates, including the research area, which is located in Kupang Regency, East Nusa Tenggara (Figure 1).The area has a semi-arid climate with two seasons, where the dry season lasts longer than the rainy season.The dry season lasts 7 -9 months, while the rainy season lasts around 3 -4 months, making the study area vulnerable to meteorological drought [1,2].Water scarcity is a significant obstacle to sustainable development and the most important environmental challenge in semi-arid regions [3,4].Surface water cannot be used optimally throughout the year to meet the community's water needs, especially at the peak of the dry season.Water crises often occur, making it difficult for people to access clean water for their daily needs.Drought also impacts agricultural conditions, where the planting season only occurs once a year [2].
All the above conditions make groundwater one of the main reliable solutions for meeting the community's water needs, both from springs and production wells.Groundwater has the potential to fulfill the water demands of the community.Groundwater is one of the water resources that has an important role, especially in maintaining the balance and availability of water for several purposes, such as domestics, agriculture, livestock, and industry [6,5,7,8].
According to the hydrogeological regional map by the Directorate of Environmental Geology Indonesia, the study area is dominated by non-exploitable groundwater regions [9].This condition of the non-exploitable groundwater region is related to the condition of geology in the study area.As shown by the geological regional map of Kupang by the Geological Research and Development Centre Indonesia, the geological conditions of the area are dominated by the Bobonaro Complex, followed by the Maubisse Formation, Bisane Formation, Mutis Complex, and Cablac Formation [10].The Bobonaro Complex consists of two main components, namely scaly clay, which is a matrix of exotic blocks of various sizes.These exotic blocks include mica sandstones, limestones, flint, ultrabasic rocks, pillow lavas, and metamorphic rocks.The Maubisse Formation consists of interbedded limestone and pillow lavas.There is also a variety of limestone, the Cablac Formation, scattered locally in the study area, namely calcilutite, oolitic limestone, and dense limestone.The Mutis Complex consists of low to highdegree metamorphic rocks, including slate, phyllite, schist, amphibolite, amphibolite schist, quartzite, genes amphibolite, and granulite.From the rocks of these formations, it can be inferred that the research area has non-permeable outcrops (non-aquifer lithologies) composed of sedimentary rocks, including hard sedimentary rocks, metamorphic and igneous rocks [10].Based on all facts, it is important to know which areas have a possibility of groundwater potential in the study area that have yet to be previously conducted.This study aims to map groundwater potential zones using the Groundwater Potentiality Index (GPI) method.Mapping the potential of groundwater can be the basis for determining locations for groundwater exploration and exploitation [4,6,11,12] and IOP Publishing doi:10.1088/1755-1315/1311/1/0120243 is expected to be one of the solutions to overcoming drought in the study site.It can aid the proper development and utilization of groundwater resources to mitigate water scarcity and enhance the agricultural system [8,13].This research can also determine the suitability of the GPI method used to be applied to areas with similar conditions or characteristics to the research area in mapping groundwater potential zones.

Materials and Methods
The study area is Sulamu, West Fatuleu, and Central Fatuleu, Kupang Regency, East Nusa Tenggara, Indonesia.The study site consists of 13 villages and one sub-district with a total area of 579.39 km 2 (Figure 1).
The method used in this study is the Groundwater Potentiality Index (GPI40) by Ettazarini [14] and the Groundwater Potentiality Index (GPI81) by Ettazarini and Jakani [3].These methods are applied in arid and semi arid areas with impermeable outcrops.The GPI method uses five parameters, namely fractures (F), lithology (L), topography (T), drainage (D), and rainfall (R).Each parameter is given a notation based on its effect on groundwater presence, as seen in Table 1.Furthermore, the five parameters are assigned a weight that describes their contribution to groundwater availability, formulated in Equations 1 and 2 below [3,14].(2) Field observations were carried out for primary data collection, consisting of geological and hydrogeological observations.Geological observations include lithology and structure identification [8], while hydrogeological observations include the distribution of springs and drilled wells and their discharge [7,15].In addition, secondary data was also collected, including the Topographical Map of Indonesia (RBI) obtained from the Geospatial Information Agency's geoportal, Digital Elevation Model (DEM) obtained from DEMNAS with a spatial resolution of 0.27-arcsecond vertical datum EGM 2008, DEM SRTM, hydrogeology map scale 1: 250,000 [9], geological map scale 1: 250,000 [10], geological map of remote sensing imagery scale 1: 50,000 [16], and annual rainfall data.Processing of the thematic maps for each parameter up to the delineation of the groundwater potential zone map is using the Geographic Information System (GIS) technique integrated with Remote Sensing (RS) and other primary and secondary data [3,6,7,8,11,13,14,17].Processing of the GPI thematic parameter map is as follows: a. Fracture parameters (F) were analyzed by combining geological structure from the observed geological maps and lineament interpretation from DEMNAS imagery.Lineament interpretation involves assessing topographic lineaments such as valleys and ridges, which are suspected as the lineament of geological structures.Fractures were analyzed using a buffer technique with a distance of 125 m, resulting in a grid size of 250 x 250 m [3,14].The GPI notation is attributed to each fracture condition in Table 1.b. Lithology parameters (L) were analyzed from the observed geological map as a lithology map.
The GPI notation is given to each lithology unit of the study area according to the description in Table 1.c. Topography parameters (T) are processed from contour maps sourced from RBI maps and then into slope maps in percent units.GPI notation is then given according to the slope conditions in Table 1.d. Drainage parameters (D) are processed from DEM SRTM imagery data to generate a stream order map, classified using the Strahler method (1957).Drainage parameters are obtained by dividing the study area into sub-basins in the same order [3,14].The GPI notation is attributed to each subbasin based on the stream order that flows through it, according to Table 1.e. Rainfall parameters (R) were analyzed using average annual rainfall data from BMKG Kelas II Kupang and Balai Wilayah Sungai Nusa Tenggara 2 Kupang.The average annual rainfall data is plotted at the coordinates of the rainfall observation stations and then interpolated to obtain rainfall conditions in the study area.GPI notation is then attributed according to Table 1.Each analyzed thematic parameter map is then overlaid according to the rules and formula of the GPI method in Equations 1 and 2, resulting in the groundwater potentiality index (GPI).The index values are divided using equal intervals and categorized into five zones: very low, low, moderate, high, and very high [3].The groundwater potential zone map is then verified using data from springs and drilled wells, both from discharge and its occurrence [2,14].

Fracture (F).
Fractures are void spaces within massive rocks that serve as potential spaces for the movement and occupancy of water [5,6,8,14].The presence of fractures or joints can increase the permeability and function as secondary porosity [3,5,6,8,17].Aperture, spacing, interconnection, and fracture plane orientation play an important role in the presence and movement of groundwater in fractured aquifers [14].Zones with open, dense, and interconnected fractures are more likely to contain groundwater [3,6,8,14].Fractures that occur structurally in rocks with linear or curved features are referred to as lineaments [6,8,17,19].Lineaments represent fracture systems, discontinuity planes, and rock faults [17].
The fracture parameters combine geological structure data and lineament interpretation [3], as shown in Figure 2(a).Geological structure data is obtained from geological observation results and regional geological maps [10].There are several geological structures in the study area: faults, strike-slip faults, thrust faults, and folds.The fault is located in the southern part of the study area with a West-East (W-E) orientation.Strike-slip faults have Northwest-Southeast (NW-SE) and Northeast-Southwest (NE-SW) orientations.Thrust faults have a Northeast-Southwest (NE-SW) orientation.Meanwhile, the folds are anticlines and synclines with Northeast-Southwest (NE-SW) and Northwest-Southeast (NW-SE) orientations.Furthermore, the lineament interpretation was analyzed from DEMNAS data with a spatial resolution of 0.27-arcsecond by assessing the topographic lineaments in valleys and ridges in the study area, which are suspected as geological structure lineaments.

Lithology (L).
Lithology affects the presence of groundwater in an area [3,17].Lithology influences the permeability of rocks and the distribution of fracture patterns [3,8,14,17].Massive rock without fractures has little effect on groundwater occurrence unless secondary porosity develops through weathering and fractures in the rock, making it potential for groundwater availability [6].
The lithology map of the study site is a map that is processed based on the results of geological observations, regional geological maps [10], and geological maps of remote sensing imagery interpretation [16].The study site is dominated by fragmented claystone from the Bobonaro Complex.Then various kinds of limestone are distributed in the study area, including crystalline, coraline, and clastic limestone.In addition, there are metamorphic rocks in the form of schist from the Mutis Complex.The lithology map of the study site can be seen in Figure 3(a).
The lithology in the research area is then given a notation following the GPI description in Table 1.Ultrabasic rocks are given notation 1, schist and fragmented claystone are given notation 3, interbedded 6 shale and carbonate sandstones and sandy claystone are given notation 4, notation 5 is assigned to limestone with fractures, notations 8 is given to conglomerates with fractures, and notation 9 is given to alluvial.The lithological map that has been given GPI notation is then converted into raster form based on the notation values, resulting in a lithological thematic map, which can be seen in Figure 3

Drainage (D).
Drainage reflects the characteristics of surface and subsurface conditions [6].The drainage parameter is a classification that shows the surface runoff branching system as a stream order.Classification of stream orders based on Schumm (1956) and Strahler (1957), where a river without tributaries has order one and the confluence of two rivers with order n produces a river of order n + 1.The highest order indicates the importance of drainage in a watershed.The discharge of water will increase as the order of the river increases [14].The surface runoff impact is insignificant in impermeable and unfractured rocks [14].Drainage, which has a high density, makes the runoff move faster.While the drainage density is low, the runoff is slower, so the possibility for infiltration and recharge will be higher [6,17].
Drainage parameters were analyzed from the DEM Shuttle Radar Topographic Mission (SRTM) with a resolution of 30 m, producing a river order map according to the classification of the Strahler method (1957), which can be seen in Figure 4(a).Processing is carried out using GIS techniques with hydrology tools.Drainage in the study site reaches up to order 5.The river with the highest order is in the north of the study site.The result of the drainage (stream order) map is assigned a GPI notation to each sub-basin drained by the same stream order.The classification and notation of stream order are adjusted according to Table 1, where notation 7 is given to stream order 5 and low notation (3 -4) is given to streams with order < 3. The drainage map given a GPI notation is then converted into raster form based on the notation values assigned to each sub-basin with the same order, resulting in a drainage thematic map that can be seen in Figure 4

3..1.4. Topography (T).
Slope maps are used to assess the impact of topography on water infiltration into fractured aquifers [3].The slope is the main factor affecting surface runoff and infiltration due to the gravity effect [6,14,17].Based on the relationship between flow velocity, slope, and infiltration, the greater the slope, the higher the velocity of surface runoff, causing a shorter time for water to infiltrate and vice versa [13,17].The gentle slope makes the water flow more slowly so there is enough time for water to infiltrate the fractured aquifer [3,6,8,14,17].High groundwater potential areas are in gentle to flat slope areas [8,17].Topographic parameters are processed from contour maps sourced from RBI maps and analyzed into slope maps in percentage units (%), as shown in Figure 5(a).The topography conditions of the study site with gentle slopes (< 6%) are found in areas close to the coast in the west and south of the study site.The study site's highest slope (54-65%) is found on the limestone mountains.Furthermore, the GPI notation is given to the topography map, according to the percentage of slope shown in Table 1.High notation is given to gentle slopes, while low notation is assigned to steep slopes.The range of GPI notations for the research area is the lowest, with a value of 2, and the highest is 10.The topographic map that has been given the notation is then converted into raster form based on the notation value, resulting in a topographic thematic map, which can be seen in Figure 5

Rainfall (R).
Rainfall is the primary source of recharge for groundwater [6,8,14,17].The annual rainfall data is for the last 10 years, namely 2013 to 2022.Based on these data, the lowest annual rainfall of 837 mm/year occurred in 2016 in the study site.Rainfall parameters were then analyzed from annual rainfall data for the last 10 years.The average annual rainfall data at each observation point is input and interpolated with the IDW method to obtain the rainfall conditions in the study area, as shown in Figure 6(a).The average annual rainfall in the last 10 years in the study area ranges from 1300 mm/year to 1700 mm/year, which is way lower than the annual rainfall in Indonesia.
The highest classification and notation in the GPI method for rainfall parameters is the average annual rainfall > 900 mm/year.Based on the GPI notation in Table 1, the research area is given the notation of 10 for rainfall with a value of >900 mm/year.The rainfall map that has been given notation is then converted into raster form based on the notation values, resulting in a rainfall thematic map, which can be seen in Figure 6(b).

Groundwater Potentiality Index (GPI)
Groundwater Potentiality Index (GPI) is a method used for the relative assessment of groundwater potential zones by integrating all factors of presence and movement of groundwater, such as fracture, rainfall, slope, runoff, infiltration, lineament, drainage, and others [8,14].Each factor is weighted according to its effect on the presence and movement of groundwater.The groundwater potential zone classification is evaluated based on the total weighting of the Groundwater Potential Index (GPI) parameters to assess groundwater availability on a regional scale [8].A higher index score indicates higher groundwater potential [8].On the GPI40 method, fracture and lithology are the main parameters, while drainage, topography, and rainfall are considered as secondary parameters [14].The thematic map of each parameter in raster form is overlaid using Equation 1, producing a GPI40 index range of 53. 15 -200.4.The results of the GPI40 index were divided into equal intervals, resulting in very low zones (53.15 -82.60), low zones (82.61 -112.05),moderate zones (112.06 -141.50),high zones (141.50 -170.95), and very high zones (170.96-200.4) of groundwater potential.Groundwater potential zone map (GPI40) can be seen in Figure 7.The area of each zone class is summarized in  2, the low groundwater potential zone is the largest, covering 51.16% of the research area.Meanwhile, the smallest zone is the very high groundwater potential zone of 0.77% of the study area.In Figure 7, zones of very high groundwater potential are found along faults with a West-East (W-E) orientation in the southern part of the study area and shear faults with a Northwest-Southeast (NW-SE) orientation in the western region.Moderate to very high potential zones are located in parts of the south, east, and west of the study site.The groundwater potential zone map results in this method strongly illustrate the influence of the fracture parameters, where very high zones are found along the major structure.The GPI81 method considers fracture, lithology, and drainage as the main parameters, while topography and rainfall are secondary parameters [3].The thematic map of each parameter in raster form is overlaid using Equation 2, produces a GPI81 index range of 38.20 -180.65.The GPI81 index results were divided according to equal intervals, resulting in very low zones (38.20 -66.69), low zones (66.70 -95.18), moderate zones (95.19 -123.67),high zones (123.68 -152.16), and very high zones (152.17-180.65) of groundwater potential.The groundwater potential zone map (GPI81) shows in Figure 8 and the area of each zone class is summarized in Table 3.  3, the low groundwater potential zone is the largest zone at 48.93% of the study area.The zone with the smallest area is the very high groundwater potential zone of 0.53% of the research area.In Figure 8, we can see that zones of very high groundwater potential are found along faults with a West-East (W-E) orientation and shear faults with a Northwest-Southeast (NW-SE) orientation.Moderate to very high potential zones are located in parts of the south, east, and west of the upper part of the study site.The groundwater potential zone map results in this method also strongly illustrate the influence of fracture parameters, where very high zones are found along the major structure.
From both results, there are some characteristics of groundwater potential zone parameters.Zones with very high to high groundwater potential are located on major faults, long, dense, and interconnected fractures; lithologies of alluvium, conglomerate with fractures, and carbonate sandstone with fractures; and slopes <9%.Zones with low to very low groundwater potential are located on rare and isolated lineaments, not connected with major faults, on fragmented claystone lithologies, and slopes > 12%.

Verification
The groundwater potential zone results are verified using drilled wells and springs data based on discharge and its occurrence.Verification can confirm the level of accuracy of the groundwater potential zone map results [18].Field verification was carried out by comparing the GPI index value with the yields of boreholes and springs in the research area to determine the correlation coefficient and model reliability [14].Zones of high groundwater potential are expected to have more drilled wells and springs occurrence and more significant discharge than low and moderate groundwater potential zones.
Fifty-five wells and 25 springs are spread across the study area, collected from hydrogeological observations and secondary data obtained from Balai Wilayah Sungai Nusa Tenggara II and Dinas PUPR of Kupang Regency.Discharge data accompanied a total of 45 drilled wells and 20 springs and was then used to verify groundwater potential zones by comparing the GPI index value with discharge.Verification results by comparing the GPI index value with the yields of boreholes and springs can be seen in Figures 9 and 10, where the correlation coefficient results shown are 0.023 (GPI40) and 0.045 (GPI81).The correlation coefficient values from the two methods show that the GPI81 method provides better discharge predictions when compared to GPI40.
However, these results indicate that there is no correlation between the groundwater potential index value and wells and springs discharge.Compare this to Ettazarini (2007), who conducted research in Central Morocco and got a correlation coefficient of 0.71 by comparing the GPI40 index with water yields in 13 drilling points [14].Then, Ettazarini & Jakani (2020) did the GPI research in the Tafraoute region, Morocco, and got the correlation coefficient value by comparing the GPI81 index with 11 drilling point yields and revealed R 2 of 0.749 [3].Those results are good enough in mapping groundwater potential zones using the groundwater potentiality index.The poor correlation found in this research may be due to several reasons, such as limited discharge data and different climate conditions.Further verification is carried out by calculating the ratio between drilled wells and springs occurrence and the area of each groundwater potential zone.Verification results for the groundwater potential index (GPI40) can be seen in Figure 11a, where the results obtained are good enough to interpret the ratio of the number of drilled wells and springs to the area of the groundwater potential zones.The very high zone shows the highest ratio of 0.41 and is followed sequentially by high, moderate, low, and very low zones.For verification of the groundwater potential index (GPI81), shown in Figure 11b shows a ratio for the very high zone of 0.61, high zone of 0.15, moderate zone of 0.17, low zone of 0.12, and very low zone 0.08.Based on this verification, the GPI40 shows better results in the research area by predicting springs and boreholes per km 2 than GPI81.

Conclusion
The mapping of groundwater potential zones was processed to fulfill water needs due to the impact of drought in the research area.Groundwater potentiality index GPI40, shows very low to low groundwater potential zones covering 68.23% of the total area, moderate covering 26.9%, and high to very high 4.86% of the total study area.Meanwhile, groundwater potentiality index GPI81, shows very low to low groundwater potential zones covering 69.33% of the total area, moderate covering 27.05%, and high to very high 3.63% of the total study area.For both methods, zones of very high potential are located along major structures.Zones of medium to very high potential are found in the research area's southern, eastern, and western parts.Verification is carried out by comparing the GPI index with the discharge of drilled wells and springs and by calculating the ratio of the drilled wells and springs occurrence to the area of the zone class.The verification results show that the GPI40 method is better at predicting the presence of springs and boreholes per area of groundwater potential zone (number/km 2 ) in the research area.Meanwhile, the GPI81 method better indicates the discharge value of springs and boreholes in the study area (yields vs. index).If we look at the results of verification based on the correlation coefficient value, the Groundwater Potentiality Index (GPI) method used in this study is considered unsuitable for application in the research area or unable to represent the conditions of the research area.Except for the limited yield data, a qualitative assessment may also affect the production of a groundwater potential zone map.For future studies, local evaluation is needed, for example, for rainfall classification, so that the GPI can be appropriately applied in the research area or area with the same characteristics as the study site.Besides that, reliable data for the discharge of boreholes or springs is required to make the verification results more objective.

Figure 1 .
Figure 1.Location of the study site.
GPI 40 = 10.75 F + 5.25 L + 2.15 D + 3.20 T + 1.15 R (1) GPI81 = 10.75 F + 5.10 L + 4.75 D + 0.35 T + 0.60 R There are 4679 fracture lines distributed in the study site.Based on the rosette diagram, Northwest -Southeast (NW-SE) is the dominant fracture plane orientation.Fractures were analyzed by dividing the study area into a 250 x 250 m grid.The GPI notation in Table1is given to each fracture according to its condition on the grid.High notations(8 -9)  are assigned to geological structures, notations 4 -7 are given to dense and long lineaments close to the structure and interconnected plans, while notations 2 -3 are provided to infrequent and isolated lineaments.Notation 1 is given to an area without fracture.The buffering technique is used on the fracture line as far as 125 m, resulting in a GPI notation of fractures in a grid measuring 250 m x 250 m.The fracture map in grid form is then converted into raster form based on the given notation values, resulting in a fracture thematic map, as shown in Figure2(b).

Figure 4 .
Figure 4. Thematic map of (a) drainage parameters and (b) GPI classification.

Figure 5 .
Figure 5. Thematic map of (a) topographic parameters and (b) GPI classification.Topographic parameters are processed from contour maps sourced from RBI maps and analyzed into slope maps in percentage units (%), as shown in Figure5(a).The topography conditions of the study site with gentle slopes (< 6%) are found in areas close to the coast in the west and south of the study site.The study site's highest slope (54-65%) is found on the limestone mountains.Furthermore, the GPI notation is given to the topography map, according to the percentage of slope shown in Table1.High notation is given to gentle slopes, while low notation is assigned to steep slopes.The range of GPI notations for the research area is the lowest, with a value of 2, and the highest is 10.The topographic map that has been given the notation is then converted into raster form based on the notation value, resulting in a topographic thematic map, which can be seen in Figure5(b).
Figure 5. Thematic map of (a) topographic parameters and (b) GPI classification.Topographic parameters are processed from contour maps sourced from RBI maps and analyzed into slope maps in percentage units (%), as shown in Figure5(a).The topography conditions of the study site with gentle slopes (< 6%) are found in areas close to the coast in the west and south of the study site.The study site's highest slope (54-65%) is found on the limestone mountains.Furthermore, the GPI notation is given to the topography map, according to the percentage of slope shown in Table1.High notation is given to gentle slopes, while low notation is assigned to steep slopes.The range of GPI notations for the research area is the lowest, with a value of 2, and the highest is 10.The topographic map that has been given the notation is then converted into raster form based on the notation value, resulting in a topographic thematic map, which can be seen in Figure5(b).

Figure 7 .
Figure 7. Map of groundwater potential zone (GPI40) and the distribution of wells and springs.

Figure 8 .
Figure 8. Map of groundwater potential zone map (GPI81) and the distribution of wells and springs.

Figure 11 .
Figure 11.The ratio of drilled wells and springs to the area of the groundwater potential zone using (a) the GPI 40 and (b) GPI 81 .

Table 2 .
Area of groundwater potential zone using the GPI40 method.

Table 3 .
Area of groundwater potential zone using the GPI81 method.