Toward the Sustainable Use of Groundwater Springs: A Case Study from Namibia

: The water supply in drylands mainly relies on groundwater, making it a crucial resource. Springs in southern Africa are often underutilized, and are neither protected nor monitored. Thus, the aim of this study was to evaluate their quality in a sample area in northwestern Namibia and to propose solutions for the sustainable use of springs. In total, 35 springs and hot springs were evaluated in the study area located in the drier part of Namibia (P mean = 150–400 mm/year), an area highly impacted by ongoing climate change with longer and more frequent drought seasons. The springs there are mostly uncaptured and the discharge is in the form of surface runoff, which is mainly lost to the atmosphere by evaporation. Most of the studied springs were perennial, despite a severe drought period. Local communities rely on the springs mainly for livestock and human consumption, as well as for irrigation. However, 71% of the springs do not have any protective measures. The temperature, pH, conductivity and alkalinity were tested in situ. In total, 20 samples were collected and analyzed for major ions (boron, ﬂuoride, silica and strontium) and total dissolved solids (TDS). The physical and inorganic results mostly indicated good and excellent quality water for human consumption, while the hot springs tended to have poor water quality in terms of Namibian standards, indicating that the water was not ﬁt for human consumption.


Introduction
In the majority of sub-Saharan African countries, access to water, especially affordable, safe drinking water, is a major challenge. In semi-arid to arid rural areas, such as the Kaokoland in northwestern Namibia, groundwater is the main source of freshwater and is used for domestic and agricultural purposes. Unfortunately, the quality and sustainability of this resource are still uncertain and may be affected by seasonal effects, global climate change and population growth.
In this part of the country, most of the water is abstracted from shallow boreholes (<50 m) and wells. However, a surprisingly high number of springs are emerging at the surface. Some of these springs dry up during the dry season, but the majority remain perennial, sustaining wildlife, livestock and humans.

Hydro-Climatic Context
In the arid and semi-arid country of Namibia, precipitation is scarce and highly variable, both temporally and locally. In the northwestern region, rainfall events mainly occur between October and April, and the mean annual precipitation ranges between 150 mm/year in the west to 400 mm/year in the east, while the potential evapotranspiration can exceed 3000 mm/year. The average annual temperature is 20 to 25°C and can reach 35°C between September and January [20].
Except for the Namibian border rivers, no surface water is perennial in the country. The occurrence, distribution, and extension of surface runoff are seasonal and depend on precipitation [1,3,21,22]. Namibia has often experienced drought periods, especially over the past few decades [23][24][25], but one of the worst drought episodes in the country occurred in 2019 [26][27][28][29][30].

Regional Geology
The study area, situated on the western rim of the Cuvelai-Etosha Basin (CEB, Figure 1), is bounded by two orogenic belts resulting from the collision between (i) the Congo Craton and the Kalahari Craton, and (ii) the Congo Craton and the Rio de la Plata Craton,

Hydro-Climatic Context
In the arid and semi-arid country of Namibia, precipitation is scarce and highly variable, both temporally and locally. In the northwestern region, rainfall events mainly occur between October and April, and the mean annual precipitation ranges between 150 mm/year in the west to 400 mm/year in the east, while the potential evapotranspiration can exceed 3000 mm/year. The average annual temperature is 20 to 25 • C and can reach 35 • C between September and January [20].
Except for the Namibian border rivers, no surface water is perennial in the country. The occurrence, distribution, and extension of surface runoff are seasonal and depend on precipitation [1,3,21,22]. Namibia has often experienced drought periods, especially over the past few decades [23][24][25], but one of the worst drought episodes in the country occurred in 2019 [26][27][28][29][30].

Regional Geology
The study area, situated on the western rim of the Cuvelai-Etosha Basin (CEB, Figure 1), is bounded by two orogenic belts resulting from the collision between (i) the Congo Craton and the Kalahari Craton, and (ii) the Congo Craton and the Rio de la Plata Craton, during the Neoproterozoic Pan-African Orogeny [31,32], which formed the Kaoko Belt to the west and the Damara Belt to the south ( Figure 1).
The oldest (Pre-Damaran) rocks present in the area are Paleoproterozoic metamorphic complexes of volcanic, granite and gneiss rocks (Figure 2), which characterize the basement of the CEB. Largely outcropping on the western rim of the basin, eastern Kaokoland, the Neoproterozoic Damara Supergroup [32] constitutes the main succession of the Northern Otavi Platform (Figure 2). The oldest deposits of the Damara Supergroup are the synrift fluvio-clastic metasediments of the Nosib Group, followed by an important carbonate platform with deposits of the Otavi Group, and the syn-tectonic clastic deposits of the Mulden Group, which rest on the Otavi dolomites. The Karoo succession reflects climate change during the Paleozoic, with glacial tillite deposits of the Dwyka Formation along the valleys. Resting in a disconformable manner on the Karoo Formation, the Kalahari group consists of Tertiary to recent sands, gravel, calcrete, and clay formations, which can reach a thickness of 600 m toward the basin [32]. The Etosha Calcrete Formation is a groundwater calcrete cover formed by springs emerging at the base of the karstic Otavi carbonate ridges that form the western and southern margins of the basin [33,34] (Figure 2). during the Neoproterozoic Pan-African Orogeny [31,32], which formed the Kaoko Belt to the west and the Damara Belt to the south ( Figure 1). The oldest (Pre-Damaran) rocks present in the area are Paleoproterozoic metamorphic complexes of volcanic, granite and gneiss rocks (Figure 2), which characterize the basement of the CEB. Largely outcropping on the western rim of the basin, eastern Kaokoland, the Neoproterozoic Damara Supergroup [32] constitutes the main succession of the Northern Otavi Platform (Figure 2). The oldest deposits of the Damara Supergroup are the synrift fluvio-clastic metasediments of the Nosib Group, followed by an important carbonate platform with deposits of the Otavi Group, and the syn-tectonic clastic deposits of the Mulden Group, which rest on the Otavi dolomites. The Karoo succession reflects climate change during the Paleozoic, with glacial tillite deposits of the Dwyka Formation along the valleys. Resting in a disconformable manner on the Karoo Formation, the Kalahari group consists of Tertiary to recent sands, gravel, calcrete, and clay formations, which can reach a thickness of 600 m toward the basin [32]. The Etosha Calcrete Formation is a groundwater calcrete cover formed by springs emerging at the base of the karstic Otavi carbonate ridges that form the western and southern margins of the basin [33,34] (Figure 2).

Spring Occurrence and Control in a Hydrogeological Context
A spring corresponds to an emergence point where groundwater flows out to the surface [35]. In Namibia, most of the springs are located in or at the base of the carbonate and dolomitic rocks of the Otavi Group ( Figure 2) [1,33]. This group is characterized by fractured aquifers of moderate hydrogeological potential. However, the contact between the non-porous rocks of the Nosib or Mulden groups (aquitards) shows a much larger

Spring Occurrence and Control in a Hydrogeological Context
A spring corresponds to an emergence point where groundwater flows out to the surface [35]. In Namibia, most of the springs are located in or at the base of the carbonate and dolomitic rocks of the Otavi Group ( Figure 2) [1,33]. This group is characterized by fractured aquifers of moderate hydrogeological potential. However, the contact between the non-porous rocks of the Nosib or Mulden groups (aquitards) shows a much larger karstification phenomenon, and springs were mainly identified at these stratigraphic contacts, especially in the Sesfontein area [1]. The emergence of the springs can be structurally controlled at a fracture contact (Figure 3). bottom of valleys [36], some springs seep from the thick karstic calcrete deposits of the Etosha Calcrete Formation (Figure 3). In the context of very high potential evaporation, these springs have built-up a calcrete deposit [37] which locally can exceed 100 m in thickness [33,38].
Most of the geothermal springs visited emerge from deep fractures ( Figure 3) localized in one of the oldest rocks outcropping in Namibia [39,40], i.e., in the granite and/or para-to ortho-gneiss rocks from the Archaean that define the basement.

Material and Methods
Groundwater from sixteen springs and four thermal springs ( Figure 2) was sampled as close to the point of emergence as possible. These springs were sampled from 5 February to 10 February 2019. Very low amounts of precipitation were measured 3 days before the sampling took place (6 mm, Appendix A).
The following parameters were measured in situ using a portable Hach multimeter that was calibrated twice per day: temperature (T); pH; electrical conductivity (EC); and dissolved oxygen (O2). The alkalinity was measured and calculated in situ by titration. Then, 200 mL of water was sampled and filtered through a 0.1 µm filter on-site and kept in cold storage until chemical analyses were performed in the laboratory. The major ion concentrations (Cl − , SO4 2− , NO3 − , NO2 − , F − , K + , Na + , Ca 2+ , Mg 2+ ), the alkalinity (as CaCO3), the concentrations in Fe and Mn, and turbidity were analyzed at Analytical Laboratory Services in Windhoek, Namibia. The concentrations of the cations K + , Na + , Ca 2+ , Mg 2+ , Fe and Mn were analyzed using an inductively coupled plasma optical emission spectrometer (ICP-OES). The alkalinity was analyzed using the acidimetric method. NO3 − and NO2 − At the foot of the karstic dolomitic Otavi Group reliefs of Kaokoland, mostly at the bottom of valleys [36], some springs seep from the thick karstic calcrete deposits of the Etosha Calcrete Formation (Figure 3). In the context of very high potential evaporation, these springs have built-up a calcrete deposit [37] which locally can exceed 100 m in thickness [33,38].
Most of the geothermal springs visited emerge from deep fractures ( Figure 3) localized in one of the oldest rocks outcropping in Namibia [39,40], i.e., in the granite and/or parato ortho-gneiss rocks from the Archaean that define the basement.

Material and Methods
Groundwater from sixteen springs and four thermal springs ( Figure 2) was sampled as close to the point of emergence as possible. These springs were sampled from 5 February to 10 February 2019. Very low amounts of precipitation were measured 3 days before the sampling took place (6 mm, Appendix A).
The following parameters were measured in situ using a portable Hach multimeter that was calibrated twice per day: temperature (T); pH; electrical conductivity (EC); and dissolved oxygen (O 2 ). The alkalinity was measured and calculated in situ by titration. Then, 200 mL of water was sampled and filtered through a 0.1 µm filter on-site and kept in cold storage until chemical analyses were performed in the laboratory. The major ion concentrations (Cl − , SO 4 2− , NO 3 − , NO 2 − , F − , K + , Na + , Ca 2+ , Mg 2+ ), the alkalinity (as CaCO 3 ), the concentrations in Fe and Mn, and turbidity were analyzed at Analytical Laboratory Services in Windhoek, Namibia. The concentrations of the cations K + , Na + , Ca 2+ , Mg 2+ , Fe and Mn were analyzed using an inductively coupled plasma optical emission spectrometer (ICP-OES). The alkalinity was analyzed using the acidimetric method. NO 3 − and NO 2 − were determined using the colorimetric method, and the turbidity, SO 4 2− , F − and Cl − were measured using the nephelometric, turbidimetric, electrometric and argentometric methods, respectively.
One hundred and sixteen existing hydrogeochemical results for springs throughout Namibia were collected from the GROWAS II Database (DWA, Department of Water and Agriculture) of the Ministry of Agriculture, Water and Forestry (MAWF).
The Water Quality Index (WQI) was used to assess water quality in this study. The WQI is widely used and essential to evaluating the quality of surface water and groundwater to determine if it is safe for human consumption. The WQI was calculated using the weighted arithmetic water quality index, which was originally proposed by Horton (1965) [17] and developed by Brown et al. (1972) [41].
The WQI is calculated by assigning a weight (w i ) to each parameter (pH, TDS, Ca 2+ , Mg 2+ , Na + , K + , SO 4 2− , Cl − , NO 3 − and F − ) according to its relative importance to the quality of water for drinking purposes [42]. A weight between 1 and 5 was assigned to each of the parameters according to their effects on water quality [18]. Total dissolved solids, nitrate and fluoride were assigned the highest weight because of the impact they have on water quality. Chloride and potassium were given the minimum weight [18]. The computed relative weight (W i ) values for the geochemical parameters are presented in Table 1; they were calculated according to the following equation: where W i is the relative weight, w i is the weight of each parameter and n is the number of parameters. The quality rating q i was computed using the concentration of each parameter in groundwater (C i ) and the Namibian Drinking Water Standard (S i ) for each parameter [13]: where C i is the concentration of each chemical parameter for each groundwater sample (in mg/L), and S i is the Namibian Drinking Water Standard for each chemical parameter (in mg/L) according to the guidelines given by the DWA [13]. The quality rating for pH is calculated using the following equation: where pH i is the observed pH value. The WQI is finally calculated using the following equation: where q i is the rating for each parameter and W i is the relative weight of each parameter.
The computed WQI values are classified into five water quality categories (Table 2). In addition, all single parameters from the GROWAS database were used to evaluate and characterize the groundwater quality with respect to permissible values for human consumption, based on the guidelines for drinking water quality of the World Health Organization [12] and the Namibian Standards for the Evaluation of Drinking Water Quality for Human Consumption [13]. The Namibian Standards classify the water quality into four groups: A-excellent; B-acceptable; C-low risk; D-high risk or unsuitable for human consumption. Table 2 presents the permissible standards for class A water together with the WHO recommended values. This study is only based on the inorganic composition of groundwater, and it does not take into account organic or bacteriological components, which have already been analyzed in the Cuvelai Etosha Basin, notably, in hand-dug wells [43].

Field Observations
Despite a severe drought period in Namibia in 2019 and 2020 [24,28], some of the springs visited still showed flowing water, and only a few of them were dry. Of the thirty-five springs visited, only three were dry, and they were potentially seasonally dry. Sixteen of the springs were not captured or were captured in a very basic way using only a small trench dug around the emergence point (9). Twenty-five springs had some protective measures around the emergence point ( Table 3). All of these springs are mostly used for livestock watering, the irrigation of small community gardens and, when there is no borehole substitution, human consumption. Livestock usually drinks directly from the spring without any separation from the collection point for human consumption. Despite the fact that these groundwater resources are used by all the communities in the area, none of these emergence points are monitored, nor are they controlled to guarantee groundwater quality. Table 3 contains the spring capture methods observed as well as the protective measures taken for springs in northwestern Namibia, and Figure 4 provides some photos of field observations. Table 3. Spring capture methods and protective measures observed during the study. At locations with unprotected springs, where the emergence is open air and easily accessible, the water is turbid and presents suspended matter. Some springs distinctly show the flowing of clear water, but some of them present lower flow rates or low seep. Three of the thermal springs visited were used for human consumption, livestock watering and irrigation, and two of them were used as hot pools for tourism.

Physico-Chemical Parameters
The temperature of the studied springs in northwestern Namibia varied from 20.1 to 32.0°C, while the temperature of the thermal springs varied from 41 to 44.1°C. The electrical conductivity of the samples was variable; the electrical conductivity must be between 701 and 1271 µS/cm for class A springs (water with excellent quality). Two of the springs were above the limit of 1000 µS/cm given by the guidelines of the World Health Organization. The EC varied from 1124 to 2740 µS/cm in geothermal springs, and many were not in class A but were still in class B (water with acceptable quality). The pH of all the springs sampled varied between 6.75 and 7.89. The limits, medians and averages of the physicochemical field parameters are given in Table 4. At locations with unprotected springs, where the emergence is open air and easily accessible, the water is turbid and presents suspended matter. Some springs distinctly show the flowing of clear water, but some of them present lower flow rates or low seep. Three of the thermal springs visited were used for human consumption, livestock watering and irrigation, and two of them were used as hot pools for tourism.

Physico-Chemical Parameters
The temperature of the studied springs in northwestern Namibia varied from 20.1 to 32.0 • C, while the temperature of the thermal springs varied from 41 to 44.1 • C. The electrical conductivity of the samples was variable; the electrical conductivity must be between 701 and 1271 µS/cm for class A springs (water with excellent quality). Two of the springs were above the limit of 1000 µS/cm given by the guidelines of the World Health Organization. The EC varied from 1124 to 2740 µS/cm in geothermal springs, and many were not in class A but were still in class B (water with acceptable quality). The pH of all the springs sampled varied between 6.75 and 7.89. The limits, medians and averages of the physico-chemical field parameters are given in Table 4.

Major Ions
Figures 5 and 6 present the chemistry of the springs in the form of a Piper diagram and Schoeller diagram, which were used to identify the different water facies observed in the study area (in black, Figure 5) and those from the GROWAS database (in grey, Figure 5).

Major Ions
Figures 5 and 6 present the chemistry of the springs in the form of a Pip and Schoeller diagram, which were used to identify the different water facies o the study area (in black, Figure 5) and those from the GROWAS database (in grey The twenty sources sampled for this study covered similar geochemica shown by the springs from the Namibia-wide GROWAS dataset. The domina cies of the springs in northwestern Namibia (black dots in Figure 5 and blue lin 6) and all Namibian springs (GROWAS) was principally Ca-Mg-HCO3. Some not show dominant ions and presented a mixed water type. Potassium and s the major ions that had the lowest concentrations in the springs (Figure 6).
The majority of the thermal springs (black triangles in Figure 5 and oran lines in Figure 6) were located in the Na-Cl-HCO3 area, except for the La Ro Spring, which had no dominant ions. Magnesium, potassium and calcium wer ions that had the lowest concentrations in hot springs.
The Schoeller diagram in Figure 6 clearly distinguishes the Ca-Mg-HCO3 w of the springs from the Na-Cl water facies of the thermal springs, except for chelle Hot Spring, which presented mixed facies. The thermal springs also ge a much higher SO4 content.   Table 5 presents the concentrations of major ions in the springs and thermal springs sampled in the area of interest (see Figure 1 for locations) compared to the Namibian Standards for the Evaluation of Drinking Water for Human Consumption [13]. Almost all the springs visited in northwestern Namibia showed ion concentrations that fell into class A (excellent quality). Three of these springs had sodium, magnesium or calcium concentrations slightly above the class A threshold (134 mg/L, 83 mg/L and 159 mg/L), but they did fall within class B, indicating water with an acceptable quality (in yellow, Table 5).
All the thermal springs (TSs) had at least two ion concentrations higher than the class A limits. Sodium and fluoride concentrations ranged between 191 and 522 mg/L and between 1.6 to 8.7 mg/L, respectively; all of these values exceeded the class A limits. Some of these hot springs fell into class B, water with an acceptable quality (in yellow, Table 5), and some of them into class C (water with a low health risk, in orange). The majority of the fluoride concentrations fell into class D, water with a high health risk (in red). The sulfate and chloride concentrations were also above the class A limits for three and two, respectively, of the thermal springs analyzed.
The turbidity measured in the springs was extremely variable; one spring was classified as class D (water with a high health risk) with regard to turbidity (in red, Table 5), one was in class C (water with a low health risk, in orange) and four samples were in class B (acceptable quality, in yellow). Only seven springs showed turbidity values that fell in class A (excellent water quality). The turbidity values measured in the thermal springs were below the class A limit.
The TDS concentrations were below the WHO limit of 1000 mg/L in all the springs sampled, but they were above this limit in the majority of the thermal springs, with the exception of Okangwati Hot Spring. The twenty sources sampled for this study covered similar geochemical facies, as shown by the springs from the Namibia-wide GROWAS dataset. The dominant water facies of the springs in northwestern Namibia (black dots in Figure 5 and blue lines in Figure 6) and all Namibian springs (GROWAS) was principally Ca-Mg-HCO 3 . Some springs did not show dominant ions and presented a mixed water type. Potassium and sulfate were the major ions that had the lowest concentrations in the springs ( Figure 6).
The majority of the thermal springs (black triangles in Figure 5 and orange and red lines in Figure 6) were located in the Na-Cl-HCO 3 area, except for the La Rochelle Hot Spring, which had no dominant ions. Magnesium, potassium and calcium were the major ions that had the lowest concentrations in hot springs.
The Schoeller diagram in Figure 6 clearly distinguishes the Ca-Mg-HCO 3 water facies of the springs from the Na-Cl water facies of the thermal springs, except for the La Rochelle Hot Spring, which presented mixed facies. The thermal springs also generally had a much higher SO 4 content. Table 5 presents the concentrations of major ions in the springs and thermal springs sampled in the area of interest (see Figure 1 for locations) compared to the Namibian Standards for the Evaluation of Drinking Water for Human Consumption [13]. Almost all the springs visited in northwestern Namibia showed ion concentrations that fell into class A (excellent quality). Three of these springs had sodium, magnesium or calcium concentrations slightly above the class A threshold (134 mg/L, 83 mg/L and 159 mg/L), but they did fall within class B, indicating water with an acceptable quality (in yellow, Table 5).
All the thermal springs (TSs) had at least two ion concentrations higher than the class A limits. Sodium and fluoride concentrations ranged between 191 and 522 mg/L and between 1.6 to 8.7 mg/L, respectively; all of these values exceeded the class A limits. Some of these hot springs fell into class B, water with an acceptable quality (in yellow, Table 5), and some of them into class C (water with a low health risk, in orange). The majority of the fluoride concentrations fell into class D, water with a high health risk (in red). The sulfate and chloride concentrations were also above the class A limits for three and two, respectively, of the thermal springs analyzed.
The turbidity measured in the springs was extremely variable; one spring was classified as class D (water with a high health risk) with regard to turbidity (in red, Table 5), one was in class C (water with a low health risk, in orange) and four samples were in class B (acceptable quality, in yellow). Only seven springs showed turbidity values that fell in class A (excellent water quality). The turbidity values measured in the thermal springs were below the class A limit. The TDS concentrations were below the WHO limit of 1000 mg/L in all the springs sampled, but they were above this limit in the majority of the thermal springs, with the exception of Okangwati Hot Spring. Table 5. Conductivity and concentrations of major ions of visited springs and thermal springs. Cells without color indicate ion concentrations in class A-excellent water quality according to the Namibian Standards. Yellow indicates ion concentrations in class B-acceptable quality. Orange indicates ion concentrations in class C-low health risk; red indicates ion concentrations in class D-high health risk. TS = thermal spring. See Figure 2 for spring locations.

Water Quality Index
The WQIs calculated for each of the springs and thermal springs in Namibia are shown in Figure 7. The majority of the springs (66%) showed excellent to good water quality (in blue and green, Figure 7), 17% had poor water quality (in yellow), 7% had very poor water quality (in orange) and 10% had water unsuitable for human consumption (in red). Table 6 shows the WQIs calculated for each of the "cold" springs (Sp) in the study area as well as those calculated for the thermal springs. In the area of interest, every cold spring (non-thermal) showed excellent or good water quality for drinking purposes. However, the majority of thermal springs (TS) had poor water quality, except for one (La Rochelle) that had good water quality.

Water Quality Index
The WQIs calculated for each of the springs and thermal springs in Namibia are shown in Figure 7. The majority of the springs (66%) showed excellent to good water quality (in blue and green, Figure 7), 17% had poor water quality (in yellow), 7% had very poor water quality (in orange) and 10% had water unsuitable for human consumption (in red).

Figure 7.
Map showing WQI for springs and thermal springs in Namibia that focuses on springs sampled in this study (blue = excellent water quality; green = good water quality; yellow = poor water quality; orange = very poor water quality; red = water unsuitable for human consumption).

Interpretation and Discussion
In northwestern Namibia, communities essentially rely on groundwater due to the lack of perennial surface water. However, a high percentage of perennial springs potentially reflect moderate-to-good groundwater potential in this arid area. The springs also present the cheapest water resource, as water from a spring can be captured at the direct emergence point without any cost.

Water Facies
Assuming that groundwater is in equilibrium with the minerals within the aquifer, as assumed in the calculation of the saturation index (SI) using the PHREEQC package [13], water facies mainly originate from water-rock reactions but could be influenced by human activities. The water facie Ca-Mg-HCO 3 is the most representative of springs in the study area, and it is very typical of carbonate and dolomitic rocks [1,[44][45][46][47]. Most of the springs have water that is in equilibrium with calcite and disordered dolomite (Appendix B); calcium and magnesium derive from the geological formations from which the majority of the springs emerge. Bicarbonate (HCO 3 ) is formed by the interaction of dissolved carbon dioxide and limestone and dolomite.
The Na-Cl water type is only recorded in the thermal springs and represents the main facies of this groundwater type in northwestern Namibia; it is strongly correlated to the high electrical conductivity values measured. Sodium and chloride are the common constituents of many minerals that are highly soluble [48], and they may be related to evaporitic rocks or prolonged water-rock reactions occurring at high temperatures at depth, resulting in highly mineralized waters; this would explain the high concentrations of dissolved salts [49].
The main source of fluoride in thermal springs is probably the magmatic granite and gneisses forming the bedrock complex [10,50]. As shown in Appendix B, the majority of the thermal springs show an equilibrium with fluoride, which is a common element widely distributed throughout the Earth's crust. High concentrations of SO 4 2− are also observed in this bedrock context, and could have originated from the deep circulation and oxidation of sulfide minerals from the Earth's crust [8,10,48].

Water Quality
Regarding the water quality, the high turbidity values of the majority of springs indicate a large amount of particulate matter, which can originate from microorganisms and/or fine inorganic material, and is unfit for human consumption. Every spring that has a turbidity higher than the class A limit does not have any protective measures at its emergence, with immediate access for communities, livestock and wild animals. Livestock breeding might be the main cause of these high turbidity values. Apart from these high turbidity values, all the inorganic constituents measured in spring waters are within the limits set by the Namibian drinking water standards, with the groundwater classified as having excellent and good water quality.
In all thermal waters that have poor and very poor water quality in terms of the WQI range, fluoride and sodium concentrations are above the class A limits (1.5 mg/L and 100 mg/L, respectively), and chloride and sulfate concentrations also fall into classes B and C in some thermal springs. At these levels, these ions do not really present a threat for human consumption, but they could affect the taste, odor or appearance of the water [12].
No elevated concentrations are found for parameters typically influenced by human activities (e.g., nitrate or potassium). Due to the very low population density in the study regions [20,22], the impact of anthropogenic activities is low; this is consistent with the observed low nitrate concentrations. In other parts of Namibia, over the past few decades, nitrate contamination due to the use of fertilizers and/or feces contamination from livestock has been documented [7,11,[51][52][53].

Recommendations for Yield and Quality Improvement, and Sustainability of the Resource
As they do not require the installation of any equipment, the springs certainly represent the cheapest groundwater resource, but their yield and durability remain uncertain. A water governance and management plan is then essential for the sustainable, equitable and affordable governance of this resource; recommendations to support these resources are given below.
Protection plan and capture improvement. The springs indicate the presence of nearsurface groundwater that is more vulnerable to contamination from the surface. Pollution may be a threat to human consumption. Emergence points with some protective and extraction measures (for example, a hand pump) allow easier water extraction and have better physical and chemical quality according to our study. Covering the direct discharge area and, when possible, deepening the pond and/or lining it will improve both water extraction and water quality. When the springs are the only resource for local communities, especially close to urban zones, an upstream borehole is an alternative in case of a decrease in the groundwater level or the seasonal dry-up of the spring. This would require some investment and operational costs, but it is also a safer option for drinking water, avoiding most common sources of contamination.
Basic treatment and water valorization. For safe human consumption, water should generally be disinfected, using either chlorination or UV treatment. Appropriate filtration methods should be employed to reduce the turbidity. The filtration of particulate matter will reduce the turbidity, reduce the number of microorganisms and improve the effectiveness of disinfection [12]. In this study, bacteriological components were not measured, but other studies [43] have revealed contamination problems that could be solved using these methods.
The thermal springs do not meet the potability requirements for human consumption, as they are highly mineralized and unfit for human consumption according to the guidelines from the WHO. However, the geothermal springs are often valued for other benefits, e.g., the health aspects of certain minerals, and they are used in medicine or for recreational purposes due to their high mineralization. Further studies are recommended from the point of view of thermalism, as mineral or medicinal water could add value to these springs and could provide a source of income for local communities.
Monitoring and surveillance. Surveillance is an important element in the improvement of the quality and quantity of the drinking water supply [12]. Water quality monitoring of both geochemistry and bacteriology (E. coli, Enterococcus, etc.) should be performed annually. The monitoring of the water level and discharge will also help to prevent a decrease in the groundwater level and the drying up of the spring.
Promoting public awareness. To inform local communities about the vulnerability of groundwater resources and the risks of human consumption, it is essential to inform them about the impact of their practices on water resources to allow them to control water abstractions and protect them from contamination, especially bacteriological contamination, to ensure safe consumption.

Conclusions
This study focused on the springs in northwestern Namibia. One of our first observations was the perennial nature of the springs, even during the dry season and during one of the most severe drought episodes recorded in Namibia in the last 90 years. This could indicate a potential permanent resource.
The major Ca-Mg-HCO 3 facies defining the majority of the springs were characteristic of the main dolomitic aquifers of the Otavi Group and the groundwater calcrete. The thermal springs showed Na-Cl facies, which may be the result of water-rock reactions at high temperatures at depth.
The groundwater quality of these springs was evaluated using the WQI method; the measured values were compared with the Namibian Standards for the Evaluation of Drinking Water for Human Consumption and the guidelines from the World Health Organization (WHO). The majority of the springs visited in Kaokoland had excellent and good groundwater quality for drinking purposes. Nevertheless, the majority of the thermal springs had a poor water quality with respect to the Namibian Standards, but they should be considered for alternative activities (thermalism, balneotherapy). The turbidity parameter represents the main element with values above the potability threshold, due to particulate matter and potentially the presence of microorganisms. In general, fluoride, sulphate, chloride and sodium concentrations in thermal springs are higher than the reference values from the Namibian Standards for drinking water. The saturation index table presented in Appendix B reveals that geogenic processes are the main factor affecting the groundwater quality in the study area. The very low population density can explain the low concentrations of chemicals typically related to human activities, such as nitrate and potassium. Proper filtration and treatment methods should be developed to decrease these physico-chemical parameters and to limit the products of pollution to acceptable concentrations for human consumption.
Due to climate change, more meteorological extremes are predicted, including increases in the duration and frequency of drought events; this may lead to the scarcity of water resources, especially in semi-arid countries such as Namibia. Thus, the use of groundwater, which is more resilient to hydrological extremes over longer time periods than surface water resources, will become more important, and if properly managed, monitored and protected, the springs in northwestern Namibia could become a true lifeline for local communities.
Some other basic recommendations, such as deepening the springs, installing concrete enclosures and covers and installing hand pumps, would considerably help to improve spring capture and to limit the direct pollution of the emergence point at a low cost. Monitoring and public awareness are essential to preserving both the quality and sustainability of the resource.
Due to sanitary conditions, the data were only collected during one mission. This study did not take into consideration the temporal variations of the inorganic parameters of groundwater. Even for arid countries such as Namibia that only experience two seasons (a dry season and a rainy season), it has been proven in the literature that some parameters can vary seasonally [7,15]. Despite the fact that this mission was undertaken during severe drought episodes, the study does not show evidence of temporal changes in the discharge flow. It would be interesting to further investigate the temporal variations of the dewatering flow rate, as well as the chemical parameters, to highlight a potential seasonal effect.  Acknowledgments: The authors would like to thank the Department of Water Affairs of the Ministry of Agriculture, Water and Forestry for sharing their data. The authors would like to thank the reviewers for all their constructive and helpful comments on this manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A
PHREEQC interactive V3.4 [54] was used to determine the saturation indices (SIs) and the equilibrium with different minerals in the groundwater. The results are shown in the following table. A positive SI indicates over-saturation of the mineral and a negative SI indicates under-saturation. An SI of 0 ± 0.15 is considered to indicate equilibrium conditions (in green). Sustainability 2022, 14, x FOR PEER REVIEW 16 of 18 Figure A1. Samples were taken from 5 February to 10 February 2019. Very low precipitation values were registered at Kaoko Otavi Station (see Figure 2 for its location) three days before.  Figure A1. Samples were taken from 5 February to 10 February 2019. Very low precipitation values were registered at Kaoko Otavi Station (see Figure 2 for its location) three days before.