variability of water table in high density low income …hydrologie.org/redbooks/a329/iahs publ...

Trends and Sustainability of Groundwater in Highly Stressed Aquifers (Proc. of Symposium JS.2 at the Joint IAHS & IAH Convention, Hyderabad, India, September 2009). IAHS Publ. 329, 2009.

Copyright © 2009 IAHS Press

47

Variability of water table in high density low income groundwater utilization area in Lagos, Nigeria SHAKIRUDEEN ODUNUGA, LEKAN OYEBANDE & IFEYINWA OKEKE Department of Geography, University of Lagos, Lagos, Nigeria [email protected]; [email protected]

Abstract This paper adopts a geo-statistical data analysis technique to model the spatio-temporal drawdown of groundwater in a high density low income area in Lagos. Weekly water level measurements for 12 wells were carried out for dry season periods between November 1995–April 1996 and November 2007–April 2008. The drawdown arising from climate variability/change and urbanization impacts between 1995 and 2008 was estimated using overlay analysis and mapped as digital terrain model (DTM). Average annual drawdown rate of 0.07 m per annum was estimated with a general drawdown of 0.8 m in the groundwater table of the dry season. This resulted in low production of water in most wells during the dry season periods. Annual maintenance of the wells during the dry season becomes imperative and this infringes on the economic wellbeing of the people. A complimentary extension of the surface water distribution network to the area is recommended. Key words water table; low income; Lagos, Nigeria INTRODUCTION

The water table is the depth below which the soil/rocks are completely filled with water; it is the top of the zone of saturation in the porous materials of the Earth (Todd, 1980). It is the contact plane between free groundwater and the capillary fringe. The water table, though rarely horizontal, reflects the surface relief due to the capillary effect in HsoilsH, HsedimentsH and other Hporous mediaH. It does not always mimic the topography due to variations in the underlying geologic structure (i.e. folded, faulted, fractured bedrock) (Mandel & Shiftan, 1981). It rises and falls with increases or decreases in HinfiltrationH. Water from the ground has many advantages for rural and urban people in both developed and developing countries. In some parts of urban areas in developing countries, it is the only source of water supply sustaining densely populated areas. Thus its variability and sustainability are crucial to the overall development of the people (Ogunkoya, 1987), especially in the high density low income area with a marginalized group that has less social power and fewer economic resources and physical capacity to anticipate, survive and recover from the effects of high levels of variability that could be triggered by climate change. The low capacity of the population of the blighted environments to adapt and recover from non-availability of shallow groundwater necessitated the need for research into the variability of groundwater as a function of environmental change. However, in Bariga a high density low income groundwater utilization area of Lagos, Nigeria, groundwater is the major source of portable water. Thus, the variability (spatio-temporal) of the saturation zone of the groundwater (water table) has significant influence on socio-economic activities of the people. Also, the pressure from an ever-increasing population of the area on the groundwater necessitated the need to understand the flow patterns and drawdown rate for sustainable developments of the groundwater resources of the area. This paper examined the spatio-temporal variation in the water table from hand dug wells in Bariga; it determined the draw down rate and flow directions and linked the observed patterns with rainfalls variability. The paper also established the impact pattern of variation of the water table on water use and socio-economic development in the micro environment. STUDY AREA

The study area is Bariga in Somolu Local Government Area (LGA) of Lagos State. Somolu LGA is amongst the 16 LGAs that make up Lagos Metropolis. Geographically, it extends from about

Shakirudeen Odunuga et al.

48

longitude 2°23′E and 3°24′30″E and latitude 6°31′30″N and 6°33′N. It covers an estimated area of about 3.65 km2. Physiographically, it has a low undulating relief with highest elevation not more than 12 m amsl. The land use is a built up area with residential and some commercial development that is less than 10% of the area covered. Major drainage systems are canals. The housing pattern is cluster and predominantly residential (Odunuga et al., 2004). The climate is basically monsoonal and experiences a contrast between dry and wet seasons. Hydrogeologically, it is part of a coastal alluvium low-lying sandy tract of recent deposit with intercalations of silts and clays. It is underlain by Holocene sediments of the Ogun River flood plain. The sediments are sandy, silty and pebbly; and are loose and poorly sorted. Figure 1 shows the study location within Lagos metropolis.

Fig. 1 Study location in Lagos metropolis.

METHODOLOGY

The methodology adopted involves: the network distribution of the monitoring wells being physiographically determined in such a way that out of the 12 monitoring wells selected, four were located on the upper plain of the topography, with elevation range between 6 m amsl and 9 m amsl. Three wells were located at medium height with elevation range between 2 m amsl to 6 m amsl, while five wells were located at the base relief with elevation below 2 m amsl. Although relief was the major factor used to determine the distribution, land use also plays a significant role. Figure 2 shows the location of the wells in the study area. The heights and locations of wells were determined from the topographic sheet. Rainfall data was sourced from Nigeria Meteorological Services (NIMET). The actual rainfall data used was the average record derived from Lagos Roof and Ikeja stations. Both stations are within Lagos metropolis and the subject area is just midway between the two stations. Monitoring of the water table variability using daily measurements in the morning hours before people start drawing water from the wells was carried out between November 2005 and

Variability of water table in high density low income groundwater utilization area in Lagos, Nigeria

49

Fig. 2 Study area and the location of the experimental wells.

April 2006 for the first phase. The same process was carried out between November 2007 and April 2008 for the same wells. The daily measurements were summarized to weekly and monthly measurements. This period is dry season when water table is expected to be at its lowest depth and the people that derived their domestic water from the shallow aquifer normally face more challenges on water supply. Probabilistic regression and correlation models as well as analysis of variance (ANOVA) based on the F test of between group variance and within group variance, a 2-way classification single observation per cell were adopted for determining the spatio-temporal variability of water table within the study area. Equation (1) was used for the stochastic analysis:

Y = a + bX ± e (1) where Y = mean monthly depth below surface in metres, b = slope, a = intercept and e = error term of stochastic factor. X = Elevation in metres (amsl) for analysis spatial variation of water table with relief and rainfall (mm) for climate connections analysis. While the correlation coefficient and coefficient of variability (CV) between static water table and elevation above mean sea level provides the spatial coefficients, the variability of static water table and the mean monthly rainfalls provides the coefficients of water table variability with climate. The monthly spatial patterns were presented using isorithmitic mapping models within the geographic information system (GIS). Average drawdown rates were determined using equation (2):

DDR = (ADT1 – ADT2)/T2 – T1 (2) where DDR = drawdown rate, ADT1 = average drawdown for 1995/1996, ADT2 = average drawdown for 2007/2008, T1 = 1995/1996 and T2 = 2007/2008. T2 – T1 = period between T2 and T1 in years. To determine the flow directions, the equal potential line between the most variable wells from upper, middle and lower crests was determined. The flow direction is perpendicular to the equal potential line. Table 1 gives the descriptions of the well environments.


50

Table 1 Description of well environment. S/N Street location Attitude Surrounding Relief Association Water condition 1 2 Jolaosho St.

Off Ilaje Rd, Bariga

6 m tarred high road and building very close

clear, no odour and tasteless

2 49 Ilaje Road, Bariga

4 m tarred medium road and building very close


3 121 Ilaje Rd, Bariga

1.5 m barred low close to lagoon greenish, no odour and tasteless

4 10 Odunsi St, Bariga

7 m tarred high under a roofing tent clear no odour and tasteless

5 49 Odunsi St, Bariga

4.5 m tarred medium closed to road and clear no odour and tasteless murdy area

6 12 James Esu of Odunsi, Bariga

2 m barred low close to swampy and brownish

7 12 Ogo Oluwa Street, Bariga

8.5 m tarred high close to road and building


8 3 Jide St, Bariga 1 m not tarred low close to building brownish 9 9 Kayode

Adebiyi, Bariga 9 m tarred high close to building clear tasteless and

odourless 10 4 Keke St, Bariga 3 m not tarred medium close to building clear and

odourless 11 27 Ewenla St,

Bariga 1 m not tarred low swampy area greenish, tasty and

smelling 12 2B Alhaji Alimi,

Bariga 0.8 m tarred low floodable area clear and

odourless Source: field work. RESULTS AND DISCUSSION

Dry season water table variability (1995/1996 and 2007/2008)

Questions relating to groundwater geography are complicated and variables to be understood theoretically and analytically are legion. Within these variables two significant drivers of the water table variability were selected for analysis. These are relief and rainfall. Table 2 and Table 3 show the mean monthly static water table (m) below the earth surface for the 1995/1996 period and 2007/2008 period, respectively. The tables show some parametric statistics of the water table variability. In Table 2 the coefficient of variability of the wells located at the higher elevation ranges between 3.42% (well 9) and 11.88% (well 1). Those of the mid-slope ranges between 9.57% (well 2) and 15.73% (well 10). The low elevation wells CV ranges between 9.79% (well 6) and 33.33% (well 11). In Table 3 the coefficient of variability of the wells located at the higher elevation ranges between 5.10% (well 7) and 12.19% (well 1). Those of the mid-slope ranges between 3.05% (well 10) and 28.96% (well 5). The low elevation wells CV ranges between 8.97% (well 8) and 24.94% (well 11). The high variability of the base wells could be attributed to the tidal effect of the lagoon which is just a few metres away from the locations of the base wells. Considering the temporal dimension, the highest CV of 86.32% was recorded in November, while the lowest CV of 77.49% was recorded in February for the 1995/1996 period. During the 2007/2008 period, the highest CV of 69.80% was recorded in February, while the lowest CV of 63.14% was recorded in March. These results reveal that while there might be low variability of individual wells during the dry season, high spatial variability is recorded over the entire study area. However, population pressure and level of the groundwater abstraction plays a significant role. Table 4 shows the results of the correlation analysis between water table (dependable variable Y and relief (independent Variable X) for the 1995/1996 and 2007/2008 periods.


51

Table 2 Mean monthly water table below the earth surface (1995/1996). Nov. 95 Dec. 95 Jan. 96 Feb. 96 March 96 April 96 Mean C.V% Well 1 4.05 4.89 4.89 5.16 5.76 5.64 5.01 11.97 Well 2 1.87 2.28 2.28 2.18 2.19 1.88 2.05 10.73 Well 3 1.09 1.15 1.13 0.84 0.95 0.99 1 13 Well 4 5.62 6.99 7.71 7.79 7.8 7.36 7.11 11.53 Well 5 2.56 2.73 3.73 5.44 5.66 4.68 4.11 29.68 Well 6 1.96 2.35 2.5 2.51 2.49 2.35 2.3 10.87 Well 7 7.51 8.02 8.6 8.64 8.78 8.49 8.29 5.55 Well 8 1.48 2 2.06 2.25 2.19 1.99 1.92 17.19 Well 9 0.93 1.17 1.13 0.99 1.01 1.04 1 14 Well 10 8.16 8.51 8.81 8.4 8.65 8.38 8.43 2.97 Well 11 0.86 1.25 2.2 2.07 2.14 1.71 1.62 33.33 Well 12 0.71 0.9 0.71 0.69 0.88 0.88 0.79 12.66 Mean 3.07 3.52 3.81 3.91 4.04 3.78 C.V% 86.32 81.03 78.48 77.49 78.57 85.67 Source: Field work 1995/1996. Table 3 Mean monthly water table below the earth surface (2007/2008). Nov.

07 Dec. 07 Jan. 08 Feb. 08 March 08 April 08 Mean C.V%

Well 1 4.81 5.42 6.21 6.22 6.68 6.6 5.99 12.19 Well 2 2.26 2.84 2.9 2.87 2.88 2.21 2.66 12.41 Well 3 1.82 1.92 1.9 1.12 1.2 1.4 1.56 23.32 Well 4 6.56 8.05 8.75 8.92 9 8.32 8.27 11.03 Well 5 3.49 3.8 3.78 6.34 6.72 5.57 4.95 28.96 Well 6 2.89 3.32 3.46 3.5 3.49 2.31 3.16 15.07 Well 7 8.49 8.92 9.54 9.65 8.76 9.36 9.12 5.10 Well 8 2.5 2.89 2.95 3.23 3.21 3 2.96 8.97 Well 9 1.78 1.97 1.98 1.65 1.82 1.84 1.84 6.73 Well 10 9.09 9.62 9.91 9.51 9.67 9.29 9.52 3.05 Well 11 1.45 2.1 2.89 2.92 3.1 2.68 2.52 24.94 Well 12 1.67 1.76 1.64 1.9 2.02 2.03 1.84 9.35 Mean 3.90 4.38 4.66 4.82 4.88 4.55 C.V% 69.80 66.18 66.56 65.72 63.19 67.91

Source: Field work 2007/2008. Table 4 Correlation analyses between mean water table and elevation for the two observation periods. Period a b R R2 e 1995/19996 1.74 0.47 0.48 0.23 0.266 2007/2008 2.64 0.47 0.47 0.22 0.275 Of all the drivers identified as being responsible for the temporal variability of water table in Bariga, rainfall was singled out for statistical verification. Table 5 shows the mean monthly rainfall The coefficient of correlation calculated as 0.48 and 0.47 for the 1995/1996 and 2007/2008 periods, respectively, reveals a linear relationship between X and Y, while the coefficient of determination (R2) calculated as 0.23% and 0.22% for the respective periods indicate that only about 23% of the spatial variability of the water table in the study area is attributed to the ground elevation. However, the mean water levels might be an over generalization of the true situation, as the monthly variability reveals a high coefficient of variability.


52

Table 6 shows the regression and correlation models results for rainfall and water table levels. The coefficient of correlation reveals positive relationships and indicates some fairly good relationships in a few places, as well as very weak relationships in many others. The R2 shows that only around well 2 does rainfall account for up to 40% of the temporal variability of water table in 1995/1996 and about 96% in 2007/2008. However, a natural physical phenomenon such as water table variability is unlikely to be an exclusive function of a single variable (rainfall). It is therefore necessary to conduct further investigation to unfold the major drivers responsible for the temporal pattern of water table in the study area. Table 5 Mean monthly rainfall (mm) in Bariga. Period Nov. Dec. Jan. Feb. March April 1995/1996 37.24 25.7 9.71 27.34 65.35 127.21 2007/2008 62 32 5 10 21 77 Source: computed from Nigeria Meteorological Agency records for Ikeja and Lagos Island. Table 6 Correlation models between mean monthly water table and mean monthly rainfall.

1995/1996 2007/2008 r R2 (%) R R2 (%) well 1 0.58 34 0.2 4 well 2 0.64 42 0.96 95 well 3 0.30 9.2 0.03 0.1 well 4 0.07 0.5 0.61 38 well 5 0.37 14 0.11 1 well 6 0.06 0.3 0.96 91 well 7 0.19 3.6 0.37 14 well 8 0.01 0.002 0.51 26 well 9 0.24 6 0.12 1.7 well 10 0.27 7.5 0.84 71 well 11 0.006 0.004 0.5 25 well 12 0.57 32.2 0.24 6.2 However, the null hypotheses that there were no spatio-temporal variations in the water table in the study area were considered true if the row effect (αi) and column effect (βi) are zero.

H0: αi = α1 = α2 = α3 = … α12 = 0 H0: β1 = β1 = β2 = β3 = … β12 = 0

with critical region (I) a = Fr > 1.976, b Fr > 2.39; (II) a = Fr > 2.66, b Fr > 3.383. The computed F (90.67, 12.58) falls in the critical region at both levels of significance (0.05 and 0.01) for hypothesis (I) therefore H0 is rejected and it is accepted that there is a highly significant spatial variation in the water table level in the high density low income groundwater utilization urban area of Bariga. Also the computed F falls in the critical region at both levels of significance (0.05 and 0.01) for hypothesis (II), the H0 that there is no temporal variation in water table in the study area is rejected. It is therefore concluded that there is high temporal variation in the water level within the study area. Water table drawdown

The overlay analysis of the two periods reveals well 4 recorded the highest average drawdown of 1.06 m while well 3 recorded the least average drawdown of 0.52 m during the periods under investigation. The records show exceptional variance whereby well 6 and well 7 recorded –0.04 in


53

April and –0.02 in March, respectively. In these cases, the negative indicates that the aquifer is gaining more, irrespective of the pressure from the groundwater abstraction. However, since these observations only occurred from two out of the 84 cells, it is considered to be negligible and that other factors rather than relief and rainfall also exert some significant influence on the variability of water table around the two wells. Table 7 shows the result of overlay analysis of the two periods. On average, a general drawdown of 0.84 m was recorded for the area with an annual drawdown rate of 0.07 m. If this trend continues, coupled with a continuous increase in population density and housing stock, it might not be too far before water supply from hand dug wells become a nightmare, with low yield from wells during the dry season. Table 7 Result of overlay analysis of the two periods. Nov. Dec. Jan. Feb. March April Total Mean Well 1 0.76 0.53 1.32 1.06 0.92 0.96 5.55 0.93 Well 2 0.39 0.56 0.62 0.69 0.69 0.33 3.28 0.55 Well 3 0.73 0.77 0.77 0.28 0.25 0.41 3.21 0.54 Well 4 0.94 1.06 1.04 1.13 1.20 0.96 6.33 1.06 Well 5 0.93 1.07 0.05 0.90 1.06 0.89 4.90 0.82 Well 6 0.93 0.97 0.96 0.99 1.00 –0.04 4.81 0.80 Well 7 0.98 0.90 0.94 1.01 –0.02 0.87 4.68 0.78 Well 8 1.02 0.89 0.89 0.98 1.02 1.01 5.81 0.97 Well 9 0.85 0.80 0.85 0.66 0.81 0.80 4.77 0.80 Well 10 0.93 1.11 1.10 1.11 1.02 0.91 6.18 1.03 Well 11 0.59 0.85 0.69 0.85 0.96 0.97 4.91 0.82 Well 12 0.96 0.86 0.93 1.21 1.14 1.15 6.25 1.04 Total 10.01 10.37 10.16 10.87 10.05 9.22 Mean 0.83 0.86 0.85 0.91 0.84 0.77 0.84

Fig. 3 Digital terrain model of the drawdown of dry season water level from between 1996/1997 and 2007/2008.


54

Figure 3 Shows the digital surface model for the drawdown of dry season water table between 1995/1996 and 2007/2008. Well 4, 12 and 10 environments recorded substantial drawdown of more than 1 m while the drawdown from other parts of the study area varies from about 0.5 m to 1 m. Water table variability (contour approach)

In November, the groundwater flow is southeast-ward (Fig. 4). There is a turbulent flow with steep gradients to the northwest around wells 7, 9 and 10. The gradient becomes gentle southward, with

Fig. 4 Water table contours in November (1995 and 2007).

Fig. 5 Water table contours in February (1996 and 2008).


55

flow becoming slower but steady. A valley-like shape is created from around well 9 to the south of well 6. This indicates areas of depressions and potential flow direction. Figure 5 shows the water table contours in February. A similar pattern was observed for both 1996 and 2008, except that 1996 is shallower than 2008. By this period, the steep gradient to the northeast has become a small gentle slope with less turbulence flow when compared with the November pattern. Flow direction remained southeast-ward. Figure 6 shows the water table contour in April. Also, a similar pattern was observed for both the 1996 and 2008 variability. The steep gradient and turbulence flow noticed in November have started to build up, not only towards the northwest, as observed in November, but also to the southwest, to a lesser degree. The flow direction remained southeast, perpendicular to the equal potential line, but this time with a more southerly inclination obeying the law of gravity.

Fig. 6 Water table contour in April (1996 and 2008).

Social-economic implication of the observed patterns

Groundwater supply accounted for about 80% of the total water supply within the study area. During the dry season, when the water table is low and most hand dug wells dry, water supply by vendor becomes a common phenomenon, especially around wells 1, 4, 7, 10 and 12. An average household of about five members spent up to 1$ daily on water supply during this period of low water table. The average income of about 40% of the bread winners in the social survey on water supply during the dry season ranges between 65$ and 75$. Therefore, 1$ daily expenses per person on water supply might affect their other expenses. Also, it costs about 12$ to maintain a well by way of deepening its depth to reach the water level during this period. All the 35% respondents that have lived in the area for more than 15 years noted that some 10 years ago, they only needed to deepen their wells by about 0.5 m to 1 m to get to a level where the wells will produce water all year round. However, the same wells now require between 1 m to 2 m achieve the same result. As a way of adapting to the period of low water table and low productivities from hand dug wells in the area, the people reduced the quantity of water used for various domestic purposes to about 20 L/day/person. This is more than a 50% reduction


56

when compared with the period of high water table of May to October when most people used up to 50 L daily. CONCLUSIONS

The study modelled the water table drawdown in a high density low income urban area of Lagos. It examines the implications of the observed patterns on the social and economic activities of the people. It was observed that while the water table variability allows for excessive/wastage of groundwater resources during the high water table, it checked the usage of this vital resource for tangible human value during its low level. It also increases household expenses and provides seasonal employment for unemployed adults who became water vendors during the low water level. The 0.8 m average drawdown computed resulted in low production of water in most wells during the dry season periods. Annual maintenance of the wells during the dry season becomes imperative, and this infringes on the economic wellbeing of the people. A complimentary extension of the surface water distribution network to the area as a means to reduce poverty level is recommended. REFERENCES Barthel, R. (2006) Common problematic aspects of coupling hydrologic models with groundwater models on the river

catchment scale. Adv. Geosci. J. 9, 63–71. Bates, B. C., Kundzewicz, Z. W. Wu, S. & Palutikof, J. P. (eds) (2008) Climate change and water. Technical Paper of the

Intergovernmental Panel on Climate Change, IPCC Secretariat, Geneva, Switzerland. Becker, M. W., Georgian, T., Ambrose, H., Siniscalchi, J. & Fredrick, K. (2004) Estimating flow and flux of groundwater

discharge using water temperature and velocity. J. Hydrol. 296(1–4), 221–233. Butler, D. (2004) Urban water-future trends and issues. In: Hydrology Science and Practice for the 21st century (ed. by

B. Webb, M. Acreman, C. Maksimovic, H. Smithers & C. Kirby), 232–242. BHS, London, UK. Christoper, B. & Francois D. (1996) ATHYS a hydrological environment for spatial modelling and coupling with GIS. In:

Application of GIS in Hydrology and Water Resources Management (ed. by K. Kover & H. P. Nachtnebel), 572–582. IAHS Publ. 211. IAHS Press, Wallingford, UK.

Mandel, S. & Shiftan, Z. (1981) Groundwater Resources: Investment and Development. Academic Press, New York, USA. Odunuga, S., Anosike, V., Fasona, M. & Tejuoso, O. (2004) Hydrology of a small urban environment. In: Hydrology Science

and Practice for the 21st century (ed. by B. Webb, M. Acreman, C. Maksimovic, H. Smithers & C. Kirby), 313–316. BHS, London, UK.

Ogunkoya, O. O. (1987) Potential groundwater discharge and safe yield of drainage basin in southern Nigeria. J. Africa Earth Sci. 6(6), 773–779.

Todd, D. K. (1980) Groundwater Hydrology. John Wiley, New York, USA. Winter, T. C. (1995) Recent advances in understanding the interaction of groundwater and surface water. Rev. Geophys. 33(s1),

985–429.

variability of water table in high density low income …hydrologie.org/redbooks/a329/iahs publ...

Documents