Hydrological classification of the Besease inland valley bottom in Ghana for crop production using the water table fluctuation method

The shapes and forms of piezometric hydrographs arising from the recharging and discharging of unconfined aquifers offer a hydrological tool for the classifications of inland valley bottoms in Ghana for crop production. A two-year measured water table fluctuations at Besease wetlands were plotted on a reference scale of time in months on the x-axis and hydraulic head on the y-axis. The water table fluctuation method was used to evaluate the seasonal and annual variations in the water level rise and to estimate the groundwater recharge. The monthly slopes segment of the water table fluctuations were used as a base for the classification of the heads. Results from the study showed that the estimated recharge for the study area ranged from 133 to 467 mm for the fourteen (14) piezometers, representing 9 to 31% of 2009 annual rainfall and 47.6 to 427.9 mm in 2010 representing 4 to 34% of the annual rainfall. The Results also showed that most of the piezometers had their monthly slopes dominated by the acute segment followed by the obtuse segment, flat segment and right-angled segment in that order. It can be inferred that most of the piezometric areas dominated by acute forms become relatively dry during the dry season; however, these areas may still have some water to support crops. The hydrograph representation of the monthly slopes were employed to classify the studied Inland Valley Bottoms into three hydrological regimes as a management tool for developing wetlands for crop production. The regimes were Water table fluctuation (WTF) Class I acute slopes segment varying from 0 to 30%, WTF Class II acute slopes segment varying from 30 to 45% and WTF Class III acute slope segment > 45%. It is concluded that a controlled water table offers a distinguishing criterion for the development of Inland Valley Bottoms for year round crop production in Ghana.


INTRODUCTION
Small land holding farmers along the Oda River catchment in Ejisu-Besease in the Ashanti Region of Ghana practice inland valley bottom cultivation as a form of supplementary irrigation.The Government of Ghana through the Agricultural Sector Rehabilitation Programme of the Ministry of Food and Agriculture (MoFA) and the Crops Research Institute (CRI) are actively encouraging floodplain cultivation, which can be practiced in the dry season using pumped water as a source of irrigation.Wetlands hold a lot of water in its phreatic zone which alternate under variable recharge conditions from rainfall, runoff from uplands and rivers and seepage from streams.In the wet seasons, the water tables fluctuate close to the ground surface when rainfall inputs are high.However, they also fluctuate at lower depths in the dry season and at times installed piezometers and wells go dry (Bradley, 2002).
West African inland valley bottoms which are under utilisation for crop production receives surface water through irrigation canal in the dry season.Thus, when the water table is expected to fall, it would rather be rising due to induced groundwater recharge from the canal.On the other hand, water table which is expected to rise would be observed to be falling under a considerable groundwater abstraction due to pumping.Therefore, understanding the temporal and spatial hydrological processes of inland valley bottoms is key to capturing the behaviour of the catchment (Nyarko, 2007).Initial storage, antecedent moisture, volume and intensity of rainfall and surface cover will help further understand the water table fluctuations of unconfined aquifers.The watertable fluctuations even under normal conditions of dry and wet periods or when subjected to management scenarios depicts different forms and shapes of piezometric and well hydrographs which can be used to classify wetlands hydrologic regimes.Raj (2004) classified well hydrographs based on their shapes and forms from slopes made by water table fluctuations of the wells over more than one hydrological cycle in south eastern Peninsular India.He further reported that the differences in the shapes of the hydrographs were attributed to changes in the climatological pattern, reflection of the underlying aquifer characteristics and lithology and also management practices.Ogban and Babalola (2009) also classified inland valley bottoms on the basis of their drainage densities according to their watertable depths in the dry season in Ayepe, South Werstern Nigeria as design criteria for crop production.Knowledge about the shapes and forms of piezometric hydrographs is therefore necessary in the hydrological classification of inland valley bottoms.This aids in the optimal estimation of groundwater table elevation due to an available weather forecast in order to determine an adequate irrigation depth for maximum crop yield (Serrano and Unny, 1987).Kyei-baffour et al. (2013)  This study sought to analyse the hydrological regimes of water table fluctuations in the Besease inland valley bottom catchment as a management tool for crop production.

Study area
Besease is a farming area in the Ejisu-Juaben Municipal District of the Ashanti Region in Ghana.The site lies within Latitude 1°15' N and 1°45' N and Longitude 6°15' W and 7°00' W. The study area covers about 72 ha of the valley bottom lands at Besease (Figure 1).The climate of the study area is mostly related to the semidecedious type.The region is characterised by two distinct seasons, the wet season which begins from March and ends in November while the dry season extends from the month of November to March.The wet seasons can be categorised less than two rainy seasons.
The major rainy season which ranges from mid-March to July and the minor rainy season starts from September to mid-November.The mean annual rainfall is 1420 mm; mean monthly temperature is 26.5°C, the relative humidity ranges from 64% in January to 84% in August.The average monthly maximum and minimum evapotranspiration (ET o ) for the study area were 127.5 mm and 64.7 mm, and has an annual ET o of 1230 mm.The area is drained by the Oda River which is seasonal and whose basin is about 143 km² (Kankam- Yeboah et al., 1997).
The study area is located in the moist semi-deciduous forest zone.Grass species prominently found in the valley bottom are Santrocema trifolia, Chromolaeve ordorata, Imperata cylindrical, Mimosa pigra,Ceiba patendra, Centrosema pubescens and Mariscus flabelliformis.Plant species like Raphia hookeri (Raphia palm), Alstonia boonei, Malotus oppositifolius and Pseudospondias microcarpa extends along the margins of the Oda River.Soils of the Ejisu-Besease can be found in the soil map of Kumasi area.The study area lies in the Offin soil series which are grey to light brownish grey, poorly drained alluvial sands and clays developed within nearly flat but narrow valley bottoms along streams.The series have very slow internal drainage, very slow runoff, rapid permeability and moderate water holding capacity.The geology of the watershed is relatively heterogeneous and mainly composed of phyllites, quartzite, shale, Tarkwain and Voltaiansandstone and limestone.
The phyllites which underlie 59% of the area consist of upper and lower Birimian rocks.Very few rock outcrops were encountered in the survey as the rocks are deeply weathered.The weathered phyllite is soft and easily broken, with recognizable pieces and is *Corresponding author.E-mail: thomas.atta-darkwa@uenr.edu.gh.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License typically found at 2 to 3 m below surface.Soils found within the Oda River catchment are grouped as those derived from granites, sandstones, alluvial materials, greenstone, andesite, schist and amphibolities.Specifically the soils are Orthi-ferric Acrisol, Eutric Fluvisol, Gleyic Arenosols, Eutric Gleysols and Dystri-Haplic Nitisol.The Besease aquifer is composed of heterogeneous sequence of layers which is dominated by sand, clayey sand and silts.The valley bottom is developed by small land holding farmers who cultivate rice in the wet season and also grow vegetables like cabbage, lettuce, sweet pepper, cauliflower, cucumber and okra and other cereals like maize in the dry season when the water table is low.

Groundwater level monitoring
The water table fluctuation method (Meinzer, 1923;Hall and Risser, 1993;Rasmussen and Andreasen;1959;Healy and Cook, 2002;Risser et al, 2005) was used for estimating recharge.This method was based on the premise that the rise in groundwater levels in unconfined aquifers was due to recharging water arriving at the water table.Recharge was calculated using the formula: (1) where, R= Recharge (mm/month), Sy= Specific yield, dh or Δh = Change in water table height (mm), dt or Δt = Time interval (month).
Wetland groundwater level fluctuations was monitored through a network of 14 piezometers installed using a hand auger along a longitudinal and transverse transect at the Besease site as shown in Figure 1.The piezometers consisted of PVC pipes of 7.62 cm diameter screened over the bottom 20 cm with holes of 0.3 cm diameter.The depth of the pipes ranged from 1.8 m to 3 m.Sand was packed around the screens and the rest of the annulus hole was backfilled with auger cuttings and then grout placed on the top to prevent surface water entry.The cup covering the top of the pipes were not hermetically closed to prevent build up of pressure in the piezometer during phases of groundwater rise.Depth to water table was measured for every two days with greater frequency during rain events by inserting a measuring tape down into the piezometers and observing when it encountered the water surface.The elevations of the piezometers were surveyed to benchmarks to allow adjusting the water levels in the wells to the local datum.

Shapes and classifications of piezometeric heads
Two-year measured water table fluctuations were plotted on a reference scale of months on the x-axis and hydraulic head on the y-axis.Hydrographs plotted to this scale were divided into monthly segments and the slope of each segment was then used as a basic element for classification of hydrographs.According to Raj (2004), slopes are classified as flat (segment's inclination < 20°), obtuse (segment's inclination between 20° and 45°), acute (segment's inclination between 45° and 80°), right angled (segment's inclination between 80° and 90°) and homoclinal (when the hydrograph cannot be divided into rising and falling segments and are either rises or falls during one complete water year).Homoclinal segments showing a rise are suffixed as rising, while homoclinal segments that show a fall are referred to as homoclinal-falling.Figure 2

Slopes of piezometric hydrographs
Summarised results from Table 1 show that most of the piezometers had their monthly slopes dominated by  August-September in 2010 water year.P11 and P14 had the lowest acute form of hydrograph representation of 13% and 30%.P12 and P13 had the highest form of acute slopes representing 75 and 61%.Also P11 had the highest obtuse segment of 57% and the lowest (Table 1) was recorded at P12 and P4.The highest form of flat segment was recorded at P14 representing 40% of the slopes in the hydrographs followed by P10 and P4.P3 was the only piezometer which showed a right-angled segment slope.Summing the obtuse and the flat segments revealed P11 and P14 recording 87 and 70% respectively as the highest slopes from the hydrograph presentation.However P12 and P13 had the lowest form of the combination of obtuse and flat of 25% and 39% respectively.The piezometric point twelve (P12) which is closer to the river became empty in most of the dry season possibly due to the lithology of the strata.

Water level rise
When one takes into account all observation    The water level rise measured at P12 and P2 were rather high and may have been influenced by lateral flow due to its close proximity of 33 m and 66 m to the Oda River and P 14 at a low topographic height also experienced a high water level rise (Figure 1).

Groundwater recharge estimation
The groundwater recharge rate for each of the observation wells was calculated by multiplying the water level rise with the specific yield values ( Recharge rate in the month of March 2009 was very high in the entire observational piezometric network.The highest recharge rate of 160 mm occurred in P 14 which is located at a relatively low topographic height (Figure 1) with a shallower water table.However, the water table fluctuation estimated recharge rate increases.One possible reason for this increase in recharge rate may be that it takes proportionately less time for water to travel through a thinner unsaturated zone, thus bringing the water to the saturated zone before it can be transpired by plants.The topographic low height at the site of P 14 coupled with the horizontal movement of subsurface groundwater (West-East) at the location of piezometer 14 gives the field a better point to locate a well to irrigate the field.

Hydrograph representation
Comparison of hydrographs provides an insight into the nature of the aquifer (Raj, 2004).The most common form of the hydrograph is the acute-obtuse slopes of the quarters.The acute slopes suggest a higher fluctuations surge from high rainfall records whiles that of obtuse indicates a lower fluctuation also from a moderate rainfall.Right-angled rising segments or departures towards steeper-rising segments occur due to high rainfall in a short duration, either in one spell or in several closely spaced precipitation events (Raj, 2004).
The shape of the rising limb is influenced by the intensity, interval and duration of precipitation.Rainfall appears to primarily determine the shape of the rising limb, as even a hydrograph of a borehole in a good aquifer in an alluvial tract shows an acute or right-angled (steep) rising limb during a period of particularly good rainfall and conversely well hydrographs tend to be obtuse in a poor raining year which is also a characteristic of the aquifer.A higher percentage of acute recession slope segments with either acute rising segments or right-angled rising segments is due to relatively poor unconfined aquifers.These are observed in aquifers comprising weathered residual of vispar, gneisis and shale which gives off gritty or coarser grained sand.
The dominant flat segment slopes are noted in aquifers with low hydraulic conductivity, high silt and clay, and low topographic site of the study area and aquifers with granites and sandstones rocks weathered to give fine grained sand.The dominant obtuse recession and rising slopes reflect characteristics like that of the flat segment which is often accompanied by poor rainfall amount in a water year.Also linear recession in the flat segment slopes could be due to rapid discharge from the aquifer.In the drying periods the dominant recession slope  segment was the acute which suggest a sharp decrease and a higher fluctuation depth and in most cases leads to the drying of the borehole.

Water table slopes segment fluctuation and areal extent
The small size of the inland valley bottoms (IVBs) is attributed to deeper incision of the landscape and convex nature of the valleys.The amount of water flowing into the valley depends not only on rainfall amount but also on catchment size.Killian and Teissier (1973) have reported that water capture would be too small for a catchment size less than 400 ha.Therefore, the extent of the study area of 72 ha suggest that much of precipitation runs off the surface resulting in flooding of shallow depth and short duration.However, despite the topographic condition and the high drainage density and texture, the rainwater does infiltrate and readily recharges the groundwater, causing a sharp rise in the regional groundwater table and seepage flows, resulting in the seasonal or perennial wetness condition which prevails in Inland Valley Bottoms (Ogban and Babalola, 2009).

Watertable fluctuation and classification of inland valley bottom
Time series of the hydrograph shows seasonal variations in the observations among the boreholes of the study sites.About 14 % of the area had their monthly slopes from 13 to 30% of the acute segment.This was observed in P11 and P14.Also, 14% of the area had their slopes ranging between 40 to 45% acute segment which occurred in P10 and P5.However, 72% of the monthly slopes had their acute slopes greater than 45% in the aerial domain.It can be explained that most of the piezometric areas dominated by acute forms become relatively dry during the dry season; they may, however, still have some water to support crops.These results also point to some management practices in Inland valley bottoms in developing them for crop production.Figure 6 and 8 shows that the water table (WT) is at or near the soil surface for more than 6 months in the valley system IVBs with the monthly segment slopes dominated by obtuse and flat forms of 70 to 87%.This also indicates that with higher records of monthly rainfall amount, water table would fluctuate near the ground surface for a longer period of time in the wetlands.
According to Ogban and Babalola (2009) high watertables fluctuating near the ground surface are also attributed to poor surface and subterranean drainage outlets.Hekstra and Andriesse (1983) and Andriesse (1986) have reported that IVBs in the West African subregion have excellent conditions for more than one crop growing season (150 days or more) especially wetland crops, for example, rice.It can be inferred that wetlands hold a rich potential for food production owing to the availability of water coupled with the fertile nature of the inland valley bottoms.On the other hand, the water table recedes in the beginning of the dry season which varies in the inland valley system accounting for high evapotranspiration rates.Effective cultivation can be enhanced by planting dryland crops like maize, cassava, sweet potato and vegetable crops specifically when the water table (Figure 9) has attained its mean lowest depth in February.High evaporation processes and increase in internal drainage reduces the pore water pressure, extends or pushes the phreatic surface downwards and improve aeration for crops with aerobic edaphic requirements.
Thus, a rise in the level of the WT decreases the zone of unsaturation, increases the pore water pressure, reduces the hydraulic gradient and increases the drainage load, and creates waterlogging conditions that inhibit cultivation and crop growth for dryland crops but enhances rice cultivation.On the other hand, the receding water table reduces waterlogging conditions, or creates unsaturated conditions or re-establishes the agricultural zone of the soils for dryland crops.These alternating conditions explain the alternate fallow and farming in the IVBs (Ogban and Babalola, 2009).Consequently, the decrease of the phreatic surface depth in the dry season is a critical predictive criterion because it defines the effective rooting depth (ignoring the extent of the capillary fringe), the soil water storage depth and drainage requirements, and a distinguishing characteristic for classifying the Inland valley bottoms into soil and water management regimes (Ogban and Babalola, 2009).Three hydrological regimes have been developed to classify wetlands from Figure 3 to 8.These regimes are: WTF Class I Acute slopes segment varying from 0 to 30%, WTF Class II acute slopes segment varying from 30 to 45% WT Class III > 45%.
The distinguishable factors describing the fluctuation classes are as follows.

Conclusions
Developments of Inland Valley Bottoms of unconfined aquifers for crop production have been classified into three hydrological regimes based on the intensity of their acute slope segments of the water table fluctuations.The regimes for Besease and other IVBs are: WTF Class I for acute slope segments varying from 0 to 30% WTF Class II for acute slope segments varying from 30 to 45% WT Class III > 45%.
The results show that most of the piezometric areas dominated by acute forms become relatively dry during the dry season; they may, however, still have appreciable water to support crops.It was also revealed that a rise in the level of the WT decreases the zone of unsaturation, increases the pore water pressure, reduces the hydraulic gradient and increases the drainage load, and creates waterlogging conditions that inhibit cultivation and crop growth for dryland crops but enhances rice cultivation.The lowest recharge rate in the dry periods of the two study years showed a decrease in groundwater storage which lowered the phreatic water level.It meant that irrigation water should be applied to obtain optimal moisture content and water table levels.It is concluded that a controlled water table will offer a distinguishing criterion for the development of Inland Valley Bottoms for a year round crop production.This study could be extended to other IVBs in the remaining agro-ecological zones in Ghana.

Figure 9 .
Figure 9. Example of graphical approach for estimating recharge for P 1.

Table 1 .
Slopes exhibited by the hydrographs.

Table 2
some time to reach the groundwater table.That implied the rainfall infiltrated to replenish soil moisture deficit.
mean annual rainfall.The difference in the recharge values for the two study years could be attributed to the variability in the annual rainfall distribution and intensity.The recharge estimate obtained in this study is similar to estimates from groundwater studies done elsewhere in the world, using the water table fluctuation method.Sibanda et al. (2009) estimated the recharge rate of rainfall takes

table fluctuation class II Acute
slopes segment ranges from 30 to 45%.Water table is intermediate between the ground surface and the base of the borehole and piezometric water table recuperates in March in the dry season.Duration of high water table is about 4 months suitable for wetland (rainy season) and dryland (dry season) crop production but with little soil and water conservation getting to the latter part of the dry season using residue mulch in the middle of the dry season for roots to follow the receding WT.Water table Fluctuation class III Acute slopes segment ranges > 45%.Water table is close to the ground in the wet seasons and most of the piezometric water table recuperates in April throughout the water year.Duration of high water table is about two months suitable for wetland (rainy season) and dryland (dry season) crop production but with soil and water conservation using residue mulch together with early planting for roots to follow the receding WT.