Development and application of runoff model for ater harvesting in North East Nigeria.

Rainfall in the northeast arid zone of Nigeria is limited and has been found to be declining over the last three decades (Hess et al., 1995). This problem of inadequate rainfall is further exacerbated by runoff especially on the degraded lands (fako). Such runoff-prone lands are potential areas for water harvesting. In dry years, harvested runoff water can considerably improve the environmental conditions for plant growth and can make the difference between death and survival. However, water conservation techniques will only show a benefit if the soil is able to hold the extra water within the root zone of the crops. A clear understanding of soil properties and moisture variations on these lands should provide the baseline information needed for applied soil and water management research. Aerial photographs and photo mosaics were used to identify several fako lands out of which three sites were selected for detailed study. The three sites were located at Jawa (12° 48.71’ N, 11° 02.21* E), Zurkaya (12° 49.15’ N, 11° 05.52’ E) and along Dumburi road (12° 54.31!N, 11° 07.49!E). The fako lands have compacted loamy and clay loamy soils with low infiltration rate and hydraulic conductivity. Dry bulk density values range between 1.34 gem¯³ to 1.61 gem¯³ and saturated hydraulic conductivity obtained were between 1 mm/h and 6 mm/h. Water retention characteristic curves revealed that the fako soils have good water holding capacity. Slopes are generally gentle and range from 0.1% to about 1%. Volume of water harvested depends on the runoff yield of an area. Models can be used to estimate runoff on the fako lands. The model to be used will depend upon the available information, the required accuracy and the resolution of the output and the time resources that can be directed at the modelling exercise. As the rainfall data available for the area is in daily time step, a model that can use daily rainfall as input to estimate runoff is required. In order to develop such a model, accurate rainfall-runoff records for several years should be obtained. High-resolution rainfall data for 13 site-years were collected between 1992 and 1994. The EUROSEM model (Morgan et al., 1992) was applied to simulate the rainfall events and partition them into overland flow and infiltrated water. The model was however calibrated and validated before being used for the simulations. For the calibration and validation, a rainfall simulator (USDA, 1972) was constructed and 32 rainfall-runoff events at intensities of between 25 mmh¯¹ and 169 mmh¯¹ were artificially generated on the three sites. A graph of measured versus simulated runoff events showed good agreements in both calibration and validation. Coefficients of determination and efficiency were 0.82 each in calibration and 0.83 and 0.74 in validation. Predicted runoff by the EUROSEM model was regressed against daily rainfall to obtain a linear regression model for predicting runoff from daily rainfall for the fako lands of north-east Nigeria. The model can be regarded as an integral expression of the physiographic and climatic characteristics that govern the relations between rainfall and runoff on the fako areas. Runoff coefficient and threshold value obtained for the area were 0.44 and 16 mm respectively. The linear model was compared to the curve number model and the runoff estimates by both models were similar. The developed linear model was combined with a water balance model, BALANCE (Hess, 1994) and applied to microcatchment water harvesting investigations. The BALANCE model was used to estimate the water balance components for the area. The Ritchie equation in the BALANCE model was calibrated and validated with field data. Measured versus predicted soil moisture plot gave R² values of 0.90 and 0.89 in calibration and validation respectively. Growth of neem tree (Azadirachta indica) on a typical fako land was simulated with different microcatchment sizes in three categories of years (dry, average and wet). Simulation results indicated that augmenting rainfall through runoff water harvesting technique could provide enough water to sustain growth and ensure rapid establishment of the neem tree seedlings. However, due to reduced dry spell some deep percolation may result during the peak of the rainy season (August - mid September) especially in wet years. A microcatchment size of 12m² (basin-runoff area ratio of 1: 2) was found to sustain year round survival of the tree and minimum drainage for all categories of years. Complementary to water harvesting in the conservation measure, effect of three soil cover treatments (bare, perforated polyethylene cover and solid polyethylene cover) on soil evaporation was investigated. As expected, the solid cover was found to have the most effect in reducing soil evaporation but this is not suitable in the current situation because it limits infiltration of rain and free air circulation. The perforated cover treatment is preferred as it can also significantly reduce soil evaporation and at the same time allow unrestricted infiltration of rain and free exchange of gases between the soil and atmosphere.