The effect of climate variability and land use land cover on the runoff characteristic of Limpopo Basin, Botswana.

Estimation of likely changes in runoff characteristic such as runoff coefficients (ROC) for a semi-arid river is a trivial task particularly in the wake of climate change and changes in the land use. To assess this water balance model and a land use based model have been used. Runoff coefficients estimated through these models have been compared to assess the likely contribution due to land use /land cover changes. To assess the future scenario of ROCs with the present land use conditions, an optimal architecture of Artificial Neural Network has been developed and used to forecast ROCs up to 2020. The study undertaken over a period of 1971-1998, revealed that marginal increase in the ROCs was predominantly due to land use land cover contributions, which remained more or less constant since the 1990s. ROCs forecasted using ANN with the present land use conditions, showed that there can be gradual increase in the ROCs, which may be attributed to possible decrease in the rainfall characteristics in the Limpopo basin.