Optimal groundwater management under uncertain climate and its implication on irrigation water availability in the coastal North-Niayes region of Senegal

In Senegal, irrigated agriculture particularly horticultural crops mostly grown in the Niayes region, have attracted less attention in terms of climate change or variability studies. In the Niayes, farmers use almost exclusively groundwater resource for irrigation needs. Contrary to cereals, the effect of climate on irrigated crops is rather indirect as it mostly affects crop water requirements and irrigation water availability. Research has shown the negative effect of climate on aquifer recharge and depth on localized parts of the Niayes (Aguiar, 2010; DaSylva,2005, 2009). Combined with human use of water resources, climate variability may threaten irrigation water availability.In this paper, we evaluate optimal patterns of farmers' groundwater extraction under climate uncertainty and its implication on irrigation water availability. We also assess the potential gains from improved groundwater management.We use an integrated approach that combines a dynamic hydro-economic optimization model to evaluate farmers' withdrawals and groundwater levels in the myopic and central planner cases; a farm production model calibrated to data from the Niayes region by using the Positive Mathematical Programming approach of Howitt(1995) that embodies a Mitscherlich-Baule endogenous yield function to portray the yield response to water and a first order Markov chain to define a transition probability matrix and project rainfall levels through a simulation model. We use the Standardized Precipitation Index-SPI (McKee et al, 1993) to characterize climate conditions.Results illustrate that in a drought situation, farmers extract less water and aquifer lift is higher. However, the difference between a wet and dry situation is very low. Gains from managing the resource are also very low. Finally, results show that in a drought scenario, farmers tend to decrease the area allocated to crops with somecrops having greater decreases. We establish a baseline for economic efficiency in resource management, by solving an optimization problem which captures the social planner’s decision-making problem under uncertainty and limited foresight. We construct a stochastic dynamic programming model of resource management to maximize the sum of current benefits together with the net present value of future benefits from groundwater extraction for irrigation – which also takes into account groundwater extraction for other usages and the stochastic levels of rainfall that affect aquifer recharge. Results from this forward-looking optimization problem are compared to the myopic optimization behavior that farmers might typically display, under different climate states (normal, wet and dry), in order to assess the gains from improved resource management. We use an agricultural production model that is calibrated to data from this part of Senegal, using the Positive Mathematical Programming approach of Howitt (1995), and taking into account the costs of water extraction. We use the outputs of the agricultural production model to estimate the demand for water within the agricultural sector, and characterize the climate conditions with data on precipitation from the National agency of meteorology. Last but not least, the data on hydrological aspects are drawn from the literature (Gaye, 1990; Faye, 1995; El Faid, 1999; Tine, 2004; DGPRE, 2005, 2009) and the direction of management and planning of water resources (DGPRE) of Senegal. Our results illustrate the value of improved groundwater management in the horticultural sector of Niayes, and suggest the importance of including resource management in the plans for adaptation of agriculture to climate change for this region of Senegal.