APPLICATION OF HYDRAULIC FLOODING MODELING FOR PLANT SELECTION AND LAND MANAGEMENT IN THE OUÉMÉ DELTA (BENIN): CASE OF ADJOHOUN DISTRICT

Abstract

This study assessed the causes of human and agricultural losses and damages and presents a unique solution for optimal plant selection and land-use planning. In terms of agricultural potential, the Lower Ouémé Valley is the second richest after the Nile in the world, but due to flooding this potential is being lost or destroyed. The district of Adjohoun is located in this valley and is crossed by the Ouémé River. This research aims to (i) characterize hydroclimatic events and the occurrence of exceptional hydro-rainfall phenomena for return periods of 2 and 5 years, (ii) delimit flood-prone areas to characterize the maximum height of water submersion, (iii) evaluate the best model for statistical forecasting of seasonal flood discharges and (iv) develop a web-based tool for optimal plant selection and assistance with thoughtful territorial planning. Results showed that: (i) the Adjohoun valley is influenced by two sources of flooding capable of destroying crops and other properties; often the rains in July and the increase in the discharge of the Ouémé river during the month of September. Values of rainfall 115 mm/day and 780 m3 /s are expected each year with a 50% probability of occurrence, compared with 153 mm/day and 1150m3 /s for a 5-year return period under Gumbel's distribution; (ii) at the advent of these observations, under a rainfall of 115mm/day about 17.05% (53.92 km²) of the surface area will be affected, compared to 50.87% (160.61 km²) of area for a rainfall of 153 mm/day. Despite a maximum submersion height of around 396.24 cm (60 m²), only 1.81% (3.72 km²) of the valley is unsuitable for the production of plants that do not tolerate a maximum water level below 0.9 m (Solanum lycopersicum, Oryza sativa, Capsicum). The contrary observation is found during river flooding (month of September) when the majority of floods inundate crops on the right bank of the Ouémé River for a proportion of 28.32% (89.81km²) of area with predominant submersion height classes of 30.48 cm (~30 km²); 60.96 cm (~29 km²), 91.44 cm (~15 km²) and 121.92cm (~5 km²). (iii) To alert the valley's population and farmers on a seasonal forecast that describes the extent of flooding, the best statistical prediction model following the current discharge regime is the Holt Winter Exponential Smooth (HWES) model [RMSE=104.97, MAE=60.13, MAPE=27.70]; and (iv) Finally, a web-based assistance tool has been developed for the broadcasting of seasonal forecast information and for 3D visualization of flood-prone areas, giving information on maximum water submersion height of each area, optimal crop type, etc., by using a smartphone/computer and Google Earth. However, we recommend farmers to opt for flood-resistant plants and local authorities to envisage the installation of hydraulic structures to prevent annual flooding.