12A.2 Modeling Hydrological Regimes of the Logone-Lake Chad River Basin, Africa under CMIP5 GCMs Projections

Wednesday, 10 January 2018: 1:45 PM
616 AB (Hilton) (Austin, Texas)
Asmita Murumkar, Ohio State Univ., Columbus, OH; and M. T. Durand, B. Mark, A. Fernández, M. Moritz, I. Hamilton, S. Phang, S. Laborde, and A. Shastry

The Logone-Lake Chad River basin corresponds to one of the extensive floodplain systems in sub-Saharan Africa, which support human populations of millions. These areas are particularly vulnerable to climate change due to high levels of poverty and low adaptive capacity of its inhabitants. The Logone River is one of the most important tributaries of Lake Chad. In this region, impacts of global climate changes interact with traditional modification of the local hydrology. The work we present is part of a larger, NSF-funded project in which we are investigating the feedbacks between social processes (e.g. pastoralism), economic activities (e.g. river fishing), land-cover changes, and hydrometeorological changes in this region. Small changes in the Logone watershed rainfall have historically produced large variations in the streamflow and annual flooding on the Logone floodplain. The aim of this study was to model and analyze the future hydrologic regimes to better understand the long-term climatic change in the Logone-Lake Chad basin.

NASA Earth Exchange Global Daily Projections (NEX-GDDP) data (precipitation and minimum and maximum temperatures) were used as inputs of multi-GCMs of CMIP5 for two RCP scenarios (viz. medium mitigation RCP 4.5 and high emission RCP 8.5) for duration 2006-2100. Macro-scale hydrological Variable Infiltration Capacity (VIC) model was used to simulate the projected changes in seasonal and annual streamflow as affected by changes in precipitation and temperatures in the Logone-Lake Chad river basin. About 13 years of historical monthly and annual streamflow data observed at Bongor station (250 km upstream of basin outlet) were used for calibration and validation of VIC model. The model performance was evaluated using Nash-Sutcliffe Efficiency (NSE) and coefficient of determination (R2). Overall, VIC model showed good prediction accuracy for monthly streamflow during calibration (NSE = 0.77, R2 = 0.88) as well as validation (NSE = 0.62, R2 = 0.82). Preliminary results suggest that basin will experience greater future streamflow in response to increased precipitation. However, the magnitude of increase may be dampened by increased evapotranspiration in response to warmer minimum and maximum temperatures. A greater spatio-temporal variability of precipitation and temperatures will also affect the hydrologic regimes of the basin. Performance of multi-model ensemble mean is being examined in simulating the hydrologic regimes using historical observations. Furthermore, uncertainty associated with individual model projections is being analyzed using skill scores. Results based on a comprehensive analysis of modeled hydrologic regimes of the Logone River-Lake Chad basin will be presented. Changes in streamflow may affect the flooding patterns, which shape the vegetation and fish populations in the floodplain, which directly affect the livelihoods of herders and fishers respectively.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner