Climate or Land Use?—Attribution of Changes in River Flooding in the Sahel Zone

Valentin Aich,Eva Müller,Stefan Liersch,Jafet Andersson,Fred Hattermann,Tobias Vetter

doi:10.3390/w7062796

Abstract

This study intends to contribute to the ongoing discussion on whether land use and land cover changes (LULC) or climate trends have the major influence on the observed increase of flood magnitudes in the Sahel. A simulation-based approach is used for attributing the observed trends to the postulated drivers. For this purpose, the ecohydrological model SWIM (Soil and Water Integrated Model) with a new, dynamic LULC module was set up for the Sahelian part of the Niger River until Niamey, including the main tributaries Sirba and Goroul. The model was driven with observed, reanalyzed climate and LULC data for the years 1950–2009. In order to quantify the shares of influence, one simulation was carried out with constant land cover as of 1950, and one including LULC. As quantitative measure, the gradients of the simulated trends were compared to the observed trend. The modeling studies showed that for the Sirba River only the simulation which included LULC was able to reproduce the observed trend. The simulation without LULC showed a positive trend for flood magnitudes, but underestimated the trend significantly. For the Goroul River and the local flood of the Niger River at Niamey, the simulations were only partly able to reproduce the observed trend. In conclusion, the new LULC module enabled some first quantitative insights into the relative influence of LULC and climatic changes. For the Sirba catchment, the results imply that LULC and climatic changes contribute in roughly equal shares to the observed increase in flooding. For the other parts of the subcatchment, the results are less clear but show, that climatic changes and LULC are drivers for the flood increase; however their shares cannot be quantified. Based on these modeling results, we argue for a two-pillar adaptation strategy to reduce current and future flood risk: Flood mitigation for reducing LULC-induced flood increase, and flood adaptation for a general reduction of flood vulnerability.

Highlights

Catastrophic flooding in the Sahelian part of the Niger basin has become an increasing threat during the last decades, leading to more than ten million people affected since the year 2000 [1].Tarhule et al (2005) [2] were some of the first to bring the topic into academic research, referring to it as the “other Sahelian hazard”
A LOESS curve with a minimum point is added as a dashed line and the Theil-Sen estimators for the discharge trends are plotted as bold lines, beginning at the minimum of the observed discharge points
The SWIM model showed in this study, like SWAT model in other land use and land cover changes (LULC) studies in the region [33], that the process-based models are generally able to reproduce the effects of LULC

Summary

Introduction

Catastrophic flooding in the Sahelian part of the Niger basin has become an increasing threat during the last decades, leading to more than ten million people affected since the year 2000 [1].Tarhule et al (2005) [2] were some of the first to bring the topic into academic research, referring to it as the “other Sahelian hazard”. Aich et al (2014) [1] recently published a comprehensive overview of flooding characteristics within the entire Niger basin, including a review of existing literature and damage statistics from different sources. They found that the Sahelian part of the Niger basin was affected by catastrophic floods, with an almost exponential increase in people affected over the last decades. They showed that the increasing flood risk was related to the extreme population growth, the increasing vulnerability of the population, and an increase in flood magnitude.

Methods

Results

Discussion

Conclusion