In this WP, unique rainfall and discharge databases will be statistically exploited, meteorological
forcings of extreme precipitation events will be disentangled, and loss and damage data will be
collected locally and from international data bases for the focal regions. By combining the centurylong
data bases of daily rainfall from KIT-IMK and UoA-IG to one of the largest worldwide and by
correcting the GPM IMERG data set (Hou et al. 2014) using the unique rainfall network around
Kumasi, station-based and gridded estimates for different extreme quantiles, return periods and
accumulation periods (1, 2, 5 and 10 days) will be calculated (Paeth et al. 2011, Engel et al. 2017).
The long daily records will allow for testing the applicability of Clausius-Clapeyron scaling.
Selected extreme events will be investigated dynamically. Prime candidates are coastal extreme
events related to land-sea breeze convection. In addition, new copula-based approaches (Lorenz
et al. 2018) will be further developed to generate precipitation fields in high spatiotemporal
resolution for extreme events as input for HN modelling (WP2). The statistics will be derived from
the Kumasi rainfall and other networks such as the WASCAL and AMMA observatories. Moreover,
a quality-controlled historical discharge database with long-term daily measurements for Guinea
Coastal countries and river basins will be created based on a variety of existing data sources (e.g.
GRDC, GLOWA, SIEREM, AMMA and WASCAL). Case studies of urban (e.g. September 2017
in Kumasi) and widespread river flooding (e.g. 2007, 2010 and 2018) will be conducted that include
ground surveys and interviews to figure out flood extent, impacts and mitigation approaches as
well as available satellite-based information on the extent of the floodplains. In addition, the flood
loss and damage database compiled in WASCAL-1 will be complemented with new information
from the countries of the FURIFLOOD focal areas (Ghana, Togo and Benin). This will be done in
close collaboration with the national Sendai focal points and the national disaster management
agencies to support these countries in reporting on and implementing of the SFDRR.

Deliverable 1: Catalogue of precipitation extremes in high spatial resolution for the WASCAL
countries and flood events including impacts and mitigation measures

References:

Hou, A. Y., and Coauthors, 2014: The global precipitation measurement mission. Bull. Amer.
Meteor. Soc., 95, 701–722, https://doi.org/10.1175/BAMS-D-13-00164.1

Paeth H.,A. H. Fink, S. Pohle, F. Keis, H. Mächel, and C. Samimi, 2011: Meteorological
characteristics and potential causes of the 2007 flood in sub-Saharan Africa. Int. J. Climatol.,31,
1908-1926,DOI: 10.1002/joc.2199

Engel, T., A. H. Fink, P. Knippertz, G. Pante, and J. Bliefernicht, 2017: Extreme precipitation in
the West African cities of Dakar and Ouagadougou – atmospheric dynamics and implications for
flood risk assessments. J. Hydromet., 18, 2937-2957, doi: 10.1175/JHM-D-16-0218.1

Lorenz, M., Bliefernicht, J., Haese, B., Kunstmann, H., 2018: Copula-based downscaling of daily
precipitation fields. Hydrological Processes, 32, 3479– 3494. https://doi.org/10.1002/hyp.13271

In WP2 flood risk indicators are determined through (i) stochastic generation of extreme rainfall,
(ii) flood simulation and hazard mapping and (iii) derivation of indicators of exposure and
vulnerability for the integrative risk assessment. The investigations will be performed for the
FURIFLOOD river basins focusing on Kumasi and Cotonou. We will consider past extreme events
for an in-depth hazard assessment and model validation, as well as so-called design events of
given return periods (e.g. 10yr, 100yr) for generating flood hazard maps. 2D-HN flood modeling is
done based on the high resolution Digital Elevation Model TanDEM-X (Wessel et al. 2017). The
open-source software Telemac-2D (Ata 2017) will be used to include runoff generation in HN
simulations. For past extremes and design events high-resolution sub-hourly precipitation
scenarios based on stochastic approaches of WP1 will be generated as input for 2D-HN modeling
to simulate flood scenarios. Model validation will be performed with observed flood information
such as satellite data compiled in WP1 and additional on-site investigations to gather further
qualitative information of recent flood events (i.e. extent, depth). This combination is a promising
way to identify which magnitudes of hazard are prone to become a serious risk for the population
in the focal areas. In addition, indicators of exposure and vulnerability will be derived. For this,
WASCAL-1 outcomes are used and adapted for the study areas based on a participatory approach
with relevant stakeholders from the FURIFLOOD river basins. The most important indicators will
be selected and integrated into a quantitative and spatially explicit risk assessment, interlinked
with flood scenarios.

Deliverable 2a: Design rainfall events and flood hazard maps including explanatory reports
Deliverable 2b: Risk maps including list of most relevant flood risk indicators

References:

Wessel, B., Huber, M., Wohlfart, C., Marschalk, U., Kosmann, D., Roth, A., 2017: Accuracy
assessment of the global TanDEM-X Digital Elevation Model with GPS data. ISPRS Journal of
Photogrammetry and Remote Sensing, Volume 139, May 2018, Pages 171-182, doi:
10.1016/j.isprsjprs.2018.02.017

Ata, R., 2017: Telemac2d. User Manual - Version 7.2

In WP3 RCM precipitation scenarios from CORDEX, WASCAL-1 and potentially new PRT1
scenarios (e.g. LANDSURF proposal) will be merged in an advanced statistical downscaling
approach that preserves RCM signals and removes biases (e.g. Bárdossy and Pegram, 2011). By
making use of the precipitation database of WP1, we will combine this approach with stochastic
geostatistical-based approaches (e.g. Mamalakis et al. 2017) to provide reliable high resolution
8
RCM precipitation scenarios for West Africa (< 10 km). These improved RCM scenarios will be
the basis to generate maps of projected precipitation extremes and future trends for West Africa
using the statistical methods from WP1. This approach will be complemented by novel pseudoglobal
warming simulations. The German ICON model will be integrated at convection-resolving
resolution for selected extreme events (1- to 2-day and 5- to 10-day) driven by mid-21st century
CMIP6 or new PRT1 scenarios. The approach will be methodologically extended by considering
tropical wave-forced cases at an intermediate resolution using spectral nudging. While this cannot
provide spatio-temporal statistics, it can remediate specific RCM errors and provide answers to
which extent future extreme rainfall amounts can exceed those estimated from RCMs or Clausius
Clapeyron scaling arguments. Future flood hazard will be estimated based on the WP2 flood
scenarios conditioned on future precipitation extremes. Potential trends of exposure and
vulnerability indicators will be explored using a participatory, scenario-based approach. This
information will be interlinked with future flood hazard to determine future flood risk and its drivers.

Deliverable 3a: Plausible future extreme rainfall estimates at 1-2 and 5-10-day accumulations
Deliverable 3b: Report on future flood hazard scenarios & roadmap for analysing future risks

References:

Bárdossy, A., Pegram, G., 2011: Downscaling precipitation using regional climate models and
circulation patterns toward hydrology. Water Resources Research, 47, 4,
DOI:10.1029/2010wr009689

Mamalakis, A., Langousis, A., Deidda, R., Marrocu, M., 2017: A parametric approach for
simultaneous bias correction and high-resolution downscaling of climate model rainfall. Water
Resources Research, 53, 3, 2149–2170, DOI: 10.1002/2016wr019578