17 Strengthening Flood and Drought Risk Management Tools …
393
the Global Runoff Data Centre (GRDC) and (iii) inclusion of representation of Lake
Chad, as the most important hydrological feature of the region, via a component
of the modelling system and using satellite imagery. These are discussed in detail
below.
2.3 Details of Each Component
a. Historic and real-time climate data
Historically, the system is driven by Version 2 of the Multi-Source WeightedEnsemble Precipitation (MSWEP) dataset (Beck et al. 2019) that significantly
enhances Version 1 (Beck et al. 2017). MSWEP is a precipitation product (1979–
present) recently completed with a focus on use in regional monitoring systems, with
the potential to be used globally. Its new (and unique) features include: (i) optimal
merging of an unprecedented broad range of gauge, satellite and reanalysis precipitation products; (ii) high spatial (0.1°) and temporal (three hourly) resolution; (iii)
fully global coverage; (iv) correction for distributional precipitation biases by probability matching; (v) correction for long-term terrestrial precipitation biases using
discharge observations from 13,762 stations around the globe; (vi) incorporation of
daily (rather than monthly) gauge observations from 66,993 gauges around the globe
and (vii) accounting for regional differences in the 24-hour accumulation period of
gauges. Version 2 of the software is described and validated by Beck et al. (2019). A
near-real-time extension has also been developed and is used to force the CHADFDM
operationally.
The VIC model is driven by MSWEP plus other meteorological variables
(minimum and maximum air temperature and wind speed) that are derived in the
same way as is the AFDM, that is, historically they are based on downscaled and
bias-corrected data from the Princeton Global Forcings (PGF; Sheffield et al. 2006),
which is a hybrid reanalysis-satellite-gauge dataset. In real time, they are based
on bias-corrected analysis fields from the US National Oceanic and Atmospheric
Administration (NOAA) NCEP Global Forecast System (GFS).
b. VIC model, lake model and streamflow routing
The VIC model is a land surface hydrological model that was originally developed
as the land component of coupled climate models (Liang et al. 1994), but has evolved
through its use in applied hydrological research to be a fully distributed hydrological
model (Sheffield et al. 2014; Hamman et al. 2018). Given inputs of precipitation
and other near-surface meteorology, the model solves the surface water balance and
estimates the partitioning of precipitation into evapotranspiration, surface runoff,
baseflow (subsurface flow) and the change in soil moisture content. The model is
parameterized in terms of the soil type and land cover spatial fields and associated
soil and vegetation parameters. These have been updated to align with the AFDM.
The updated soil data is from SoilGrids, a global collection of updatable soil property
393
the Global Runoff Data Centre (GRDC) and (iii) inclusion of representation of Lake
Chad, as the most important hydrological feature of the region, via a component
of the modelling system and using satellite imagery. These are discussed in detail
below.
2.3 Details of Each Component
a. Historic and real-time climate data
Historically, the system is driven by Version 2 of the Multi-Source WeightedEnsemble Precipitation (MSWEP) dataset (Beck et al. 2019) that significantly
enhances Version 1 (Beck et al. 2017). MSWEP is a precipitation product (1979–
present) recently completed with a focus on use in regional monitoring systems, with
the potential to be used globally. Its new (and unique) features include: (i) optimal
merging of an unprecedented broad range of gauge, satellite and reanalysis precipitation products; (ii) high spatial (0.1°) and temporal (three hourly) resolution; (iii)
fully global coverage; (iv) correction for distributional precipitation biases by probability matching; (v) correction for long-term terrestrial precipitation biases using
discharge observations from 13,762 stations around the globe; (vi) incorporation of
daily (rather than monthly) gauge observations from 66,993 gauges around the globe
and (vii) accounting for regional differences in the 24-hour accumulation period of
gauges. Version 2 of the software is described and validated by Beck et al. (2019). A
near-real-time extension has also been developed and is used to force the CHADFDM
operationally.
The VIC model is driven by MSWEP plus other meteorological variables
(minimum and maximum air temperature and wind speed) that are derived in the
same way as is the AFDM, that is, historically they are based on downscaled and
bias-corrected data from the Princeton Global Forcings (PGF; Sheffield et al. 2006),
which is a hybrid reanalysis-satellite-gauge dataset. In real time, they are based
on bias-corrected analysis fields from the US National Oceanic and Atmospheric
Administration (NOAA) NCEP Global Forecast System (GFS).
b. VIC model, lake model and streamflow routing
The VIC model is a land surface hydrological model that was originally developed
as the land component of coupled climate models (Liang et al. 1994), but has evolved
through its use in applied hydrological research to be a fully distributed hydrological
model (Sheffield et al. 2014; Hamman et al. 2018). Given inputs of precipitation
and other near-surface meteorology, the model solves the surface water balance and
estimates the partitioning of precipitation into evapotranspiration, surface runoff,
baseflow (subsurface flow) and the change in soil moisture content. The model is
parameterized in terms of the soil type and land cover spatial fields and associated
soil and vegetation parameters. These have been updated to align with the AFDM.
The updated soil data is from SoilGrids, a global collection of updatable soil property