The African and LAC Flood and Drought Monitors

The African Flood and Drought Monitor--using available satellite remote sensing and in-situ information, a hydrologic modeling platform and accompanying web-based user interface--has been developed, in collaboration with UNESCO, for operational and research use over Africa.  It is based out of Princeton University and is available in English, French, Spanish, Arabic and Chinese. A Latin American Flood and Drought Monitor has also been available since 2015.

As described in a Bulletin of the American Meteorological Society article, the AFDM is based on macro scale hydrologic modeling, ingests available data to provide a real-time assessment of the water cycle and drought conditions, and puts this in the context of the long-term record back to 1950. The data is made available online for drought research and operational use to augment on-the ground assessments of drought.

The monitoring system comprises two parts:

i) First, a historic reconstruction (1950 – 2008) of the water cycle forced by a merged reanalysis/observational meteorological data set; this forms the climatology against which current conditions are compared.

ii) Second, a real-time monitoring system (2009 – present) driven by remote sensing precipitation and atmospheric analysis data that tracks drought conditions in near real-time.

The historic and real-time data are calculated using the Variable Infiltration Capacity (VIC) land surface hydrological model (Liang et al., 1994), which is run at a daily time step and ¼ degree spatial resolution for the whole of Africa. For the historic reconstruction, the model is forced by meteorological data derived from a blending of reanalysis (NCEP/NCAR) and gridded observation-based datasets including: the Climate Research Unit TS3.1 monthly precipitation and temperature data set, the NASA Tropical Rainfall Measurement Mission (TRMM), the monthly gridded precipitation and temperature data set of Willmott and Matsura, and the Global Precipitation Climatology Project (GPCP) monthly data set (Sheffield et al., 2006).

For the real-time monitor, the VIC model is forced by a mixture of observations and modeled/remotely sensed meteorology to produce updates of water cycle variables (e.g., soil moisture, runoff, and evapotranspiration).

Daily mean wind speed and daily maximum and minimum temperatures are taken from NOAA’s Global Forecast System (GFS) analysis fields, which ingest data from multiple sources including remote sensing and in-situ observations in real-time (Parrish and Derber, 1992); this is a reliable source for real-time data over large-scales.

Daily precipitation comes from the NASA TRMM Multi-satellite Precipitation Analysis (TMPA) data set (Huffman et al., 2007) when available, otherwise from GFS. Both TMPA and GFS are bias-corrected to ensure consistency with the historical meteorological data set.

Drought is estimated primarily in terms of low soil moisture, which is given as a drought index based on soil moisture percentiles. The index is calculated by determining the percentile of the daily average of relative soil moisture at each grid cell with respect to its empirical cumulative probability distribution function provided by the historical simulations (1950 – 2008). The drought index (and all hydrological variables and meteorological forcings) is available for the entire record between 1950 and real-time.

Using the daily land surface model output, multiple hydrologic variables and derived drought products are estimated for more than 800 catchments over Africa that correspond to the Global Runoff Data Center (GRDC) network and Food and Agriculture Organization (FAO) reservoir database. The variables include: simulated discharge and basin averaged water cycle variables including precipitation, evaporation, surface runoff, baseflow, and the soil moisture drought index.

The development of the AFDM was one of the accomplishments cited in the awarding of the Sixth Prince Sultan Bin Abdulaziz International Prize for Water (2014) to its developers at Princeton University, Eric Wood and Justin Sheffield.  The AFDM has already been applied to applied studies of drought resilience, impact of irrigation dams, human migration, and health and epidemiology