Author: Erin Coughlan de Perez, Red Cross Red Crescent Climate Centre and FATHUM team member
In order to get the most time for early warning early action (e.g. for Forecast-based Financing projects), we are particularly interested in seasonal rainfall forecasts, which are issued before a rainy season starts and can give an indication of whether the next three months are likely to be unusually wet or dry. However, in our guidance materials for the use of seasonal forecasts, we clearly mention that these are not flood forecasts (see IFRC maproom), because they are only forecasting total rainfall over a three-month period. The rain could come all at once, or it could be nicely spread out over the three months.
In 2008 in West Africa, we successfully used a seasonal rainfall forecast to prepare for flooding, and Red Cross teams were grateful to have taken action before widespread regional flooding happened that year. Therefore, we decided to analyze whether there is a clear link between total seasonal rainfall and flooding – maybe these seasonal forecasts can be a good indicator for flood preparedness.
In our recent paper, we use the Global Flood Awareness System hydrological model to compare the total seasonal rainfall that went into the model with the “floodiness” that came out of it. We analyzed all the modeled rivers in sub-Saharan Africa above a certain size, therefore we did not consider flash flooding.
Our results showed that in many places, the association between total seasonal rainfall and floodiness is not very strong. Especially in wetter regions of West and Central Africa, it seems that more total rainfall in a season only marginally increases the chances of flooding that season. In East Africa and Southern Africa, the relationship was slightly stronger.
Recommendations for disaster managers
Our results show that:
- even if forecasters could say with complete certainty that rainfall would be in the “above normal” tercile category, this doesn’t mean that there is certainty over whether it will flood
- forecasts can usually only give slight increases in the likelihood of “above normal” rainfall, this means that in reality there is even lower certainty over where it will flood.
- so in practice, current seasonal rainfall forecasts are unlikely to ever provide high enough confidence in whether there will be flooding for disaster managers to be comfortable taking flood preparedness actions.
- future work may change that though! Some scientists are working to get more useable information on flood likelihood during El Nino / La Nina events, and based on the current temperature anomalies in the Indian Ocean. Others have been developing seasonal hydrological forecasting systems.
Recommendations for seasonal forecasters
In our paper we explored several alternative options to make seasonal rainfall forecasts more informative for flood preparedness. We have have the following recommendations for institutions providing operational seasonal forecasts:
- Make seasonal total rainfall forecasts for river basins, in addition to the current gridbox forecasts.
- Develop skillful seasonal forecasts of the upper extremes of total rainfall (e.g. top 10th percentile) would already be more useful than terciles for anticipating floods.
- Forecast the number of 3-day extreme rainfall events and seasonal rainfall intensity is more informative for flood preparedness than total rainfall.
- Invest further into seasonal hydrological forecasts, in which seasonal rainfall forecasts are passed through a hydrological model to forecast seasonal flooding, would also be a very useful step forward (see the freely available new GloFAS model for this).
- Continue to provide the valuable information on forecast skill
Coughlan de Perez, E., Stephens, E., Bischiniotis, K., van Aalst, M., van den Hurk, B., Mason, S., Nissan, H., and Pappenberger, F.: Should seasonal rainfall forecasts be used for flood preparedness?, Hydrol. Earth Syst. Sci., 21, 4517-4524, https://doi.org/10.5194/hess-21-4517-2017, 2017.
Available to read for free here: https://www.hydrol-earth-syst-sci.net/21/4517/2017/