Motivations
Communities worldwide are impacted by climate variabilities such as droughts or floods. Advance predictions of seasonal precipitation may provide information to aid water resource management decisions in various sectors, yet there is a disconnect between the spatial scale of skillful predictions and the decision-making scale. Thus, we ask: are skillful hydroclimate forecasts at the local, sectoral decision-making scale possible for enhanced value to the end-user?
Forecast Predictors and Framework
Global and local climate variables are correlated with observed June – September (JJAS) precipitation timeseries at varying lead times (October – May) using correlation maps (shown below).
Taken together, these climate variables provide information on regional and global teleconnections that influence moisture flow to the regions as well as inter-annual variabilities in precipitation.
A principle component regression, cross-validated framework is then used to forecast JJAS precipitation for two local regions, Koga and Finchaa, as shown below:
Forecast Verification and Interpretation
Deterministic (correlation) and categorical (Hit Score, Extreme Miss Score, Rank Probability Skill Score) metrics were used to evaluate forecast performance for each local region and forecast issue date. As shown, nearly all forecasts have a positive RPSS score, indicating skill over climatology. Perhaps unsurprisingly, predictions fare better closer to the season of interest, with the inclusion of additional climate variable information.
Observed precipitation (bold) and probabilistic precipitation predictions issued June 1 for the Koga Region are shown below as an example of a hindcast.
Precipitation predictions align modestly well with observed precipitation over the 36 year historical record, showing promise for bridging the gap between the spatial scale of skillful forecasts and the decision-making scale.
For Additional Information:
Alexander, S., Shu, W. and Block, P. Development and Evaluation of Season-Ahead Precipitation Predictions for Sectoral Management in the Blue Nile Basin, Ethiopia. American Geophysical Union (AGU) Fall Meeting, New Orleans, L.A., December 2017.