Generating Spatio-Temporal Precipitation Forecast Fields Based on the Output of the Global Ensemble Forecast System (GEFS) (Core Science Keynote)

Hydrological forecasts strongly rely on predictions of precipitation amounts and temperature as meteorological inputs to hydrological models. Ensemble weather predictions provide a number of different scenarios that reflect the uncertainty about these meteorological inputs, but are often biased and underdispersive, and therefore require statistical postprocessing. In hydrological applications it is crucial that spatial and temporal (i.e. between different forecast lead times) correlations are adequately represented by the recalibrated forecasts. We present a study with precipitation forecasts over different river basins in California that are postprocessed with the censored, shifted gamma distribution approach by Scheuerer and Hamill (2015). For modelling spatial and temporal dependence we follow Scheuerer et al. (2017) who propose a variant of the Schaake Shuffle (Clark et al., 2005) that uses spatio-temporal trajectories of observed precipitation as a dependence template, and chooses the historic dates in such a way that the divergence between the marginal distributions of these trajectories and the univariate forecast distributions is minimized. We then discuss how this approach can not just be used for modeling spatio-temporal correlations, but also permits an alternative modeling strategy where probabilistic precipitation forecasts are first generated on a coarser spatial and temporal scale, and then downscaled using the selected historic observation
trajectories. A case study is presented where a combination of the two strategies is used to generate high-resolution space-time forecast fields of precipitation based on the output of the (lower resolution) Global Ensemble Forecast System (GEFS).

References:

Clark, M., Gangopadhyay, S., Hay, L., Rajagopalan, B. and Wilby, R. (2004): The Schaake shuffle: A method for reconstructing space-time variability in forecasted precipitation and temperature fields. Journal of Hydrometeorology, 5(1), 243-262.

Scheuerer, M. and Hamill, T.M. (2015): Statistical post-processing of ensemble precipitation forecasts by fitting censored, shifted Gamma distributions. Monthly Weather Review, 143(11), 4578-4596.

Scheuerer, M., Hamill, T.M., Whitin, B., He, M. and Henkel A. (2017): A method for preferential selection of dates in the Schaake shuffle approach to constructing spatio-temporal forecast fields of temperature and precipitation. Water Resources Research, 53(4), 3029-3046.