Estimating the economic added-value of improved seasonal forecasts for the Jucar river basin (Eastern Spain)

16:15 Wednesday 29 May

SS033 • OC196

Room S10


Hector Macian-Sorribes (Spain) 1; Ilias Pechlivanidis (Sweden) 2; Louise Crochemore (Sweden) 2; Manuel Pulido-Velazquez (Spain) 1

1 - Universitat Politècnica de València (UPV); 2 - Swedish Meteorological and Hydrological Institute (SMHI)

Here, we estimate and compare the economic value of several seasonal inflow forecasting systems. As economic value is defined the difference between the economic benefits obtained in a base case and the ones associated with the seasonal forecast analysed. The base case considers the absence of seasonal inflow forecasting systems, and hence the historical inflow average value is used instead. The Jucar river basin (Eastern Spain) is our case study.

The seasonal inflow forecasting systems analysed are:

  1. the base case with historical average values instead of forecasts;
  2. the hydrological forecasts provided by the pan-European E-HYPE model post-processed (bias-adjusted) using fuzzy rule-based systems;
  3. a locally developed stochastic model (ARMA), used by the Jucar River Basin Authority;
  4. a 1-month perfect forecast, in which the streamflow for the following month is known in advance; and
  5. a full-period perfect forecast, in which streamflow is known for the whole planning horizon.

Skill is assessed using the Mean Absolute Error (MAE) and the CRPSS metrics. Then, these forecasting systems are embedded into a hydro-economic optimization model of the Jucar river system, built using Stochastic Dual Dynamic Programming (SDDP). The model evaluates the inflow forecasts for the given month and makes an optimal decision in response to them, targeting the maximization of the economic revenues in the system. The comparison between forecasting systems was made for the 1998-2009 period, which includes one of the most severe droughts affecting the Jucar river (2005-2008).

Results showed that E-HYPE seasonal forecasts were more skilful than using the average historical values, but the ARMA outperformed E-HYPE’s skill for the summer period. However, the increase in skill of E-HYPE with respect to the historical average does not turn into increased economic profit since both alternatives show a rather similar revenue level. This is caused by a lack of E-HYPE skill during 2007. Nevertheless, the ARMA model shows an increase in economic profit compared to the historical average and E-HYPE (around 2.50 M€/year), with the perfect forecasts offering the upper bound as expected. It can be concluded that the ARMA yielded the highest benefits due to being the most accurate inflow forecasting system in the summer months.


This study has been supported by the European Union’s Horizon 2020 research and innovation programme under the IMPREX project (grant agreement no: 641.811) and co-funded by the postdoctoral program of Universitat Politècnica de València (UPV)