Abstract
This paper presents an evaluation of the climate forecast in the upper valley of the river Jaguaribe, State of Ceara-Brazil. The forecast was performed using the global circulation model (ECHAM 4.5) boundary conditions to the regional model (RAMS 6.0). The model simulates the monthly scale rainfall in order to aggregate it in the regional rainy season (January to June). We used the Heidke Skill Score (HSS) as a metric of forecast skill. The simulations were performed for the wet season, covering the period from 1979 to 2010. The HSS was calculated for intervals of 1, 2, 10, 15, 21, 30, 45, 60, 120 and 180 days in order to find the Time Interval of Maximum Predictability (ITEMP). The obtained curve, HSS vs. time, presented a maximum HSS of 0.62 for 45 days. This result indicates that climate prediction even with its limitations (initial data of soil moisture and used parameterization) could be an important tool for the reservoir management in Brasilian Semiarid region.
Keywords:
climate forecast; Heidke Skill Score; reservoir management
