Forecast Change in Dry and Humid Periods in South America, Using the CORDEX RCM-SMHI with Bias Correction

Abstract:

Extreme precipitation events will be more common in the future due to climate change. In this work, precipitation deviations were computed using the Standardized Precipitation Index (SPI), which is broadly recognized because its simplicity. The SPI associates the precipitation data with standardized deviations, so long series are required to fit the distribution, then transformed into a standard normal variable. The SPI positive values point out precipitation surpluses, while negative values show precipitation deficits.

Monthly precipitation data from Delaware University (UDATP), and the Swedish Meteorological and Hydrological Institute Regional Climate Model (SMHI-RCM), were used. Both databases cover South America with 0.5° spatial resolution, and they have a shared temporary window between 1951-2005. The forecast time window of SMHI-RCM spans between 2006-2100, and the chosen Representative Concentration Pathway is the RCP-4.5.

Prior to calculation of SPI12 (12 months-SPI), a bias correction of RCM precipitation data was done, using the period 1961-1990 as reference. The same period was used as reference for the SPI12; in this way, the forecasted values of SPI12 were compared with those computed between 1961-1990.

The appended figure shows the predicted changes in dry and humid spells in South America. The computed values in the reference period 1961-1990, the forecast period 2071-2100, and the change between them are presented for a) yearly average precipitation, b) number of months with SPI12<-1, and c) number of months with SPI12<+1.

According to forecast, a rise of the number of dry periods are expected in Chile, the Gran Chaco region, the Andean mountain range in Peru, Ecuador and Colombia, and the Amazonian forest between Peru and Colombia. On the other hand, a rise of the number of surplus periods are expected in northeast Brazil, and the Caribbean Sea coast of the continent. This analysis is concordant with the RCM-SMHI forecast of precipitation trends.