Assessing long-term discharges of the Plata River

SAURRAL, R.; MEZHER, R.; BARROS, V.

Foz do Iguaçu

Conferencia; 8th International Conference on Southern Hemisphere Meteorology and Oceanography; 2006

American Meteorological Society

Most of the Plata River discharge originates in the north of the basin, namely over Brazil and Paraguay, where the runoff accounts for only 30% of the long-term mean precipitation volume. Thus, runoffs are very sensitive to changes in either precipitation or evaporation. Since many climate scenarios are projecting several degrees of warming over this region for this century, the enhanced evaporation would also lead to considerable drop in river discharges. Given that the runoff response to 4 or 5°C of warming could be important, a simple model with relative small percent error may be a tool to get a first order assessment of future discharges. The region was divided in 5 sub-basins: the Upper Paraguay (the Pantanal) with closing point at Ladario (Brazil), the rest of the Paraguay basin excluding the Chaco where the runoff is very small, a third sub-basin on the east of Brazil from the sources of the tributaries of the Paraná River to Jupiá (Brazil), another sub-basin from Jupiá to the junction of the Paraná River with the Paraguay River (Argentina) and the Upper Uruguay River basin. The data of precipitation include 3009 stations of Brazil, Paraguay, Uruguay, Bolivia and the northeast of Argentina, while the data of mean monthly temperatures were obtained from NCEP/NCAR reanalysis. The simple model consists in calculating the monthly potential evapotranspiration (PET) utilizing a second degree formula depending on temperature and latitude. This was made for each sub-basin and the runoff was calculated monthly from the balance between precipitation and evaporation. Whenever the available water was greater than the PET, it took its value; if it was lower, then it was equal to the available water. The field capacity and the PET coefficients of each sub-basin were adjusted by iteration to match the long term runoff of the basin with the observed discharge value. For the adjustment phase, the period was 1990-1999, and the validation was made with 1963-1969, 1970-1979 and 1980-1989 data. In the validation periods the errors in the long-term mean discharge of the Plata Basin were less than 11%. This value was considerably lower than the drops in runoff calculated by the model for 2ºC and 5°C warming that would be 21% and 51% respectively.