ENSO drought onset prediction in northeast Brazil using NDVI

El Nino Southern Oscillation (ENSO) indices and satellite-recorded Normalized Difference Vegetation Index (NDVI) were used to construct a drought onset prediction model for northeast Brazil (NEB) using a multiple linear regression technique. Monthly NDVI and ENSO indices anomaly data for the period January 1981 to December 1993 were used to develop the model, while those of 1951 to 1998 were used to simulate the NDVI anomaly time series for model validation. Three different regression models were constructed using the NDVI anomaly as dependent variable and various ENSO indices anomalies including: Sea Surface Temperature in the Pacific Ocean area (5 degreesN-5 degreesS and 120 degreesW-170 degreesW, called Nino 3.4), Southern Oscillation Index (SOT), North Atlantic Sea Surface Temperature (NATL), South Atlantic Sea Surface Temperature (SATL) and Dipole 2 (DIP2=SATL-NATL), as independent variables. Model I was constructed using 12-month NDVI data while Models 2 and 3 used data from only four months (September to December). The results showed that R-2 values of 0.38, 0.62 and 0.79 at a significance level of 1% were obtained for Model 1, Model 2 and Model 3 respectively. Simulated NDVI anomaly values agreed quite well with observed values for all three models but Model 3 had a better intensity estimate. The simulated dynamic evolution of the NDVI anomaly of 1951 to 1998 showed that the predicted NDVI anomalies coincided with historical ENSO induced drought events reported in the literature. It is concluded that the use of satellite-recorded NDVI instead of rainfall data improved the correlation with ENSO indices. Drought onset Model 3, based on the dataset with high anomaly values of NDVI and ENSO indices, predicted drought onset in NEB four months before its occurrence with reasonable success (68%). Combined use of ENSO indices and NDVI inferred drought may provide a better alternative to the construction of an ENSO drought onset prediction model for other regions. Further studies will be carried out to investigate the ENSO drought and flood onsets in the southeastern South America.