MJO Modulation of Station Rainfall in the Semiarid Seridó, Northeast Brazil

Abstract

The influence of the Madden Julian Oscillation (MJO) on station rainfall over the Seridó region of Rio Grande do Norte state, Northeast Brazil is examined based on 17 raingauge daily data over 30-year period (1 January-31 December, 1981-2010). The Seridó is one of the driest regions in Northeast Brazil and is recognized as particularly vulnerable to desertification by the United Nations Convention to Combat Desertification. Firstly, daily anomalies were calculated by removing the 30-year daily climatology. Then, to distinguish the MJO signal from other patterns of climate variability, the daily anomalies were band pass filtered in the frequency domain (20 - 90 days) by applying Fast Fourier Transform (FFT). Composites of rainfall anomalies were computed for each of the eight phases of the MJO during February-May (FMAM) rainy season, based on the Jones-Carvalho MJO index. Only days classified as MJO events were considered in the composites. For each phase composite, statistical significance tests were computed independently at each individual station by applying a two-tailed Student’s t-test at 5% significance level. Preliminary results showed that the rainfall anomalies have a spatial coherence throughout the MJO cycle. Extreme positive (negative) anomalies occurred in phase 2 (phase 5), where 13 (12) out of the 17 stations showed statistically significant anomalies in the range of 0.9 - 1.9 mm/day (0.8 - 1.7 mm/day). The typical difference between the wet MJO phase 2 and dry phase 5 represented at least 50% modulation of the daily mean rainfall.