The impacts of the Madden‐Julian oscillation (MJO) on precipitation over Northeast Brazil (NEB, also known as Nordeste) are evaluated based on daily raingauge data from 492 stations over 30‐year period (1981–2010). Composites of precipitation, outgoing longwave radiation and moisture‐flux anomalies are performed for each phase of the MJO, and over all four seasons, based on the Jones–Carvalho MJO index. To distinguish the MJO signal from other patterns of climate variability, daily data are filtered using a 20–90 day band‐pass filter; only days classified as MJO events are considered in the composites. The results show strong seasonality of the MJO's impact on precipitation. The most spatially coherent signals of precipitation anomalies occur in the austral summer, when about 80% of the raingauge stations showed increased precipitation in phases 1–2 and suppressed precipitation in phases 5–6 of the oscillation. Although the MJO impacts precipitation on intraseasonal timescales in all seasons in most locations, these impacts vary in magnitude and depend on the phase of the oscillation. Precipitation anomalies over NEB are explained by the interaction of convectively coupled Kelvin‐Rossby waves with the dominant climatic features in each season. During the austral summer and spring, westerly regimes increase precipitation over most NEB. In the austral winter and fall, precipitation anomalies exhibit more complex spatial variability. In these seasons, precipitation anomalies in coastal areas depend on the strength of the South Atlantic anticyclone, which is largely modulated by Rossby waves. The strengthening of the anticyclone intensifies the convergence of the trade winds in coastal areas and precipitation windward of the coastal range. Conversely, the intensification of the subsidence is responsible for precipitation deficits in the lee side of the range. These conditions are typically observed when easterly regimes dominate over tropical South America and NEB decreasing moisture flow from the Amazon.

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.