Previsibilidade climática do Nordeste do Brasil assistida pelo indicador técnico MACD, com lastro em ciclicidades astronômicas em escalas interanual a multidecadal

Abstract

Time-series of precipitation, collected from five rainfall stations across the Northeast region of Brazil (NEB) – representatives of its different climatological subregions and comprising more than a century of data – were submitted to statistical analysis via the technical indicator/oscillator Moving Average Convergence Divergence (MACD), with recognized efficiency in forecasting trends, momentum and reversals in the financial market, having already been applied to other fields of the natural sciences. Given the versatile nature of this technique, its parameters were easily adapted for implementation in this study, with the purpose of assisting in climate predictability of a region whose most of the area has been historically affected by colossal droughts, but also by the occurrence of slashing floods and landslides in the coastal cities. The periods assigned to the exponential moving averages (EMA) are allusions to astronomical cycles. Two Lunar orbital cycles – the Nodal (18.6-year) and Apsidal (8.85-year) Precessions – as well as the Schwabe Solar Cycle (11-year), were employed. A comprehensive examination of the literature has revealed that there is a growing understanding that lunar orbital variations alter gravitational forces that strengthen or weaken the ocean currents, as well as the tidal ranges, changing the vertical mixing, the heat transport to the poles and, in turn, the sea surface temperatures (SST). Even more abundant in literature are findings regarding the influence of solar magnetic variations in climate variability, with signals of 11-year and 22-year having also been found in the precipitation over NEB. Indeed, the solar magnetism is known to produce effects on the terrestrial climate through several proposed physical mechanisms, such as variations in Total Solar Irradiance (TSI), in the UV spectrum, changes in the electric global circuit and the modulation of the supply of Cloud Condensation Nuclei (CCN) which is modulated by the Galactic Cosmic Ray (GCR) flux. With grounds on the aforementioned cycles, the MACD was assessed regarding its applicability to rainfall data and effectiveness, being proven to be consistently successful for most of the analysed past periods, foreseeing reversals up to three years earlier and trends that would last for about a decade, or even two, in the series of Petrolina/PE and Barreiras/BA. The several graphical parameters that make up the indicator were observed to vary in efficiency. Therefore, a cautious analysis, contemplating all of them, is recommended. Na analogous thought is due to the discrepancies found between the two tested sets of EMA, i.e., they should be preferably consulted simultaneously. Also, an analysis of climatological nature, similar to the fundamental analysis in econometrics, should also be carried on, in order to refine and supplement its technical counterpart. Predictions of rainfall behaviour in the forthcoming years were also formulated for each city. São Luís/MA, Natal/RN and Barreiras/BA should experience wetter rain seasons in the near future, leaving definitely behind the long drought that started in 2012 and culminated with the El Niño of 2015-16. The same can be inferred about Petrolina/PE. The analysis for Maceió/AL was more challenging, possibly owing to conflicting influences from the Pacific and the Atlantic.