Relationship between interdecadal fluctuations in annual rainfall amount and annual rainfall trend in a southern mid-latitudes region of Argentina

An amplifying fluctuation is detected in the time series of annual rainfall, starting about mid-1930s in the province of Córdoba (central Argentina). The timescale of this fluctuation initially had a value approximate to 10 years, and increased to a value of about 20 years. This fluctuation is structured as a train of centers of negative and positive rainfall perturbation alternating in time. A strong positive trend in annual rainfall amount started simultaneously to excitation of amplifying fluctuation. Analyses of contribution from bands of wavelet timescale to reconstruction of time series of perturbation of annual rainfall amount indicate that trend is produced by fluctuations with timescale larger than 10 years. Before 1935, annual rainfall had a stationary mean. After that year, mean annual rainfall in this region is increasing at a rate of 5 mm/year. During the period 1935–1983, the trend produced by contribution from fluctuations with timescale greater than 10 years is 4.6 mm/year. The remaining 0.4 mm/year of trend is produced by fluctuations with timescale smaller than 10 years, and it does not have statistical significance. The amplifying fluctuation in the bands of fluctuations with timescale 10–17 years and 17–27 years, is clearly detected in the 3-month periods November to January, and February to April. These are also the only two 3-month periods with trend statistically significant. Regression analysis of seasonal rainfall on year indicates that there is a discontinuity in trend between the periods 1873–1934 and 1935–1983. In periods November to January and February to April, fluctuations with timescale larger than 10 years produce a statistically significant trend after 1935. Fluctuations with timescale smaller than 10 years do not contribute in a significant way to trend in these seasons. During the 60 years before wave excitation (period 1873–1934), wavelet analysis does not show another amplifying fluctuation of such strong intensity. Finding what triggered this gargantuan amplifying fluctuation in annual rainfall is an important question for understanding multidecadal climate variability.