Padrões climáticos de precipitação e a produção de soja na Amazônia

Abstract

In this research we studied the modulation of precipitation through ocean-atmosphere phenomena of large scale active in the Amazon region adjacent oceans and their intensities were known as represented by climate indices (SOI, PDO, NAO) numerical indicators. From a series of 32 years (1979-2010) of rainfall data in the GPCP grid points, we applied Empirical Orthogonal Function not rotated for monthly and annual scale, where spatial and temporal patterns and rainfall were obtained. The monthly EOF1 (58%) had a southern pattern associated rainfall in most part, to the action of synoptic scale phenomena. The monthly EOF2 (21%) also showed different patterns of precipitation, but orthogonal to the first, reflecting the effects of the cold phase of the North Atlantic Oscillation. However, in the monthly EOF3 (5%) the spatial pattern is represented by a zone precipitation pattern, reflecting the effect of ENOS and ODP. For the annual scale, both EOF1 (36%), EOF2 (16%) and EOF3 (10%) show zonal and meridional spatial patterns, which also reflect the effects of the Atlantic and Pacific Oceans. The three climate indices studied showed were present in the modulation of precipitation in the Amazon region. Among the municipalities belonging to the Hall of soybeans, Sinop-MT and MT-Smile were the locations that were more sensitive to climate variability in precipitation, both annual averages, as for the average rainy seasons each year analyzed, mainly by because of surface conditions. Itaituba-PA, Santarém-PA and PA-Paragominas showed positive growth trend of rainfall, with the latter located in eastern Pará, showed high correlation between local precipitation and the number of IOS, with a lag of 1 month and 2 months, by selected case studies.