The aim of this study was to estimate the probable values of rainfall in the first and second halves of the months and its spatial distribution in Guinea-Bissau. Distribution functions Log-normal 2 parameters were used to estimate the probable values of rainfall adopting the probability level of 75%. The spatial interpolation (kriging) was applied by semi-variogram modeling demonstrating that the probable rainfall estimates have high spatial variability, except in the month of September, when the behavior is homogeneous for both first and second halves. The lowest probable values of rainfalls under 50 mm occur in June and October throughout the country, supporting the use of supplemental irrigation to avoid compromising the periods of sowing (June) and minimize damage at harvest (October, November) for long cycle crops.
spacial variability of precipitation; rainfall; mapping
