Monthly Rainfall and its Inter-Annual Variability (February)

Monthly rainfall totals are necessary to many water resources as well as agricultural problems and decisions for which MAPs, or even wet/dry season precipitation totals be they high or low, are of relatively little consequence, because an intra-year distribution of rainfall is required (Schulze, 1997). Monthly rainfall values then serve as an important tool in describing such an intra-year distribution. It should, however, be borne in mind that the use of the calendar month is but a time step of convenience for describing temporal patterns of rainfall, in that it breaks up annual precipitation into components of time long enough to smooth out many of the irregularities of daily rainfalls (Schulze, 1997). Nevertheless, large differences in rainfall can exist from one month to the next. Some of these differences result from major rainfall generating mechanisms changing from one month to the next. How the maps of median monthly rainfall were derived since time series of monthly rainfall are more variable than those of annual rainfall, the raster surfaces of median monthly rainfall at 1 arc minute spatial resolution (i.e. at 1` x 1` latitude/longitude spacing; 1.7 x 1.7 km; with 429 700 raster points making up South Africa) were calculated by expressing the median rainfall value of a given month at each qualifying rainfall station as a ratio of the MAP surface which was generated by Lynch (2004) using Geographically Weighted Regression. These ratios were then interpolated by Inverse Distance Weighting (IDW) onto the rectangular raster of 1 arc minute. This interpolated raster was then multiplied by the raster of MAP values to give 1` x 1` values of that month's median rainfall. The procedure was then repeated for each of the 12 months of the year (Lynch, 2004).