Modelling variation of rainfall with altitude in Usambara Mountains

To estimate precipitation properly, it is necessary to have optimally distributed locations Of rain gauges, and to apply an appropriate technique for spatial distribution estimation. However, most cases, Me network of the rainfall measuring stations is sparse and available data are insufficient to characterize the highly variable rainfall spatial distribution. A study was canned out to estimate and model the variation of rainfall with altitude in the Usambara Mountains. It was observed that rainfall did not show a strong correlation with altitude in some cases hence it was necessary to first subdivide the region into sub regions (West or Elevated and East or coastal areas Usambara) so as to capture the spatial pattern of rainfall at a finer spatial scale and secondly introduce a variable that would improve the model efficiency. In this case distance from the sea was introduced as a second independent variable that was hypothesized to affect the rainfall spatial variation. The result of this approach was a development of rainfall interpolation models for Mean Annual Rainfall (MAP) and seasonal rainfall. Performances of the developed models were compared using the explained variance criteria (R2 %). This comparison resulted into an output of five (5) recommended models, one for MAP and one for each of ne four seasons with R2 varying 6om 55% to 97%. The recommended models were applied on a 6km x 6km grids developed from the 1km x1km DEM of Usambara Mountains and used to develop isohyetal maps of M.AP and seasonal rainfall for each season. The study has provided a description of the spatial pattern for annualand seasonal rainfall in Usambara Mountains using models that account for spatial and typographical parameters.