A Pearson Product Moment Correlation matrix for four years of daily ra
infall from 96 sites on the island of Mallorca has been used to derive
and map correlation fields and construct composite correlation fields
for different parts of the island based on previously defined daily r
ainfall affinity areas. Significant variations in rainfall organisatio
n over the island are revealed. These reflect the dominance of uplift
and exposure by upland areas in the generation of rainfall, evidenced
in strongly anisotropic correlation fields. Further anisotropy along l
owland coasts normal to the relief trend indicate the role of local ra
infall generation along sea-breeze fronts. Near isotropic fields for c
entral island locations highlight the importance of local convection a
nd local convergence between opposing sea breeze systems and inland co
nvection. Linear correlation distance-decay trends are determined for
the island and for the predetermined rainfall affinity areas, includin
g estimates of distance-decay rates along field axes. A suite of expon
ential cone functions is derived to model the empirically observed fie
lds, representing their typical bidimensional structure. Use of the fi
tted cones to model a 'typical' rainfall day, where the rainfall distr
ibution mimics quite closely the longer term average, yields quite goo
d results and suggests that their use may be more appropriate than app
lying more conventional general interpolation techniques in the mappin
g of spatial rainfall distributions.