Models for forecasting airborne Cupressaceae pollen levels in central Spain.

The influence of meteorological variables on airborne Cupressaceae pollen levels in central Spain was analyzed, and prediction models based on polynomial and multiple regressions were used to predict pollen counts throughout the pollen season. The Cupressaceae pollen type was selected in view of both its abundance in the atmosphere of the central Iberian Peninsula (particularly from January to March) and its allergenic importance. Sampling was performed uninterruptedly over a 5-year period, using a Hirst volumetric sampler and the sampling method established by the Spanish Aerobiology Network. Temperature displayed the strongest (positive) correlation with Cupressaceae pollen counts. Polynomial and multiple regression analysis showed that maximum temperature was the most influential variable included in prediction models. The prediction equations obtained for the study period were reasonably satisfactory, accounting for 48% and 59% of the variation in airborne pollen levels.