Forecasting spore concentrations: a time series approach

Area of Science:

• Aerobiology
• Allergen monitoring
• Statistical modeling

Background:

• Fungal spores, specifically basidiospores and Cladosporium, are abundant in Dublin's air.
• These spores possess known allergenic properties, impacting public health.
• Accurate forecasting of spore concentrations is crucial for managing allergies.

Purpose of the Study:

• To develop a predictive model for airborne fungal spores (basidiospores and Cladosporium) in Dublin.
• To assess the efficacy of a simple time series model combined with diurnal rhythm estimation for short-term spore forecasting.
• To quantify the reduction in prediction error variance achieved by the model.

Main Methods:

• Development of a predictive model integrating estimated diurnal patterns with a one-parameter time series model.
• Application of the model to forecast concentrations of basidiospores and Cladosporium.
• Evaluation of model performance by calculating the one-step prediction error variance.

Main Results:

• The developed model provided effective short-term forecasts for both spore types.
• A significant reduction in one-step prediction error variance was observed: 88% for Cladosporium spores.
• An even greater reduction of 98% in one-step prediction error variance was achieved for basidiospores.

Conclusions:

• A simple predictive model combining diurnal rhythms and time series analysis is effective for forecasting airborne fungal spores.
• The model demonstrates high accuracy in short-term prediction, significantly reducing forecasting errors.
• This approach offers a valuable tool for allergen monitoring and allergy management in urban environments.