Search

Pollens of Poaceae are among the most allergenic pollen in Europe with pollen of birch. It is therefore useful to elaborate models to help pollen allergy sufferers. The objective of this study was to construct forecast models that could predict the first day characterized by a certain level of allergic risk called here the Starting Date of the Allergic Risk (SDAR). Models result from four forecast methods (three summing and one multiple regression analysis) used in the literature. They were applied on Nancy and Strasbourg from 1988 to 2005 and were tested on 2006. Mean Absolute Error and Actual forecast ability test are the parameters used to choose best models, assess and compare their accuracy. It was found, on the whole, that all the models presented a good forecast accuracy which was equivalent. They were all reliable and were used in order to forecast the SDAR in 2006 with contrasting results in forecasting precision.

The pollen of Ambrosia, highly allergenic, can be found in high quantities in the French regions of Burgundy, Auvergne and mostly a great part of the Rhône-Alpes region. The aim of this study was to establish some forecast models of the Starting Date of the Pollen season (SDP) of Ambrosia, in order to provide early information to allergists and allergic people. The pollen data came from the Hirst trap set up in Lyon since 1987 by the RNSA, and the meteorological data for the same period were provided by Météo-France. Two models were used, a sum of daily average temperatures and a multiple regression on all the 10-day period meteorological data. The results were quite accurate, except for 2003 with the temperature sums, probably because of very special meteorological conditions. So the models will have to be updated in this way.

The french network of aeropalynology includes traps very close together and vast areas without any trap as well. A study was launched in order to optimise this network, without the number of traps increasing too much. Five criteria are successively reviewed to rationalise the spatial cover. Equi-distance of sites has to be balanced by the number of concerned people. But the need for having long time series is incentive to delete only as a last resort the traps which are reliable for a long time. Moreover, apart from the general network, some local or regional features make indispensable a trap in such-and-such a place (health resort), or justify a higher density of traps in such-and-such an area (the Rhône-Alpes concerning ragweed, the Mediterranean South concerning cypress).