AOT40 is the present European standard to assess whether ozone (O-3) pollution is a risk for vegetation, and is calculated by using hourly O-3 concentrations from automatic devices, i.e. by active monitoring. Passive O-3 monitoring is widespread in remote environments. The Loibl function estimates the mean daily O-3 profile and thus hourly O-3 concentrations, and has been proposed to calculate AOT40 from passive samplers. We investigated whether this function performs well in inhomogeneous terrains such as over the Italian country. Data from 75 active monitoring stations (28 rural and 47 suburban) were analysed over two years. AOT40 was calculated from hourly O-3 data either measured by active measurements or estimated by the Loibl function applied to biweekly averages of active-measurement hourly data. The latter approach simulated the data obtained from passive monitoring, as two weeks is the usual exposure window of passive samplers. Residuals between AOT40 estimated by applying the Loibl function and AOT40 calculated from active monitoring ranged from +241% to -107%, suggesting that the Loibl function is inadequate to accurately predict AOT40 in Italy. New statistical models were built for both rural and suburban areas by using non-linear models and including predictors that can be easily measured at forest sites. The modelled AOT40 values strongly depended on physical predictors (latitude and longitude), alone or in combination with other predictors, such as seasonal cumulated ozone and elevation. These results suggest that multi-variate, non-linear regression models work better than the Loibl-based approach in estimating AOT40. (C) 2014 Elsevier Ltd. All rights reserved.
New functions for estimating AOT40 from ozone passive sampling
Paoletti E
2014
Abstract
AOT40 is the present European standard to assess whether ozone (O-3) pollution is a risk for vegetation, and is calculated by using hourly O-3 concentrations from automatic devices, i.e. by active monitoring. Passive O-3 monitoring is widespread in remote environments. The Loibl function estimates the mean daily O-3 profile and thus hourly O-3 concentrations, and has been proposed to calculate AOT40 from passive samplers. We investigated whether this function performs well in inhomogeneous terrains such as over the Italian country. Data from 75 active monitoring stations (28 rural and 47 suburban) were analysed over two years. AOT40 was calculated from hourly O-3 data either measured by active measurements or estimated by the Loibl function applied to biweekly averages of active-measurement hourly data. The latter approach simulated the data obtained from passive monitoring, as two weeks is the usual exposure window of passive samplers. Residuals between AOT40 estimated by applying the Loibl function and AOT40 calculated from active monitoring ranged from +241% to -107%, suggesting that the Loibl function is inadequate to accurately predict AOT40 in Italy. New statistical models were built for both rural and suburban areas by using non-linear models and including predictors that can be easily measured at forest sites. The modelled AOT40 values strongly depended on physical predictors (latitude and longitude), alone or in combination with other predictors, such as seasonal cumulated ozone and elevation. These results suggest that multi-variate, non-linear regression models work better than the Loibl-based approach in estimating AOT40. (C) 2014 Elsevier Ltd. All rights reserved.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
