Limits...
A geostatistical approach to assess the spatial association between indoor radon concentration, geological features and building characteristics: the case of Lombardy, Northern Italy.

Borgoni R, Tritto V, Bigliotto C, de Bartolo D - Int J Environ Res Public Health (2011)

Bottom Line: Firstly, we mapped indoor radon concentration in a large and inhomogeneous region using a geostatistical approach which borrows strength from the geologic nature of the soil.Secondly, knowing that geologic and anthropogenic factors, such as building characteristics, can foster the gas to flow into a building or protect against this, we evaluated these effects through a multiple regression model which takes into account the spatial correlation of the data.This allows us to rank different building typologies, identified by architectonic and geological characteristics, according to their proneness to radon.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics, University of Milan-Bicocca, Via Bicocca degli Arcimboldi 8, 20126 Milan, Italy. riccardo.borgoni@unimib.it

ABSTRACT
Radon is a natural gas known to be the main contributor to natural background radiation exposure and second to smoking, a major leading cause of lung cancer. The main source of radon is the soil, but the gas can enter buildings in many different ways and reach high indoor concentrations. Monitoring surveys have been promoted in many countries in order to assess the exposure of people to radon. In this paper, two complementary aspects are investigated. Firstly, we mapped indoor radon concentration in a large and inhomogeneous region using a geostatistical approach which borrows strength from the geologic nature of the soil. Secondly, knowing that geologic and anthropogenic factors, such as building characteristics, can foster the gas to flow into a building or protect against this, we evaluated these effects through a multiple regression model which takes into account the spatial correlation of the data. This allows us to rank different building typologies, identified by architectonic and geological characteristics, according to their proneness to radon. Our results suggest the opportunity to differentiate construction requirements in a large and inhomogeneous area, as the one considered in this paper, according to different places and provide a method to identify those dwellings which should be monitored more carefully.

Show MeSH

Related in: MedlinePlus

Maps of predicted IRC.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC3108118&req=5

f7-ijerph-08-01420: Maps of predicted IRC.

Mentions: In order to produce maps representing IRC on the regional territory, KED and conditional simulations have been adopted. As mentioned above, geological classes were used as a secondary variable in KED for spatial prediction. To produce maps of IRC, a dense regular grid of 2,515 prediction points covering the Lombardy territory was preliminarily built. Figure 6(a) shows the grid used to produce the maps. To derive the geological class necessary for KED prediction, this grid was superimposed to the geological map reported in Figure 4 by a spatial join in a GIS environment. Table 4 shows the number of prediction points by geological classes. As far as the implementation of the kriging prediction is concerned, this requires knowing the variogram which has to be estimated on the data. To this end we detrendized the data by estimating the mean function (2) via OLS first and then estimating the variogram of the residuals. A weighted least square estimate of an isotropic exponential variogram was used for this, see Figure 6(b). Figure 7 shows the surface of IRC as estimated by KED. To evaluate the performance of the prediction method, we adopted a one-leave-out cross-validation technique, removing one sample observation at a time and predicting this observation by the rest of the sample. Although we have not reported the results in detail, we here mention that the cross-validation residuals showed a reasonably good performance of KED since their average was close to zero and their histogram symmetrically shaped. A good correlation between observed and predicted value was also found.


A geostatistical approach to assess the spatial association between indoor radon concentration, geological features and building characteristics: the case of Lombardy, Northern Italy.

Borgoni R, Tritto V, Bigliotto C, de Bartolo D - Int J Environ Res Public Health (2011)

Maps of predicted IRC.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC3108118&req=5

f7-ijerph-08-01420: Maps of predicted IRC.
Mentions: In order to produce maps representing IRC on the regional territory, KED and conditional simulations have been adopted. As mentioned above, geological classes were used as a secondary variable in KED for spatial prediction. To produce maps of IRC, a dense regular grid of 2,515 prediction points covering the Lombardy territory was preliminarily built. Figure 6(a) shows the grid used to produce the maps. To derive the geological class necessary for KED prediction, this grid was superimposed to the geological map reported in Figure 4 by a spatial join in a GIS environment. Table 4 shows the number of prediction points by geological classes. As far as the implementation of the kriging prediction is concerned, this requires knowing the variogram which has to be estimated on the data. To this end we detrendized the data by estimating the mean function (2) via OLS first and then estimating the variogram of the residuals. A weighted least square estimate of an isotropic exponential variogram was used for this, see Figure 6(b). Figure 7 shows the surface of IRC as estimated by KED. To evaluate the performance of the prediction method, we adopted a one-leave-out cross-validation technique, removing one sample observation at a time and predicting this observation by the rest of the sample. Although we have not reported the results in detail, we here mention that the cross-validation residuals showed a reasonably good performance of KED since their average was close to zero and their histogram symmetrically shaped. A good correlation between observed and predicted value was also found.

Bottom Line: Firstly, we mapped indoor radon concentration in a large and inhomogeneous region using a geostatistical approach which borrows strength from the geologic nature of the soil.Secondly, knowing that geologic and anthropogenic factors, such as building characteristics, can foster the gas to flow into a building or protect against this, we evaluated these effects through a multiple regression model which takes into account the spatial correlation of the data.This allows us to rank different building typologies, identified by architectonic and geological characteristics, according to their proneness to radon.

View Article: PubMed Central - PubMed

Affiliation: Department of Statistics, University of Milan-Bicocca, Via Bicocca degli Arcimboldi 8, 20126 Milan, Italy. riccardo.borgoni@unimib.it

ABSTRACT
Radon is a natural gas known to be the main contributor to natural background radiation exposure and second to smoking, a major leading cause of lung cancer. The main source of radon is the soil, but the gas can enter buildings in many different ways and reach high indoor concentrations. Monitoring surveys have been promoted in many countries in order to assess the exposure of people to radon. In this paper, two complementary aspects are investigated. Firstly, we mapped indoor radon concentration in a large and inhomogeneous region using a geostatistical approach which borrows strength from the geologic nature of the soil. Secondly, knowing that geologic and anthropogenic factors, such as building characteristics, can foster the gas to flow into a building or protect against this, we evaluated these effects through a multiple regression model which takes into account the spatial correlation of the data. This allows us to rank different building typologies, identified by architectonic and geological characteristics, according to their proneness to radon. Our results suggest the opportunity to differentiate construction requirements in a large and inhomogeneous area, as the one considered in this paper, according to different places and provide a method to identify those dwellings which should be monitored more carefully.

Show MeSH
Related in: MedlinePlus