Modelling and mapping heavy metal and nitrogen concentrations in moss in 2010 throughout Europe by applying Random Forests models

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00027073%3A_____%2F17%3AN0000093" target="_blank" >RIV/00027073:_____/17:N0000093 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S1352231017301127?via%3Dihub" target="_blank" >https://www.sciencedirect.com/science/article/pii/S1352231017301127?via%3Dihub</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.atmosenv.2017.02.032" target="_blank" >10.1016/j.atmosenv.2017.02.032</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Modelling and mapping heavy metal and nitrogen concentrations in moss in 2010 throughout Europe by applying Random Forests models

Popis výsledku v původním jazyce

This study explores the statistical relations between the concentration of nine heavy metals (HM) (arsenic (As), cadmium (Cd), chromium (Cr), copper (Cu), mercury (Hg), nickel (Ni), lead (Pb), vanadium (V), zinc (Zn)), and nitrogen (N) in moss and potential explanatory variables (predictors) which were then used for mapping spatial patterns across Europe. Based on moss specimens collected in 2010 throughout Europe, the statistical relation between a set of potential predictors (such as the atmospheric deposition calculated by use of two chemical transport models (CfM), distance from emission sources, density of different land uses, population density, elevation, precipitation, clay content of soils) and concentrations of HMs and nitrogen (N) in moss (response variables) were evaluated by the use of Random Forests (RF) and Classification and Regression Trees (CART). Four spatial scales were regarded: Europe as a whole, ecological land classes covering Europe, single countries participating in the European Moss Survey (EMS), and moss species at sampling sites. Spatial patterns were estimated by applying a series of RF models on data on potential predictors covering Europe. Statistical values and resulting maps were used to investigate to what extent the models are specific for countries, units of the Ecological Land Classification of Europe (ELCE), and moss species. Land use, atmospheric deposition and distance to technical emission sources mainly influence the element concentration in moss. The explanatory power of calculated RF models varies according to elements measured in moss specimens, country, ecological land class, and moss species. Measured and predicted medians of element concentrations agree fairly well while minima and maxima show considerable differences. The European maps derived from the RF models provide smoothed surfaces of element concentrations (As, Cd, Cr, Cu, N, Ni, Pb, Hg, V, Zn), each explained by a multivariate RF model and verified by CART, and thereby more information than the dot maps depicting the spatial patterns of measured values. RF is an eligible method identifying and ranking boundary conditions of element concentrations in moss and related mapping including the influence of the environmental factors.

Název v anglickém jazyce

Modelling and mapping heavy metal and nitrogen concentrations in moss in 2010 throughout Europe by applying Random Forests models

Popis výsledku anglicky

This study explores the statistical relations between the concentration of nine heavy metals (HM) (arsenic (As), cadmium (Cd), chromium (Cr), copper (Cu), mercury (Hg), nickel (Ni), lead (Pb), vanadium (V), zinc (Zn)), and nitrogen (N) in moss and potential explanatory variables (predictors) which were then used for mapping spatial patterns across Europe. Based on moss specimens collected in 2010 throughout Europe, the statistical relation between a set of potential predictors (such as the atmospheric deposition calculated by use of two chemical transport models (CfM), distance from emission sources, density of different land uses, population density, elevation, precipitation, clay content of soils) and concentrations of HMs and nitrogen (N) in moss (response variables) were evaluated by the use of Random Forests (RF) and Classification and Regression Trees (CART). Four spatial scales were regarded: Europe as a whole, ecological land classes covering Europe, single countries participating in the European Moss Survey (EMS), and moss species at sampling sites. Spatial patterns were estimated by applying a series of RF models on data on potential predictors covering Europe. Statistical values and resulting maps were used to investigate to what extent the models are specific for countries, units of the Ecological Land Classification of Europe (ELCE), and moss species. Land use, atmospheric deposition and distance to technical emission sources mainly influence the element concentration in moss. The explanatory power of calculated RF models varies according to elements measured in moss specimens, country, ecological land class, and moss species. Measured and predicted medians of element concentrations agree fairly well while minima and maxima show considerable differences. The European maps derived from the RF models provide smoothed surfaces of element concentrations (As, Cd, Cr, Cu, N, Ni, Pb, Hg, V, Zn), each explained by a multivariate RF model and verified by CART, and thereby more information than the dot maps depicting the spatial patterns of measured values. RF is an eligible method identifying and ranking boundary conditions of element concentrations in moss and related mapping including the influence of the environmental factors.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10511 - Environmental sciences (social aspects to be 5.7)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2017

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Atmospheric Environment

ISSN

1352-2310

e-ISSN

1873-2844

Svazek periodika

156

Číslo periodika v rámci svazku

May 2017

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

14

Strana od-do

146-159

Kód UT WoS článku

000399628900015

EID výsledku v databázi Scopus

—