Skip to main content
. Author manuscript; available in PMC: 2021 Feb 1.
Published in final edited form as: Atmos Environ (1994). 2019 Nov 14;222:117130. doi: 10.1016/j.atmosenv.2019.117130

Table A2. Overall performance of exposure assessment methods stratified by distance to closest other ACTIVE monitoring station on each day.

Correlation coefficient between PM2.5 concentration predictions and observed PM2.5 concentration in μg/m3 stratified by distance to closest other ACTIVE monitoring station on each day. The methods consdered are: raw CMAQ output, ordinary least squares (“OLS”), inverse distance weighting (“IDW”), universal Kriging(“UK”), downscaler, random forests (“RF”), support vector regression (“SVR”) and Neural networks (“NN”). Methods use either CMAQ and/or other geographic covariates (“Covs”). The last line in the table reports the mean and standard deviation, in parenthesis, of PM2.5 concentration observed at monitoring sites within each substratum.

Method Closest Station < 50 Miles Closest Station ≥50 Miles
CMAQ 0.53 0.50
OLS (CMAQ) 0.66 0.58
OLS (Covs) 0.75 0.65
OLS (CMAQ+Covs) 0.69 0.57
IDW 0.88 0.72
UK (CMAQ) 0.89 0.77
UK (Covs) 0.88 0.76
UK (CMAQ+Covs) 0.88 0.71
Downscaler (CMAQ) 0.89 0.76
RF (CMAQ + Covs) 0.75 0.64
SVM (CMAQ + Covs) 0.81 0.68
NN (CMAQ + Covs) 0.80 0.70
PM2.5 10.07 (6.12) 9.36 (6.56)