Examples reproducing the results published in the article:
GONÇALVES, K.S; WINKLER, MS; BARBOSA-BENCHIMOL, PRB; HOOGH, K; ARTAXO, PA; HACON, SS; CHINDLER, C & KÜNZLI, N. Development of non-linear models predicting daily fine particle concentrations using aerosol optical depth retrievals and ground-based measurements at a municipality in the Brazilian Amazon region. Atmospheric Environment 184 (2018) 156-165. DOI: https://doi.org/10.1016/j.atmosenv.2018.03.057
In this study, we applied a non-linear model for predicting PM2.5 concentration from AOD retrievals using interaction terms between average temperature, relative humidity, sine, cosine of date in a period of 365,25 days and the square of the lagged relative residual.
The code:
- dataset_validation.txt :: dataset used in the illustrative examples
- R_validation_Pm2.5.R :: reproduce the results of the illustrative example