AQ-CNEA-CAC Air quality dataset (2019-2020): "A machine learning approach to address air quality changes during the COVID-19 lockdown in Buenos Aires, Argentina"
Description
This dataset provides hourly air quality data collected from the National Atomic Energy Commission of Argentina (CNEA-CAC: 34° 24' S, 58° 31' W). The data spans from February 2019 to May 2020. The pollutants included are carbon monoxide (CO), nitrogen oxide (NO), nitrogen dioxide (NO₂), sulfur dioxide (SO₂) and ozone (O₃). Meteorological Variables of the Observatory Buenos Aires (OBS: 34° 35' S , 58° 29' W) used to run the Random Forest simulations (Wind speed, wind direction, 2 m temperature, 2 m relative humidity, sea level pressure and calm condition) are also provided. A full list of variables is shown below. "date" Datetime; "NO" nitrogen oxide concentration [ppm]; "NO2" nitrogen dioxide concentration [ppm]; "NOx" nitrogen oxides concentration [ppm]; "CO" carbon monoxide concentration [ppm]; "O3" ozone concentration [ppm]; "SO2" sulfur dioxide concentration [ppm]; "aqsite" air quality site; "t2" 2 m temperature (°C); "rh2" 2 m relative humidity [%]; "slp" sea level pressure [hPa]; "wspd" wind speed [degrees]; "wdir" wind direction [km/hr]; "calm" calm condition; "meteosite" meteorological site; "gas_emcycle" gas emission cycle; "aer_emcycle" aerosol emission cycle; "month" month; "U" U wind component; "V" V wind component; "daynight" [morning, day or night]; "datatype" [training: Data used for training and testing, evaluation: Data used to evaluate model skills (BLD) , lockdown: Data used to estimate relative changes during the lockdown phases (LD and PLD)]. An introductory code to use the data is also provided. The code will be run from the R script AQ_RandomForestCNEA.R. AQ_RF_functions are the auxiliary functions used and RF_parameters.csv is the file you need to provide in order to set the predictors and other parameters needed.