Replication Data for: A Machine learning approach for filling long gaps in Eddy Covariance time series data in a Tropical Dry Forest

These datasets are part of the input and output of the ML approach used to fill longer gaps in EC time series, concerning the manuscript “A Machine learning approach for filling long gaps in Eddy 2 Covariance time series data in a Tropical Dry Forest.” These files represent the Eddy-Covariance flux time series from the Principe flux site at the Santa Rosa National Park-Environmental Monitoring Superior Site (SRNP-EMSS). ‘SRNP_Principe 2013-2022-_timeseries - CO2.csv’ is the carbon flux (µmol/m2/sec) at 30-minute intervals, including gaps filled in first stage (using MissForest ML for shorter gaps), and second stage (using Prophet model for longer gaps). ‘SRNP_Principe 2013-2022-_timeseries – H.csv’ is the sensible heat flux (W/m2) at 30-minute intervals, including the outputs from the first and second stage ML time series gap-filling. ‘SRNP_Principe 2013-2022-_timeseries – LE.csv’ is the latent heat flux (W/m2) at 30-minute intervals, including the first and second-stage ML time series gap-filling outputs. ‘SRNP_Principe 2013-2022-_timeseries – RH.csv’ is the relative humidity (%) at 30-minute intervals, including the outputs from the first and second stage ML time series gap-filling. ‘SRNP_Principe 2013-2022-_timeseries – Tair.csv’ is the air temperature (℃) at 30-minute intervals, including the outputs from the first and second stage ML time series gap-filling. Finally, ‘SRNP_Principe 2013-2022-_timeseries – VPD.csv’ is the Vapor pressure deficit (kPa) at 30-minute intervals, including the first and second-stage ML time series gap-filling outputs.