ENERGY PROCESSING RESULTS DESCRIPTION

This file includes daily gapfilled Latent Heat (LE) and Sensible Heat (H), an estimation of their random uncertainty, and the version of fluxes where a correction for the energy balance closure has been applied, including estimates of the uncertainties in the corrected fluxes.

The corrected fluxes are obtained multiplying the original, gapfilled LE and H data by an energy balance closure correction factor (EBC_CF), which is calculated starting from the half-hours where all the components needed to calculate the energy balance closure were available -- measured Net Radiation and Soil Heat Flux, and measured or good quality gapfilled Latent Heat and Sensible Heat. The correction factor is calculated for each day as (NETRAD - G) / (H + LE), using the daily averages of these variables, and the timeseries is filtered removing the values that are outside 1.5 times the interquartile range and used as basis to calculate the corrected H and LE fluxes.

The correction factor and corrected fluxes are obtained using three hierarchical methods. If ECB_CF Method 1 fails for a sliding window (i.e., less than five ECB_CF values are obtained), ECB_CF Method 2 is tried. If no ECB_CF is available for being averaged in ECB_CF Method 2 (e.g., in case of long gaps), the ECB_CF Method 3 is used.

- ECB_CF Method 1:
For each day, a sliding window of +/- 7 days is used to select up to 15 ECB_CF values. The resulting ECB_CF from the selected values are then used to calculate corrected versions of the LE and H fluxes: for all days, each ECB_CF value generates one version of corrected H and LE is calculated. The 25, 50, and 75th percentiles of the corrected fluxes are then extracted (LE_CORR25, LE_CORR, LE_CORR75, H_CORR25, H_CORR, and H_CORR75) for each day.

- ECB_CF Method 2:
ECB_CF is obtained from an average of all ECB_CF values used to calculate LE_CORR and H_CORR obtained with ECB_CF Method 1 within a sliding window of +/- 5 days. With ECB_CF Method 2 LE_CORR25, H_CORR25, LE_CORR75 and H_CORR75 are not calculated.

- ECB_CF Method 3:
A sliding window of +/- 5 days is used for the same day in the previous and next years, with the current ECB_CF being obtained from the average of all available ECB_CF values in the window. With ECB_CF Method 3 LE_CORR25, H_CORR25, LE_CORR75 and H_CORR75 are not calculated.


VARIABLES DEFINITION:
LEGEND: HH (half-hourly or hourly), DD (daily), WW (weekly), MM (monthly), YY (yearly)

TIMESTAMP (YYYYMMDD): ISO timestamp - short format

DOY (DDD): Day of Year

LE_F_MDS (W m-2) Latent heat flux, gapfilled using MDS method 
    DD: average from half-hourly data
    WW-YY: average from daily data

LE_F_MDS_QC (adimensional) Quality flag for LE_F_MDS, LE_CORR, LE_CORR25, and LE_CORR75 
    HH: 0 = measured; 1 = good quality gapfill; 2 = medium; 3 = poor
    DD: fraction between 0-1, indicating percentage of measured and good quality gapfill data
    WW-YY: fraction between 0-1, indicating percentage of measured and good quality gapfill data (average from daily data)

LE_CORR (W m-2) Latent heat flux, corrected LE_F_MDS by energy balance closure correction factor 
    DD: average from half-hourly data
    WW-YY: average from daily data

LE_CORR25 (W m-2) Latent heat flux, corrected LE_F_MDS by energy balance closure correction factor, 25th percentile 
    DD: average from half-hourly data
    WW-YY: not produced

LE_CORR75 (W m-2) Latent heat flux, corrected LE_F_MDS by energy balance closure correction factor, 75th percentile 
    DD: average from half-hourly data
    WW-YY: not produced

LE_RANDUNC (W m-2) Random uncertainty of LE, from measured only data 
    HH: uses only data point where LE_F_MDS_QC is 0 and two hierarchical methods (see header and LE_RANDUNC_METHOD)
    DD-YY: from random uncertainty of individual half-hours (rand(i)) = [SQRT(SUM(rand(i)^2)) / n], where n is the number of half-hours used

LE_RANDUNC_METHOD (adimensional) Method used to estimate the random uncertainty of LE 
    HH: 1 = RANDUNC Method 1 (direct SD method), 2 = RANDUNC Method 2 (median SD method)
    DD-YY: not produced

LE_RANDUNC_N (adimensional) Number of half-hour data points used to estimate the random uncertainty of LE
    DD-YY: not produced

LE_CORR_JOINTUNC (W m-2) Joint uncertainty estimation for LE 
    HH-DD: [SQRT(LE_RANDUNC^2 + ((LE_CORR75 - LE_CORR25) / 1.349)^2)]
    WW-YY: not produced

H_F_MDS (W m-2) Sensible heat flux, gapfilled using MDS method 
    DD: average from half-hourly data
    WW-YY: average from daily data

H_F_MDS_QC (adimensional) Quality flag for H_F_MDS, H_CORR, H_CORR25, and H_CORR75 
    HH: 0 = measured; 1 = good quality gapfill; 2 = medium; 3 = poor
    DD: fraction between 0-1, indicating percentage of measured and good quality gapfill data
    WW-YY: fraction between 0-1, indicating percentage of measured and good quality gapfill data (average from daily data)

H_CORR (W m-2) Sensible heat flux, corrected H_F_MDS by energy balance closure correction factor 
    DD: average from half-hourly data
    WW-YY: average from daily data

H_CORR25 (W m-2) Sensible heat flux, corrected H_F_MDS by energy balance closure correction factor, 25th percentile 
    DD: average from half-hourly data
    WW-YY: not produced

H_CORR75 (W m-2) Sensible heat flux, corrected H_F_MDS by energy balance closure correction factor, 75th percentile 
    DD: average from half-hourly data
    WW-YY: not produced

H_RANDUNC (W m-2) Random uncertainty of H, from measured only data 
    HH: uses only data point where H_F_MDS_QC is 0 and two hierarchical methods (see header and H_RANDUNC_METHOD)
    DD-YY: from random uncertainty of individual half-hours (rand(i)) = [SQRT(SUM(rand(i)^2)) / n], where n is the number of half-hours used

H_RANDUNC_METHOD (adimensional) Method used to estimate the random uncertainty of H 
    HH: 1 = RANDUNC Method 1 (direct SD method), 2 = RANDUNC Method 2 (median SD method)
    DD-YY: not produced

H_RANDUNC_N (adimensional) Number of half-hour data points used to estimate the random uncertainty of H
    DD-YY: not produced

H_CORR_JOINTUNC (W m-2) Joint uncertainty estimation for H 
    HH-DD: [SQRT(H_RANDUNC^2 + ((H_CORR75 - H_CORR25) / 1.349)^2)]
    WW-YY: not produced

EBC_CF_N (adimensional) Number of data points used to calculate energy closure balance correction factor. Driver data points within sliding window (ECB_CF Method 1) or number of ECB_CF data points (for ECB_CF Methods 2 and 3) 
    HH: for ECB_CF Method 1 (minimum 5, maximum 93)
    DD: for ECB_CF Method 1 (minimum 5, maximum 15)
    WW--YY: fraction between 0-1, indicating percentages of half-hours used with respect to theoretical maximum number of half hours

EBC_CF_METHOD (adimensional) Method used to calculate the energy balance closure correction factor
    HH-YY: 1 = ECB_CF Method 1, 2 = ECB_CF Method 2, 3 = ECB_CF Method 3. See general description for details