Data output format
The default corresponds to the previous pvWave BGC online tool and few alternative options are provided with the web-service.
Default: Online Tool
Description of gap filling output variables
| Suffix | Description | Old suffix |
| _Thres | the threshold of uStar values used to mark insufficient conditions | |
| _orig | Original values used for gap filling | orig |
| _f | Original values and gaps filled with mean of selected datapoints (condition depending on gap filling method) | _f |
| _fqc | Quality flag assigned depending on gap filling method and window length: 0 = original, 1 = most reliable, 2 = medium, 3 = least reliable | _fqc |
| _fall | All values considered as gaps (for uncertainty estimates) | _f_unc |
| _fall_qc | All values considered as gaps (for uncertainty estimates) | _fqc_unc |
| _fnum | Number of datapoints used for gap filling | _fn, _fn_unc |
| _fsd | Standard deviation of datapoints used for gap filling (uncertainty) | _fs, _fs_unc |
| _fsdu | Standard deviation across uStar thresholds (uncertainty, bias) | |
| _fsdug | Combination of random uncertainty and uncertainty due to u*: sqrt(fsd^2 + fsdu^2) | |
| _fmeth | Method used for gap filling: 1 = similar meteo condition (sFillLUT with Rg, VPD, Tair) 2 = similar meteo (sFillLUT with Rg only) 3 = mean diurnal course (sFillMDC)) | _fmet, _fmet_unc |
| _fwin | Full window length used for gap filling | _fwin, _fwin_unc |
Bootstrapping the uStar-Threshold estimate
When bootstrapping was performed, several output columns are re-estimated with gap-Filling based on low, median, and high estimates of the uStar-Threshold.
| Suffix | Description | Old suffix |
| _uStar | estimate on the original unbootstrapped data | (empty) |
| _U05 | low estimate (5% quantile of the bootstrapped uncertainty distribution) | |
| _U50 | median estimate (50% quantile of the bootstrapped uncertainty distribution) | |
| _U95 | high estimate (95% quantile of the bootstrapped uncertainty distribution) |
A missing '_uStar' suffix in NEE, GPP, or Reco, it corresponds either to not performing uStar filtering or to uStar threshold of the non-bootstrapped data.
Description of flux partitioning output variables
| Variable name | Description | Old name |
| GPP_f | Gross primary production, i.e. influx to the land surface (μmol m-2 s-1 as NEE) (nighttime-based) | |
| Reco_f | Ecosystem respiration, i.e. outflux from the land surface (μmol m-2 s-1 as NEE) (nighttime-based) | |
| GPP_DT | Gross primary production, i.e. influx to the land surface (μmol m-2 s-1 as NEE) estimated by day-time partitioning | |
| Reco_DT | Ecosystem respiration, i.e. outflux from the land surface (μmol m-2 s-1 as NEE) estimated by day-time partitioning | |
| E0 | activation energy parameter (K) in relationship between temperature and nighttime NEE from night-time partitioning | |
| R_ref | respiration at reference temperature parameter (μmol m-2 s-1 as NEE) in relationship between temperature and nighttime NEE from night-time partitioning | |
| FP_alpha | canopy light utilization efficiency and represents the initial slope of the light–response curve (daytime partitioning) | |
| FP_beta | maximum CO2 uptake rate of the canopy at light saturation (μmol m-2 s-1) (daytime partitioning) | |
| FP_k | parameter controlling the VPD limitation of GPP (daytime partitioning) (daytime partitioning) | |
| FP_qc | quality flag of the estimated parameters: 0: good parameter fit, 1: some parameters out of range, required refit, 2: next parameter estimate is more than two weeks away (daytime partitioning) | |
| FP_dRecPar | records until or after closest record that has a parameter estimate associated (daytime partitioning) | |
| FP_GPP2000 | GPP at incoming radiation of 2000 Wm-2, more robust alternative to saturation FP_k | |
| FP_E0 | activation energy parameter (K) in relationship between temperature and nighttime NEE from day-time partitioning | |
| FP_RRef | respiration at reference temperature parameter (μmol m-2 s-1 as NEE) in relationship between temperature and nighttime NEE from day-time partitioning | |
| FP_RRef_Night | same as FP_RRef using the same FP_E0, but from intermedate step based on night-time data | |
| FP_<X>_sd | estimated standard devation of X | |
| FP_qc | quality flag of day-time partitioning: 0: good parameter fit, 1: some parameters LRC parameters were out of range, required refit, 2: next parameter estimate is more than two weeks away |
REddyProc 1.2
In difference to the Online-Tool-default,
- the _uStar suffix is handled differently:
- A missing '_uStar' suffix it corresponds to not applying filtering for u* threshold.
- For filtering with the uStar threshold estimated on non-bootstrapped data, the suffix '_uStar' is used.
Fluxnet 2015
Provides data output similar to the formatting of the 2015 fluxnet data release.
The following differences to default REddyProc 1.2. output:
- Formatted as comma-separated ASCII file.
- All rows after the first row contain data, i.e. no units in the second row
- Time (end of the measurement interval) is specified in column TIMESTAMP_END in ISO timestamp format: YYYMMDDHHMM
- During export column names are translated.
- Fluxes with suffix of used uStar threshold quantile
- "^NEE_U(\\d\\d)_f$" -> "NEE_VUT_\\1", (the pattern corresponds to the two digits quantile of the uStar threshold distribution)
- "^GPP_U(\\d\\d)_f$" -> "GPP_NT_VUT_\\1"
- "^GPP_DT_U(\\d\\d)$" -> "GPP_DT_VUT_\\1"
- "^Reco_U(\\d\\d)$" -> "RECO_NT_VUT_\\1"
- "^Reco_DT_U(\\d\\d)$" -> "RECO_DT_VUT_\\1", "GPP"
- "^Ustar_Thresh_U(\\d\\d)$" -> "USTAR_THRESHOLD_VUT_\\1"
- "^NEE_U50_fqc$" -> "NEE_VUT_USTAR50_QC"
- "^NEE_U50_fsd$" -> "NEE_VUT_USTAR50_RANDUNC"
- "^NEE_U50_fnum$" -> "NEE_VUT_USTAR50_RANDUNC_N
- further variables
- "night" -> "NIGHT", "Rg_f" -> "SW_IN_F_MDS", "Rg_fqc" -> "SW_IN_F_MDS_QC", "PotRad_NEW" -> "SW_IN_POT", "Tair_f" -> "TA_F_MDS", "Tair_fqc" -> "TA_F_MDS_QC", "VPD_f" -> "VPD_F_MDS", "VPD_fqc" -> "VPD_F_MDS_QC"
- Fluxes with suffix of used uStar threshold quantile