Known Issues

FLUXNET2015 Dataset: Known Issues

[pdf version]

Currently known issues and limitations of the FLUXNET2015 Dataset are documented here. These can include data issues, processing issues, and notes from tower teams.

Known Issues affecting November 3, 2016 update

FLUXNET Site IDs

  • Country code changes for FLUXNET site IDs:
    DK-ZaF  →  GL-ZaF
    DK-ZaH  →  GL-ZaH
    DK-NuF  →  GL-NuF
    NO-Adv  →  SJ-Adv
    NO-Blv  →  SJ-Blv

    AmeriFlux and the European Database regional networks are adopting the ISO-3611-2 standard for determining country codes for FLUXNET site IDs, affecting both new and existing sites.

    Five sites in the FLUXNET2015 dataset are affected by this new guidance, with three sites in Greenland previously identified with the Denmark country code DK now using GL, and two sites in Svalbard previously using the Norway country code NO now using the Svalbard and Jan Mayen country code SJ.

    In the regional network database, the FLUXNET site IDs were already updated, changing DK-ZaF to GL-ZaF, DK-ZaH to GL-ZaH, DK-NuF to GL-NuF, NO-Adv to SJ-Adv, and NO-Blv to SJ-Blv.

    For the identification of these sites in FLUXNET2015, the implementation of the updates is underway. We are taking into account the usage of the dataset, in both current and future works, and aim at preserving all usage and citation information. Any changes will be documented in this page and the site pages of the affected sites.

 

Data

  • [Solved with patch file, variables LE_F_MDS_QC and H_F_MDS_QC in patch file replace original versions]Latent and Sensible Heat flux quality control variables (LE/H QC). The original version of the LE_F_MDS_QC and H_F_MDS_QC variables provided in the regularly downloaded FLUXNET2015 files are correct, however they were only reported for periods when LE_CORR and H_CORR were present (LE_CORR and H_CORR are the versions LE and H with the correction for the energy balance closure). In fact there are long periods where H and LE (and the gap-filled versions LE_F_MDS and H_F_MDS) are present but the _CORR version not because one of the variables for the energy balance calculation is missing (i.e. NETRAD and/or G). In these cases the _QC variables provided in the regular files have -9999 as value, even though there is a valid value in LE_F_MDS and H_F_MDS. The version of the _QC variables provided here are complete and provide QC flags for all records with a valid value of LE_F_MDS or H_F_MDS. Download patch file with complete variables here [ftp://ftp.fluxdata.org/.fluxnet_downloads_67934/FLUXNET2015/FLUXNET2015_PATCH1.zip]

 

Known Issues affecting July 12, 2016 update

Data

  • [Solved in the November 3, 2016 update] [DK-NuF] there was a problem creating the downscaled variables TA_ERA, SW_IN_ERA, VPD_ERA. These have an incorrect range of values, affecting their corresponding gapfilled variables (_F) and fluxes products (GPP, RECO, LE_CORR, H_CORR, etc.)
  • [Solved in the November 3, 2016 update] [CA-TP1] there is a number of non-zero values for nighttime PPFD_IN, potentially affecting downscaling (_ERA) and gap-filling (_F) of PPFD_IN and fluxes products (GPP, RECO, LE_CORR, H_CORR, etc.)

 

Known Issues affecting December 31, 2015 release

Data

  • [Solved in the July 12, 2016 update] [CA-Qfo, US-MMS, RU-Vrk] ERA downscaling processing step failed, affecting all downscaled (_ERA) and gapfilled (_F) variables, as well as all fluxes products (GPP, RECO, LE_CORR, H_CORR, etc.)
  • [Solved in the July 12, 2016 update] [AU-ASM, AU-Ade, AU-Cpr, AU-Cum, AU-DaP, AU-DaS, AU-Rig, AU-Tum, AU-Whr, CN-Cha, CN-Cng, CN-Dan, CN-Din, CN-Ha2, CN-HaM, CN-Qia, CN-Sw2, IT-Lav, MY-PSO] Soil Water Content (SWC) variables have the wrong units (0-1 instead of 0-100 range)
  • [Solved in the July 12, 2016 update] [ALL-SITES] Aggregation of meteo _QC variables in temporal resolutions WW, MM, YY are incorrect. Recomputing from HH resolution is necessary before using.
  • [Solved in the July 12, 2016 update] [ALL-SITES] Aggregation of soil temperature (TS) and water content (SWC) variables in temporal resolutions WW, MM, YY are incorrect. Recomputing from HH resolution is necessary before using.
  • [Solved in the July 12, 2016 update] [ALL-SITES] Aggregation of _QC variables for NEE in temporal resolutions WW, MM, YY are incorrect. Recomputing from HH resolution is necessary before using.

Auxiliary Data

  • [Solved in the July 12, 2016 update] [ALL-SITES] AUXMETEO files: Variable LW_IN_JSB mislabeled as LW_IN_LSB in AUXMETEO files
  • [Solved in the July 12, 2016 update] [ALL-SITES] AUXNEE files Variables NEE_CUT_USTAR50, NEE_VUT_USTAR50 mislabeled as NEE_USTAR50_CUT, NEE_USTAR50_VUT, NEE_CUT_REF, NEE_VUT_REF, respectively in AUXNEE files