QCF: overall quality control flag

Contents

QCF: overall quality control flag#

Note

The QCF (Quality Control Flag) is calculated from multiple flags from single tests.

Description#

The overall quality of each 30-minute flux data record was assessed by combining quality test results from the individual Level-2, Level-3.2 and Level-3.3 quality tests into a single overall quality control flag (QCF), calculated separately for each flux. Fluxes were categorized into three quality levels: highest quality (QCF=0) fluxes successfully passed all individual quality checks; medium quality (QCF=1) fluxes had no individual tests that outright rejected the data (test flag=2) and a maximum of two tests indicating moderate quality (test flag=1); low quality (QCF=2) fluxes were characterized by at least one individual test rejecting the record or by three or more tests marking the record as moderate quality. Low-quality flux records (QCF=2) were removed from the dataset in all cases. Moderate quality fluxes (QCF=1) were retained, with the exception of nighttime NEE where moderate quality NEE fluxes were also rejected. Resulting quality-filtered time series were used in subsequent processing steps.

Details#

  • QCF is a flag that shows the overall quality of the respective data point.

  • QCF can be: 0=best data, 1=OK data, 2=bad data

  • QCF is calculated from single test flags summed together:

    • For records where the sum > 2: QCF=2

    • For records where the sum = 2 from a single flag with value 2: QCF=2

    • For records where the sum = 2 and no single flag is 2 (i.e., the sum comes from two single flags with value 1): QCF=1

    • For records where the sum = 1 (i.e., only on single flag with value 1): QCF=1

    • For records where the sum = 0 (i.e., all single flags are zero): QCF=0 (best quality fluxes)

  • Flags for single tests are created in Level-2, Level-3.2 and Level-3.3.

  • Genereally, single tests can be:

    • Hard flags: Some tests yield either 0 or 2, but have no flag=1. This is the case if the test is so crucial that if the test fails, data are considered bad. If one test flag=2, then the QCF is automatically 2, no matter the other tests. Flags from such tests are also called hard flags.

    • Soft flags: If a test results is flag=1, it can still be OK data for some analyses. Even if there is a second test with flag=1, data might still be OK. If one or two flags are 1, and there is no flag=2, then the QCF is also 1. These flags are less crucial and are therefore soft flags.

  • QCF always uses the same logic, for both flux and meteo data, only the single tests are different.

  • This was done for all fluxes, however, for NEE the requirements were stricter during the nighttime than during the daytime. For NEE, daytime QCF flags of 0 and 1 were accepted (flag 2 = bad data), but during nighttime only QCF flags with 0 were retained (flags 1 and 2 = bad data). For all other fluxes, QCF flags of 0 and 1 were accepted during daytime and nighttime (flag 2 = bad data).

  • Example:

    • FLAG_L3.3_CUT_50_NEE_L3.1_QCF is the flag after Level-3.3, for flux NEE_L3.1 (storage-corrected) for the USTAR scenario CUT_50. This flag is applied to flux NEE_L3.1, producing the quality-filtered flux NEE_L3.1_L3.3_CUT_50_QCF. In addition, another flux variable is produced, containing only highest-quality fluxes: NEE_L3.1_L3.3_CUT_50_QCF0 (all single flags are zero).

Fig. 4 Example showing how the overall quality control flag QCF is calculated from single test flags.#

Contents