FLUXNET2015-LRQC: A Quality-Controlled FLUXNET2015 Dataset for Ecosystem Light Response Analysis

OverviewFLUXNET2015-LRQC is a quality-controlled version of the FLUXNET2015 dataset (https://fluxnet.org/data/fluxnet2015-dataset/), enhanced with seven physiologically-driven quality control (QC) flags specifically designed for ecosystem light response (LR) analysis. The dataset preserves the full structure of the original FLUXNET2015 FULLSET half-hourly data products while appending additional QC columns at the end of each record, allowing users to filter observations based on their suitability for ecosystem-scale light response parameter extraction and gross primary productivity (GPP) estimation. The suffix "_LRQC" denotes "Light-Response Quality Control."The dataset consists of 158 CSV files, each corresponding to a single FLUXNET2015 site. All original FLUXNET2015 data values remain unchanged; only the new QC columns have been appended.File Naming Convention FLX_{SITE_ID}_FLUXNET2015_FULLSET_HH_{TIME_RANGE}_{VERSION}_LRQC.csvFile names follow the official FLUXNET2015 convention: FLX — FLUXNET prefix {SITE_ID} — Six-character site identifier (e.g., AR-SLu) encoding ISO country code and site abbreviation FLUXNET2015 — Dataset release identifier FULLSET — Data product type (all variables, no record filtering) HH — Half-hourly temporal resolution {TIME_RANGE} — Years covered by the data (e.g., 2010-2011) {VERSION} — Data sub-version (e.g., 1-4) LRQC — Light-Response Quality Control designation Novel QC Columns Seven additional QC columns are appended at the end of each record: Growing-Season QC (three levels, 90% confidence): LRS-GPP_GS_QC_broad — Broad-level growing-season assessment LRS-GPP_GS_QC_moderate — Moderate-level growing-season assessment LRS-GPP_GS_QC_narrow — Narrow-level growing-season assessment Daytime Stable-Observation QC (90% confidence): LRS-GPP_DT_QC — Identifies daytime periods with stable atmospheric and radiation conditions Combined Growing-Season + Daytime QC (three levels, 90% confidence): LRS-GPP_GSDT_QC_broad — Broad-level combined assessment LRS-GPP_GSDT_QC_moderate — Moderate-level combined assessment (recommended default) LRS-GPP_GSDT_QC_narrow — Narrow-level combined assessment QC values: 1/1.0 = passes; 0/0.0 = fails; -9999.0 = not applicable. Compared to the original unfiltered FLUXNET2015 data, light response analysis precision improves most with broad-level filtering, with diminishing incremental gains at moderate and narrow levels. Users can balance data quantity against precision according to their research needs. All three levels share a 90% confidence level, differing only in interval width (precision). Usage python import pandas as pddf = pd.read_csv('FLX_AR-SLu_FLUXNET2015_FULLSET_HH_2010-2011_1-4_LRQC.csv')df_qc = df[df['LRS-GPP_GSDT_QC_moderate'] == 1] # moderate GSDT recommended first For detailed documentation, refer to the README files included in the dataset.