Using Real-Time Eddy Covariance Data for Timely Validation of Operational Remote Sensing Products: I. Setting Workflow for QC and GPP Partitioning

In order to assure the quality of operational satellite products, there is a strong demand for timely available in-situ flux data. Typically, the Earth Observation Community has to rely on publicly available data processed and distributed by flux networks such as the European Fluxes Database Cluster, AmeriFlux Network, and other major networks globally.

While the centralized processing systems employed by the major networks provide exceptional advantages for long-term data quality, reproducibility and comparability, to date these result in 1-5 year delays between the time of the actual in-situ flux measurement and the publicly online availability of processed and quality controlled data, especially for derived parameters such as Gross Primary Production (GPP) often used by Earth Observation experts. Such delays hamper the use of in-situ fluxes for timely (and ultimately near-real-time) operational satellite product monitoring, envisioned and often required by the Earth Observation Community.

Within the European Copernicus Global Land Service (CGLOPS), a validation protocol is in place for each publicly available satellite product. One of the elements is the yearly Scientific Quality Evaluation (SQE), where data of the most recent calendar year are quality checked within the three months after the end of the year. This implies that in-situ data should be available within this timespan in order to be included in the operational quality monitoring. Recently, a set of new tools to collect, process, analyze, partition, time- and space- allocate and share time-synchronized flux data from multiple flux stations were developed and deployed globally. These new tools can be effective in solving the time delay issues listed above without sacrificing quality, reproducibility and comparability of the in-situ flux data.

The fully automated remotely-accessible microcomputer, SmartFlux, utilizes EddyPro software  to calculate fully-processed fluxes in near-real-time, alongside supporting data and flux footprints. All data are merged into a single quality-controlled file timed using GPS-driven PTP time protocol to assure a microseconds-scale time synch between  the instruments within each station and between different stations.

The flux data analysis software, Tovi, can seamlessly ingest the data from the SmartFlux stations to allow a non-micrometeorologist analyze and interpret the flux data. Specifically, it allows rapid execution of the QC/QA and data analysis steps using interactive GUI, including advanced QC and gap fill schemes, footprint calculations and flux apportioning, NEE (Net Ecosystem Exchange) flux partitioning, automated generation of specific lists of references for each workflow, etc. All processing routines and analysis steps are reproducibile and intercomparable to other SmartFlux stations across the globe.

Based upon the timely needs for the in-situ flux data and the newly available technical tools, a pilot initiative was set-up to test the viability of using 2019 data generated by multiple SmartFlux stations and Tovi analysis software to quality control, gap fill, and partition NEE into GPP product to support the quality assurance analysis of the global Copernicus Dry Matter Productivity (DMP) product. This presentation will show the actual established workflow, and demonstrate the detailed post-processing of in-situ flux data for timely operational satellite product monitoring.