Lessons learned from FLUXNET

FLUXNET data are important for a variety of research communities.  Groups of studies have revealed:

1. How annual photosynthesis and respiration co-vary with variations in climate, plant functional type, drought, and heat;
2. How length of the growing season and phenology modulates annual photosynthesis;
3. How peak photosynthesis acclimates with temperature;
4. How light use efficiency increases with the fraction of diffuse light;
5. How photosynthetic capacity adjusts with time of season;
6. How rain induces pulses in ecosystem respiration;
7. How net carbon exchange varies as a function of time since disturbance;
8. How biophysical variables, like albedo, vary with plant functional type and nutrition.
9. What the relative controls among physical and biological variables are on net and gross carbon fluxes.
10. How interannual variability in weather and climate modulate carbon and water fluxes
11. How trends in carbon dioxide are affecting ecosystem scale water use efficiency
12. How much water is evaporated from ecosystems, as opposed to that inferred as the residual of water budgets.

Figure 1. The distribution of FLUXNET publications by field

What do we contribute to other communities?

In its current and future role, FLUXNET is proving to be a critical partner in efforts to produce information on carbon and water fluxes that is ‘everywhere, all of the time’.  This work is being done by collaborating with the remote sensing and earth system modeling communities.  The FLUXNET community serves the broader biogeoscience community by producing direct flux measurements at integrated time steps (days, months, years) that are of value to modelers for parameterization and validation.  In turn, the remote sensing community is able to infer gross and net carbon dioxide and water vapor fluxes at high spatial resolution by using satellite-based sensors that detect reflected light from Earth.  But, these algorithms must use and rely on carbon dioxide flux measurements to parameterize and validate themselves.  Biogeochemical and ecological models that diagnose and forecast the state of the land’s carbon and water budgets depend upon a network of ‘super-sites’ measuring a suite of site characteristics to identify or quantify important biophysical processes and parameterize mechanistic algorithms.  Another group of models need a dense network of less-intensive flux measurement sites that are sampling representative climate and ecological spaces.

These scientists digest flux data, remote sensing and climate data and produce maps of net and gross carbon and water fluxes at regional, continental and global scales using machine learning techniques (e.g., neural networks, regression trees or genetic algorithms).  Moreover, maps produced by machine learning technique have the potential to be used to test and improve upon land-atmosphere flux algorithms that are used in climate models, if these data products are integrated to the coarse scale (e.g., 0.5 to 1 degree) of such models.  Or, gridded flux products can be used as ‘priors’ in inverting carbon fluxes based on measurements of carbon dioxide concentration fields from the global carbon dioxide flask network or orbiting carbon dioxide satellites.  And in the future, spatially integrated flux data will aid the validation of new efforts to measure global photosynthesis from space with solar-induced fluorescence (SIF).

FLUXNET data are also being used by a diverse and wider set of scientists, outside the FLUXNET community the suite of mass and energy flux, meterological, soil and plant data to:

1. Test fields of incoming and outgoing radiation measurements derived from satellites;
2. Upscale spatial/temporal fields of net and gross photosynthesis with remote sensing
3. Define and evaluate Evaporation and Water Budgets, at annual time scales
4. Quantify and refine theories on Phenology
5. Provide data on biophysical properties of the land surface
6. Provide data and information on soil moisture and groundwater coupling
7. Quantify water use efficiency, and evaluate trends in time
8. Identify and discover emergent scale properties of ecosystems
9. Provide data on atmospheric chemistry and VOC fluxes
10. Advance the field of eddy Covariance flux measurement theory and application and refine methods and quantify uncertainty
11. Provide data for Land Surface Modeling and Land Atmosphere interactions
12. Provide data on Canopy Structure and Function
13. Provide data and new insights on Ecosystem Ecology and EcoPhysiology
14. Serve as a model for Big Data and Data Systems

For more lessons learned, please see archived versions of FluxLetters, a quarterly newsletter published from 2008 – 2013.