River Metabolism Estimation Tools (RiverMET) with Demo in the Illinois
River Basin
Abstract
Ecosystem metabolism quantifies the rate of production, maintenance, and
decay of organic matter in terrestrial and aquatic systems. It is a
fundamental measure of energy flow associated with biomass production by
photosynthesizing organisms and biomass oxidation by respiring plants,
animals, algae, and bacteria (Bernhardt et al., 2022) . Ecosystem
metabolism also provides an understanding of energy flow to higher
trophic levels that supports secondary and tertiary productivity, as
well as helping to explain when aquatic ecosystems undergo
out-of-balance behaviors such as harmful algal blooms and hypoxia.
Recent advances in sensor technology and modeling capabilities have
enabled estimation of aquatic system metabolism and gas exchange over
long time periods in rivers, streams, ponds, and wetlands where oxygen
sensors have been deployed. Here we present RiverMET, a framework for
estimation of river metabolism, with workflows to streamline data
preparation, run a stream metabolism model, assess the model
performance, and flag and censor final output data. The workflows are
specifically tailored to use streamMetabolizer, a model for one-station
calculations of stream metabolism that calculates gross primary
productivity (GPP), ecosystem respiration (ER) and the air-water gas
exchange rate constant (K600). We advise potential users of RiverMET to
review core publications for the streamMetabolizer model (Appling et
al., 2018 a, b, c) to ensure best practices that produce the most useful
results. We encourage feedback about our workflows, although issues
regarding the streamMetabolizer model itself should be referred to the
model authors. We tested RiverMET by calculating GPP, ER, and K600
across 17 river sites in the Illinois River basin (ILRB). Each river had
between one and nine years of sensor data appropriate for modeling
metabolism. In total, metabolism was modeled on 15,176 days between 2005
and 2020. Overall confidence in the results was rated as high at nine
river sites, medium at six river sites, and poor at two river sites.
Twenty-nine percent of the total modeled days had performance metrics
that triggered flags. Metrics used for daily flagging are provided with
the final output, with an option to only retain the censored daily
outputs with high confidence (representing 72 %, i.e., 10,938 days, of
the total days modeled). This work was completed as part of the U.S.
Geological Survey Proxies Project, an effort supported by the Water
Mission Area (WMA) Water Quality Processes program to develop estimation
methods for harmful algal blooms (HABs), per- and polyfluoroalkyl
substances (PFAS), and metals, at multiple spatial and temporal scales.