A Multiscale Approach to Timescale Analysis: Isolating Diel Signals from Solute Concentration Time Series

Area of Science:

• Environmental Science
• Hydrology
• Ecology

Background:

• Solute concentration time series reveal hydrological and biological influences across various frequencies.
• Understanding dynamic ecosystem conditions and water quality requires disentangling these complex signals.
• Inferring biogeochemical processes from diel (daily) solute variations is challenging due to background noise.

Purpose of the Study:

• To develop and test a new time-series filtering method for isolating subtle diel signals in river solute concentrations.
• To overcome limitations of conventional analyses that assume stationary background variability.
• To enable better inference of biogeochemical processes and ecosystem dynamics.

Main Methods:

• Developed a time-series filtering method using empirical mode decomposition (EMD) to break down concentration data into intrinsic mode functions.
• Filtered decomposed modes based on periodicity, phase, and coherence with neighboring data, incorporating mechanistic knowledge.
• Validated the method on synthetic datasets and a year of high-frequency (15-min) river solute concentration data from three distinct US rivers.

Main Results:

• The method successfully isolated diel signals in real-world data that correlated with gross primary productivity.
• While the isolated diel signal strength was lower than in synthetic data, derived process-model estimates were comparable to other methods.
• The decomposition method preserves information valuable for further process modeling.

Conclusions:

• The developed filtering technique effectively isolates diel variations in river solute concentrations, even amidst complex background variability.
• This approach offers a valuable tool for understanding diel ecosystem processes and transient water quality issues.
• The method provides an alternative to Fourier and wavelet analyses, with different underlying data assumptions.