Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Reddy: An open-source toolbox for analyzing eddy-covariance measurements in heterogeneous environments

Laura Mack, Norbert Pirk

TL;DR

The paper presents Reddy, an open-source, modular toolbox for eddy-covariance data processing and turbulence analysis in heterogeneous environments, addressing challenges posed by advective fluxes and non-stationarity. It details data processing, a suite of turbulence diagnostics, and advanced analysis/visualization tools (spectra, MRD, ogives, quadrant analysis, anisotropy, flux footprints) with interfaces for model outputs and context. Three Norwegian field applications (Finse, Langtjern, Iškoras) demonstrate how Reddy facilitates site-specific post-processing, averaging-time selection, and flux-variance analyses, complemented by Jupyter notebooks and Python port support. The work emphasizes reproducibility, teaching utility, and extensibility, positioning Reddy as a comprehensive extension to existing software for holistic turbulence analysis in real-world heterogeneous settings.

Abstract

Land-atmosphere exchange processes are determined by turbulent fluxes, which can be derived from eddy-covariance measurements. This method was established to quantify ecosystem-scale vertical atmosphere-vegetation exchange processes, but is also used to validate atmospheric turbulence theories with the ultimate aim to improve the representation of turbulence in numerical models. While the focus has long been on turbulence over idealized, homogeneous and flat surfaces, recent scientific developments are shifting towards investigating turbulent exchange processes in complex heterogeneous environments under non-idealized conditions, which pose particular challenges, e.g. advective fluxes between different surface types or non-stationarity of nighttime turbulence. This requires to rethink standard post-processing routines for determining turbulent fluxes from the high-frequency sonic and gas analyzer measurements. Here, we introduce the open-source R-package 'Reddy', which provides modular-built functions for post-processing, analysis and visualization of eddy-covariance measurements, including investigating spectra, coherent structures, anisotropy, flux footprints and surface energy balance closure. The 'Reddy' package is accompanied by a detailed documentation and a set of jupyter notebooks introducing new users hands-on to eddy-covariance data analysis. We showcase 'Reddy' based on measurements from three different sites in Norway: A case study during strong stratification over alpine tundra, for determining suitable averaging times during ice-cover transition at a boreal lake, and for fitting flux-variance relations for a permafrost peatland. 'Reddy' serves as extension of previously developed software packages, paving the way towards holistic turbulence data analysis in heterogeneous real-world environments.

Reddy: An open-source toolbox for analyzing eddy-covariance measurements in heterogeneous environments

TL;DR

The paper presents Reddy, an open-source, modular toolbox for eddy-covariance data processing and turbulence analysis in heterogeneous environments, addressing challenges posed by advective fluxes and non-stationarity. It details data processing, a suite of turbulence diagnostics, and advanced analysis/visualization tools (spectra, MRD, ogives, quadrant analysis, anisotropy, flux footprints) with interfaces for model outputs and context. Three Norwegian field applications (Finse, Langtjern, Iškoras) demonstrate how Reddy facilitates site-specific post-processing, averaging-time selection, and flux-variance analyses, complemented by Jupyter notebooks and Python port support. The work emphasizes reproducibility, teaching utility, and extensibility, positioning Reddy as a comprehensive extension to existing software for holistic turbulence analysis in real-world heterogeneous settings.

Abstract

Land-atmosphere exchange processes are determined by turbulent fluxes, which can be derived from eddy-covariance measurements. This method was established to quantify ecosystem-scale vertical atmosphere-vegetation exchange processes, but is also used to validate atmospheric turbulence theories with the ultimate aim to improve the representation of turbulence in numerical models. While the focus has long been on turbulence over idealized, homogeneous and flat surfaces, recent scientific developments are shifting towards investigating turbulent exchange processes in complex heterogeneous environments under non-idealized conditions, which pose particular challenges, e.g. advective fluxes between different surface types or non-stationarity of nighttime turbulence. This requires to rethink standard post-processing routines for determining turbulent fluxes from the high-frequency sonic and gas analyzer measurements. Here, we introduce the open-source R-package 'Reddy', which provides modular-built functions for post-processing, analysis and visualization of eddy-covariance measurements, including investigating spectra, coherent structures, anisotropy, flux footprints and surface energy balance closure. The 'Reddy' package is accompanied by a detailed documentation and a set of jupyter notebooks introducing new users hands-on to eddy-covariance data analysis. We showcase 'Reddy' based on measurements from three different sites in Norway: A case study during strong stratification over alpine tundra, for determining suitable averaging times during ice-cover transition at a boreal lake, and for fitting flux-variance relations for a permafrost peatland. 'Reddy' serves as extension of previously developed software packages, paving the way towards holistic turbulence data analysis in heterogeneous real-world environments.
Paper Structure (13 sections, 5 figures, 1 table)

This paper contains 13 sections, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Package Description
  3. Data processing
  4. Calculation of turbulence diagnostics
  5. Data analysis and visualization
  6. Manual and Documentation
  7. Applications
  8. Example 1: Finse
  9. Example 2: Langtjern
  10. Example 3: Iškoras
  11. Further Applications
  12. Discussion
  13. Conclusion

Figures (5)

  • Figure 2.1: Components and workflow from eddy-covariance raw data to application in 'Reddy'.
  • Figure 3.1: Overview of the studied field sites and their flux footprint climatologies: (a) Site locations, (b) Iškoras (used time period: 2019-2020), (c) Finse (used time period: 02/2018-01/2019), (d) Langtjern (during the measurement campaign). Different scales.
  • Figure 3.2: Case study from Finse, 17.03.2018-18.03.2018: (a) Temperature and wind speed, (b) Richardson number and decoupling metric (based on vertical gradients between atmosphere and surface) and anisotropy, (c) turbulence intensity and thermal heterogeneity (30 minutes and 1 minute averaging), (d) surface energy balance. Shaded areas mark bad quality flags (QC = 2).
  • Figure 3.3: Composite MRDs and ogives during ice-covered and ice-free period at lake Langtjern. The shaded areas mark the inter-quartile-range and the solid vertical line the suggested averaging time.
  • Figure 3.4: Flux-variance relations for water vapor and carbon dioxide fitted based on two years (2020-2021) of eddy-covariance measurements at Iškoras.