Random errors in carbon and water vapor fluxes assessed with Gaussian Processes

Menzer, Olaf; Moffat, Antje Maria; Meiring, Wendy; Lasslop, Gitta; Schukat-Talamazzini, Ernst Günter; Reichstein, Markus

doi:10.1016/j.agrformet.2013.04.024

Cited by 20 publications

(14 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This was only slightly higher than the range reported in the literature for forest sites (Hollinger and Richardson, 2005;Liu et al, 2009;Menzer et al, 2013), which increases our confidence in eddy covariance CO 2 sums from urban sites. We also found the half-hourly uncertainties inferred from our models to increase with the flux magnitude (i.e., error variances were heteroscedastic) and approximately Laplacian distributed, which are typical statistical properties of eddy covariance measurement error (Hollinger and Richardson, 2005;Richardson et al, 2006).…”

Section: Annual Urban Co 2 Flux Sums and Uncertaintiessupporting

confidence: 45%

“…They directly relate to ANNs as they can be considered an ANN (or RBF) with infinitely many hidden units (MacKay, 1998). GPs are widely implemented as Kriging in geostatistics, following Matheron (1963), and have more recently been applied for random uncertainty estimation in EC flux time series (Menzer et al, 2013). A GP is a stochastic process that can be defined by a mean function and a covariance function.…”

Section: Gap-filling Techniquesmentioning

confidence: 99%

See 1 more Smart Citation

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

Menzer¹,

Meiring²,

Kyriakidis³

et al. 2015

Atmospheric Environment

Self Cite

View full text Add to dashboard Cite

Section: Annual Urban Co 2 Flux Sums and Uncertaintiessupporting

confidence: 45%

Section: Gap-filling Techniquesmentioning

confidence: 99%

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

Menzer¹,

Meiring²,

Kyriakidis³

et al. 2015

Atmospheric Environment

Self Cite

View full text Add to dashboard Cite

“…In relative terms, the accumulated uncertainty varied between ±0.38 and ±2.39% of the seasonal CO 2 efflux across sites. These values of uncertainty could vary depending of the length of the data [ Liu et al ., ; Loescher et al ., ], the frequency of the measurements, the magnitude of fluxes [ Richardson et al ., ], and the method to calculate random errors [ Menzer et al ., ; Richardson and Hollinger , ]. For example, the Stu site had not only the highest uncertainty due to random errors but also the highest magnitude of F (Figure ), the highest total annual sum (Table ), and the highest daily variability (Table ) that is translated into a large standard deviation in the annual sum (Table ).…”

Section: Discussionmentioning

confidence: 99%

“…There are different methods to calculate random errors in flux measurements, mainly explained in the eddy covariance literature. The “model residuals” approach [ Lasslop et al ., ; Menzer et al ., ; Richardson et al ., ] infers the errors from the difference between model predictions and measurements from instruments [ Richardson et al ., ; Vargas et al ., ]. The “paired instruments” [ Schmidt et al ., ] or “two towers” [ Hollinger et al ., ] approach uses the difference of two similar instruments/towers that measure at the “same” time and place to estimate the random errors [see Kessomkiat et al ., ].…”

Section: Methodsmentioning

confidence: 98%

A multisite analysis of temporal random errors in soil CO₂ efflux

et al. 2015

View full text Add to dashboard Cite

An important component of the terrestrial carbon balance is the efflux of CO 2 from soils to the atmosphere, which is strongly influenced by changes in soil moisture and temperature. Continuous measurements of soil CO 2 efflux are available around the world, and there is a need to develop and improve analyses to better quantify the precision of the measurements. We focused on random errors in measurements, which are caused by unknown and unpredictable changes such as fluctuating environmental conditions. We used the CO 2 gradient flux method with two different algorithms to study the temporal variation of soil CO 2 efflux and associated random errors at four different ecosystems with wide ranges in mean annual temperature, soil moisture, and soil CO 2 efflux. Our results show that random errors were better explained by a double-exponential distribution, had a mean value close to zero, were nonheteroscedastic, and were independent of soil moisture conditions. Random errors increased with the magnitude of soil CO 2 efflux and scale isometrically (scaling exponent ≈ 1) within and across all sites, with a single relation common to all data. This isometric scaling is unaffected by ecosystem type, soil moisture conditions, and soil CO 2 efflux range (maximum and minimum values within an ecosystem). These results suggest larger uncertainty under extreme events that increase soil CO 2 efflux rates. The accumulated annual uncertainty due to random errors varied between ±0.38 and ±2.39%. These results provide insights on the scalability of the sensitivity of soil CO 2 efflux to changing weather conditions across ecosystems.

show abstract

“…C5 is a low estimate, because it assumed no correlation between half-hourly errors. An improved quantification of correlations requires the full variance-covariance matrix of the LRC parameter fits (Lasslop et al, 2010;Menzer et al, 2013), which were not available for 15 the BGC online tool.…”

mentioning

confidence: 99%

Basic and extensible post-processing of eddy covariance flux data with REddyProc

Wutzler¹,

Moffat²,

Migliavacca³

et al. 2018

Preprint

Self Cite

View full text Add to dashboard Cite

Abstract.With the eddy-covariance (EC) technique, net fluxes of carbon dioxide (CO 2 ) and other greenhouse gases as well as water and energy fluxes can be measured at the ecosystem level. These flux measurements are a main source for understanding biosphere-atmosphere interactions and feedbacks by cross-site analysis, model-data integration, and up-scaling. The raw fluxes measured with the EC technique require an extensive and laborious data processing. While there are standard tools available 5 in open source environment for processing high-frequency (10 or 20 Hz) data into half-hourly quality checked fluxes, there is a need for more usable and extensible tools for the subsequent post-processing steps. We tackled this need by developing the REddyProc package in the cross-platform language R that provides standard CO 2 -focused post-processing routines for reading (half-)hourly data from different formats, estimating the uStar threshold, gap-filling, flux-partitioning, and visualizing the results. In addition to basic processing, the functions are extensible and allow easier integration in extended analysis than current 10 tools. New features include cross year processing and a better treatment of uncertainties. A comparison of REddyProc routines with other state-of the art tools resulted in no significant differences in monthly and annual fluxes across sites. Lower uncertainty estimates of both uStar and resulting gap-filled fluxes with the presented tool was achieved by an improved treatment of seasons during the bootstrap analysis. Higher estimates of uncertainty in day-time partitioning resulted from a better accounting of the uncertainty in estimates of temperature sensitivity of respiration. The provided routines can be easily installed, configured, 15 used, and integrated with further analysis. Hence the eddy covariance community will benefit from using the provided package, allowing easier integration of standard post-processing with extended analysis.

show abstract

Random errors in carbon and water vapor fluxes assessed with Gaussian Processes

Cited by 20 publications

References 39 publications

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

A multisite analysis of temporal random errors in soil CO₂ efflux

Basic and extensible post-processing of eddy covariance flux data with REddyProc

Contact Info

Product

Resources

About

Random errors in carbon and water vapor fluxes assessed with Gaussian Processes

Cited by 20 publications

References 39 publications

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

Annual sums of carbon dioxide exchange over a heterogeneous urban landscape through machine learning based gap-filling

A multisite analysis of temporal random errors in soil CO2 efflux

Basic and extensible post-processing of eddy covariance flux data with REddyProc

Contact Info

Product

Resources

About

A multisite analysis of temporal random errors in soil CO₂ efflux