Sensitivity analysis of a source partitioning method for H2O and CO2 fluxes based on high frequency eddy covariance data

Findings from field data and large eddy simulations

A. Klosterhalfen*, A.F. Moene, M. Schmidt, T.M. Scanlon, H. Vereecken, A. Graf

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)

Abstract

Scanlon and Sahu (2008) and Scanlon and Kustas (2010) proposed a source partitioning method (SK10 in the following) to estimate contributions of transpiration, evaporation, photosynthesis, and respiration to H2O and CO2 fluxes obtained by the eddy covariance method. High frequency eddy covariance raw data time series are needed, and the source partitioning is estimated based on separate application of the flux-variance similarity theory to stomatal and non-stomatal components of the regarded fluxes, as well as on additional assumptions on leaf-level water use efficiency (WUE). We applied SK10 to data from two test sites (forest and cropland) and analyzed partitioning results depending on various ways to estimate WUE from available data. Also, we conducted large eddy simulations (LES), simulating the turbulent transport of H2O and CO2 for contrasting vertical distributions of the canopy sinks/sources, as well as for varying relative magnitudes of soil sources and canopy sinks/sources. SK10 was applied to the synthetic high frequency data generated by LES and the effects of canopy type, measurement height, given sink-source-distributions, and input of varying WUEs were tested regarding the partitioning performance. SK10 requires that the correlation coefficient between stomatal and non-stomatal scalar fluctuations is determined by the ratio of the transfer efficiencies of these scalar components, an assumption (transfer assumption in the following) that could be tested with the generated LES data. The partitioning results of the field sites yielded satisfactory flux fractions, when fair-weather conditions (no precipitation) and a high productive state of the vegetation were present. Further, partitioning performance with regard to soil fluxes increased with crop maturity. Results also showed relatively large dependencies on WUE, where the partitioning factors (median) changed by around -57% and +36%. Measurements of outgoing longwave radiation used for the estimation of foliage temperature and WUE could slightly increase the plausibility of the partitioning results in comparison to soil respiration measurements by decreasing the partitioning factor by up to 42%. The LES-based analysis revealed that for a satisfying performance of SK10, a certain degree of decorrelation of the H2O and CO2 fluctuations (here, |ρq'c’| < 0.975) was needed. This decorrelation is enhanced by a clear separation between soil sources and canopy sinks/sources, and for observations within the roughness sublayer. The expected dependence of the partitioning results on the WUE input could be observed. However, due to violation of the abovementioned transfer assumption, the known true input WUE did not yield the known true input partitioning. This could only be achieved after introducing correction factors for the transfer assumption, which were known however only in the special case of the LES experiments.

Original languageEnglish
Pages (from-to)152-170
JournalAgricultural and Forest Meteorology
Volume265
DOIs
Publication statusPublished - 15 Feb 2019

Fingerprint

eddy covariance
large eddy simulation
water use efficiency
sensitivity analysis
partitioning
canopy
methodology
soil
roughness
soil respiration
method
evaporation
leaves
time series analysis
transpiration
weather
photosynthesis
longwave radiation
vegetation
foliage

Keywords

  • Flux partitioning
  • Latent heat flux
  • LES
  • Net ecosystem exchange
  • Sensitivity analysis
  • Water use efficiency

Cite this

@article{02bbd75b18fd49b2b5fbd69d8f0393a1,
title = "Sensitivity analysis of a source partitioning method for H2O and CO2 fluxes based on high frequency eddy covariance data: Findings from field data and large eddy simulations",
abstract = "Scanlon and Sahu (2008) and Scanlon and Kustas (2010) proposed a source partitioning method (SK10 in the following) to estimate contributions of transpiration, evaporation, photosynthesis, and respiration to H2O and CO2 fluxes obtained by the eddy covariance method. High frequency eddy covariance raw data time series are needed, and the source partitioning is estimated based on separate application of the flux-variance similarity theory to stomatal and non-stomatal components of the regarded fluxes, as well as on additional assumptions on leaf-level water use efficiency (WUE). We applied SK10 to data from two test sites (forest and cropland) and analyzed partitioning results depending on various ways to estimate WUE from available data. Also, we conducted large eddy simulations (LES), simulating the turbulent transport of H2O and CO2 for contrasting vertical distributions of the canopy sinks/sources, as well as for varying relative magnitudes of soil sources and canopy sinks/sources. SK10 was applied to the synthetic high frequency data generated by LES and the effects of canopy type, measurement height, given sink-source-distributions, and input of varying WUEs were tested regarding the partitioning performance. SK10 requires that the correlation coefficient between stomatal and non-stomatal scalar fluctuations is determined by the ratio of the transfer efficiencies of these scalar components, an assumption (transfer assumption in the following) that could be tested with the generated LES data. The partitioning results of the field sites yielded satisfactory flux fractions, when fair-weather conditions (no precipitation) and a high productive state of the vegetation were present. Further, partitioning performance with regard to soil fluxes increased with crop maturity. Results also showed relatively large dependencies on WUE, where the partitioning factors (median) changed by around -57{\%} and +36{\%}. Measurements of outgoing longwave radiation used for the estimation of foliage temperature and WUE could slightly increase the plausibility of the partitioning results in comparison to soil respiration measurements by decreasing the partitioning factor by up to 42{\%}. The LES-based analysis revealed that for a satisfying performance of SK10, a certain degree of decorrelation of the H2O and CO2 fluctuations (here, |ρq'c’| < 0.975) was needed. This decorrelation is enhanced by a clear separation between soil sources and canopy sinks/sources, and for observations within the roughness sublayer. The expected dependence of the partitioning results on the WUE input could be observed. However, due to violation of the abovementioned transfer assumption, the known true input WUE did not yield the known true input partitioning. This could only be achieved after introducing correction factors for the transfer assumption, which were known however only in the special case of the LES experiments.",
keywords = "Flux partitioning, Latent heat flux, LES, Net ecosystem exchange, Sensitivity analysis, Water use efficiency",
author = "A. Klosterhalfen and A.F. Moene and M. Schmidt and T.M. Scanlon and H. Vereecken and A. Graf",
year = "2019",
month = "2",
day = "15",
doi = "10.1016/j.agrformet.2018.11.003",
language = "English",
volume = "265",
pages = "152--170",
journal = "Agricultural and Forest Meteorology",
issn = "0168-1923",
publisher = "Elsevier",

}

Sensitivity analysis of a source partitioning method for H2O and CO2 fluxes based on high frequency eddy covariance data : Findings from field data and large eddy simulations. / Klosterhalfen, A.; Moene, A.F.; Schmidt, M.; Scanlon, T.M.; Vereecken, H.; Graf, A.

In: Agricultural and Forest Meteorology, Vol. 265, 15.02.2019, p. 152-170.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Sensitivity analysis of a source partitioning method for H2O and CO2 fluxes based on high frequency eddy covariance data

T2 - Findings from field data and large eddy simulations

AU - Klosterhalfen, A.

AU - Moene, A.F.

AU - Schmidt, M.

AU - Scanlon, T.M.

AU - Vereecken, H.

AU - Graf, A.

PY - 2019/2/15

Y1 - 2019/2/15

N2 - Scanlon and Sahu (2008) and Scanlon and Kustas (2010) proposed a source partitioning method (SK10 in the following) to estimate contributions of transpiration, evaporation, photosynthesis, and respiration to H2O and CO2 fluxes obtained by the eddy covariance method. High frequency eddy covariance raw data time series are needed, and the source partitioning is estimated based on separate application of the flux-variance similarity theory to stomatal and non-stomatal components of the regarded fluxes, as well as on additional assumptions on leaf-level water use efficiency (WUE). We applied SK10 to data from two test sites (forest and cropland) and analyzed partitioning results depending on various ways to estimate WUE from available data. Also, we conducted large eddy simulations (LES), simulating the turbulent transport of H2O and CO2 for contrasting vertical distributions of the canopy sinks/sources, as well as for varying relative magnitudes of soil sources and canopy sinks/sources. SK10 was applied to the synthetic high frequency data generated by LES and the effects of canopy type, measurement height, given sink-source-distributions, and input of varying WUEs were tested regarding the partitioning performance. SK10 requires that the correlation coefficient between stomatal and non-stomatal scalar fluctuations is determined by the ratio of the transfer efficiencies of these scalar components, an assumption (transfer assumption in the following) that could be tested with the generated LES data. The partitioning results of the field sites yielded satisfactory flux fractions, when fair-weather conditions (no precipitation) and a high productive state of the vegetation were present. Further, partitioning performance with regard to soil fluxes increased with crop maturity. Results also showed relatively large dependencies on WUE, where the partitioning factors (median) changed by around -57% and +36%. Measurements of outgoing longwave radiation used for the estimation of foliage temperature and WUE could slightly increase the plausibility of the partitioning results in comparison to soil respiration measurements by decreasing the partitioning factor by up to 42%. The LES-based analysis revealed that for a satisfying performance of SK10, a certain degree of decorrelation of the H2O and CO2 fluctuations (here, |ρq'c’| < 0.975) was needed. This decorrelation is enhanced by a clear separation between soil sources and canopy sinks/sources, and for observations within the roughness sublayer. The expected dependence of the partitioning results on the WUE input could be observed. However, due to violation of the abovementioned transfer assumption, the known true input WUE did not yield the known true input partitioning. This could only be achieved after introducing correction factors for the transfer assumption, which were known however only in the special case of the LES experiments.

AB - Scanlon and Sahu (2008) and Scanlon and Kustas (2010) proposed a source partitioning method (SK10 in the following) to estimate contributions of transpiration, evaporation, photosynthesis, and respiration to H2O and CO2 fluxes obtained by the eddy covariance method. High frequency eddy covariance raw data time series are needed, and the source partitioning is estimated based on separate application of the flux-variance similarity theory to stomatal and non-stomatal components of the regarded fluxes, as well as on additional assumptions on leaf-level water use efficiency (WUE). We applied SK10 to data from two test sites (forest and cropland) and analyzed partitioning results depending on various ways to estimate WUE from available data. Also, we conducted large eddy simulations (LES), simulating the turbulent transport of H2O and CO2 for contrasting vertical distributions of the canopy sinks/sources, as well as for varying relative magnitudes of soil sources and canopy sinks/sources. SK10 was applied to the synthetic high frequency data generated by LES and the effects of canopy type, measurement height, given sink-source-distributions, and input of varying WUEs were tested regarding the partitioning performance. SK10 requires that the correlation coefficient between stomatal and non-stomatal scalar fluctuations is determined by the ratio of the transfer efficiencies of these scalar components, an assumption (transfer assumption in the following) that could be tested with the generated LES data. The partitioning results of the field sites yielded satisfactory flux fractions, when fair-weather conditions (no precipitation) and a high productive state of the vegetation were present. Further, partitioning performance with regard to soil fluxes increased with crop maturity. Results also showed relatively large dependencies on WUE, where the partitioning factors (median) changed by around -57% and +36%. Measurements of outgoing longwave radiation used for the estimation of foliage temperature and WUE could slightly increase the plausibility of the partitioning results in comparison to soil respiration measurements by decreasing the partitioning factor by up to 42%. The LES-based analysis revealed that for a satisfying performance of SK10, a certain degree of decorrelation of the H2O and CO2 fluctuations (here, |ρq'c’| < 0.975) was needed. This decorrelation is enhanced by a clear separation between soil sources and canopy sinks/sources, and for observations within the roughness sublayer. The expected dependence of the partitioning results on the WUE input could be observed. However, due to violation of the abovementioned transfer assumption, the known true input WUE did not yield the known true input partitioning. This could only be achieved after introducing correction factors for the transfer assumption, which were known however only in the special case of the LES experiments.

KW - Flux partitioning

KW - Latent heat flux

KW - LES

KW - Net ecosystem exchange

KW - Sensitivity analysis

KW - Water use efficiency

U2 - 10.1016/j.agrformet.2018.11.003

DO - 10.1016/j.agrformet.2018.11.003

M3 - Article

VL - 265

SP - 152

EP - 170

JO - Agricultural and Forest Meteorology

JF - Agricultural and Forest Meteorology

SN - 0168-1923

ER -