ICOS Science Conference, Session 3

1. Assessing decoupling of above and below canopy air
masses at a Norway spruce stand in complex terrain
Georg Jochera, Milan Fischera, Marian Pavelkaa, Ladislav Šiguta, Pavel Sedláka, Gabriel Katulb
aDepartment of Matter and Energy Fluxes, Global Change Research Institute, Czech Academy of Sciences, Bělidla 986/4a,
603 00 Brno, the Czech Republic
bNicholas School of the Environment and Earth Sciences, Box 90328, Duke University, Durham, NC 27708-0328, U.S.A.

2. Schedule
• Decoupling problem at high vegetation sites
• Measurement site Bílý Kříž, Czech Republic
• Approaches to address decoupling
• Decoupling characteristics (1 year example data)
• Summary and outlook

3. Thomas et al., 2013
Decoupling problem at high vegetation sites: scheme. Decoupling yields missing flux (mainly
respiration) components in the above canopy derived EC data

4. Bílý Kříž
• Beskidy mountains, Czech Republic
• 49°30′N, 18°32′E; 800–900 m a.s.l.
• ~ 40 years old Norway Spruce stand
• LAI ~ 9 m2 m-2, stand height ~ 18 m
• EC since more than 10 years
• Tower close to a ridge
• 2017: additional EC below canopy
Measurement site Bílý Kříž, Czech Republic

5. Treating EC derived CO2 fluxes
Quality checking and flagging (Foken et al., 2004)
single-level filtering approaches:
• u*-filtering
• σw-filtering
• ………….
two-level filtering approaches:
• σw-filtering
Newly proposed:
• Telegraphic approximation raw w
• Cross correlation maximum w
Approaches to address decoupling

6. Approaches to address decoupling: σw approach
Linear range of the
relation of σw below
and above the
forest canopy:
coupled conditions.
The values of σw
which mark the
beginning of this
range can be used
as flux filtering
thresholds.

7. Telegraphic approximation (TA) for each 30 min. period:
Steps conducted for both above and below canopy vertical wind raw data
)(`
wmeanww 
assignment of 1 for positive w`, 0 for negative w`
Agreement between both above and below canopy data:
TA agreement (TAa) = sum (TA above = TA below = 1) / length data set
factor between 0 and 1 as information about the degree of coupling
Approaches to address decoupling:
telegraphic approximation

8. Cross Correlation Function Maximum (CCFmax) for each 30 min. period:
• Computation of cross correlation
function for above and below
canopy raw data (gives values in
the range -1 to 1)
• Identification of maximum of cross
correlation function around lag 0
• Value of this maximum as coupling
indicator within a given half hour
Approaches to address decoupling:
cross correlation maximum

9. Decoupling characteristics (Bílý Kříž)
median: 0.54 median: 0.09
Telegraphic approximation agreement Cross correlation function maximum

10. Decoupling characteristics (Bílý Kříž)
TAa and CCFmax follow
the yearly course of
global radiation and
air temperature
(influence of thermal
turbulence on
coupling).
No correlation with u*
was observed.

11. Decoupling characteristics (Bílý Kříž)
TAa and CCFmax follow
the daily course of
global radiation and
air temperature
(influence of thermal
turbulence on
coupling).

12. Correlation
coefficients
Rg
(W m-2)
Bf
(m s-2)
Bfr
(-)
u*
(m s-1)
shear
(°)
TAa 0.22 0.04 -0.01 -0.04 0.22
CCFmax 0.26 0.05 -0.01 0.07 0.12
Table 1. Correlation coefficients of TAa and CCFmax in relation to the
parameters Rg (W m-2), buoyancy forcing (Bf) across the canopy (m s-2),
buoyancy flux ratio (Bfr), u* (m s-1) and the wind directional shear
between above and below canopy wind direction (°).
Decoupling, as detected via TAa and CCFmax cannot be
sufficiently described by any other parameter. Consequently,
such kind of two-level analysis appears to be mandatory to
evaluate decoupling.
Decoupling characteristics (Bílý Kříž)

13. Decoupling characteristics (Bílý Kříž)
Threshold estimation
(exemplarily for
CCFmax): the rationale
is, that coupling is top-
down induced. As soon
as an increase in
turbulence above
canopy yields an
increase of turbulence
below canopy, the ratio
σw will rise (red vertical
line). If full coupling
occurs, the ratio σw will
remain constant with
rising CCFmax (blue
vertical line).
0.05 0.13

14. A proposed new flux filtering strategy:
• Best quality flags (Foken et al., 2004)
• Time lag between above and below canopy signal < ± 20 s
(site/setup specific)
• CCFmax > 0.05
Available data fraction after these steps: 26 %;
quality filtering only leaves 45 %;
(additional filtering using σw thresholds: 33 %)
Decoupling characteristics (Bílý Kříž)

15. Decoupling characteristics (Bílý Kříž)
Effect of filtering on
cumulative C
exchange
hq: filtered for quality
fc: filtered for quality
+ σw two-level
threshold
new: filtered for
quality + time lag +
CCFmax threshold
More respiration
captured with
increasing strictness
of filtering

16. Take-home message
• Single-level filtering methods like u*-filtering might not work or be
insufficient at certain sites. Two-level filtering approaches can overcome
this problem.
• Decoupling appears to be site-specific depending on canopy properties
and tower-surrounding topography.
• The proposed innovative tools TAa and CCFmax appear to function and
have the potential to be a more direct alternative to the convential σw
two-level filtering, also for shorter time scales.
• All two-level filtering approaches yield less negative CO2 fluxes in
comparison to single-level filtering, indicating that the two-level
approaches are able to capture more ecosystem respiration components.
Additional below canopy EC measurements are an easy to verify innovative
tool to reduce the uncertainty of above canopy derived ecosystem carbon
fluxes, beneficial for research infrastructures like ICOS.

17. Thanks for your
attention
This work was supported by the Ministry of Education, Youth and Sports of CR within the CzeCOS program,
grant number LM2018123. MF, LŠ and MP were supported by the project SustES - Adaptation strategies for
sustainable ecosystem services and food security under adverse environmental conditions
(CZ.02.1.01/0.0/0.0/16_019/0000797).