Measuring and modeling the impact of droughts on the Amazon

carbon cycle

Wouter Peters Gerbrand Koren, John Miller, Luciana Gatti,

Manuel Gloor, Ingrid Luijkx, Stijn Naus, Liesbeth FlorentieASICA

1

Samples collected by extractionN=1489 (+600 in 2019 )

0

3

6

9

12

15

18

19/07/2016 21/09/2016 02/12/2016 08/02/2017 08/05/2017 02/08/2017 23/11/2017 12/03/2018 18/07/2018

RBAALFSANTEFPANMANTARGET

2

¹³C16O16O12C16O16O 12C16O18O12C16O17O

~4.3 µm (2310 cm-1)

Aerodyne TILDAS dual-laser

3

Samples on Aerodyne TILDAS dual-laser(N=1285, also extracted)

4

stddev = ± 0.09 ppm

stddev = ± 0.03‰

Precision and Stability

δ¹³C of CO₂ in TARGET tank (treated as unknown)

CO₂ in TARGET tank (treated as unknown)

median = 0.008 ‰ (N=12)

median = 0.07 ppm (N=12)

95% CI of 1858 samples measured

5

12 Sep 2019 …Amazon Fires

δ¹³C of CO₂CO₂

Rio Branco profile collected by L. Gatti and co-workers, using V3 PFP including the ASICA air dryer

Aircraft Vertical Profiles from the Amazon Basin

CO₂: ±0.2 ppm δ¹³C: ± 0.03‰

6

Aircraft Vertical Profiles from the Amazon Basin

Elevated CO₂ with more negative δ¹³C => release of ‘light’ carbon with relatively much 12C

12 Sep 2019 …Amazon Fires

δ¹³C of CO₂CO₂

CO₂: ±0.2 ppm δ¹³C: ± 0.03‰

7

Aircraft Vertical Profiles from the Amazon Basin

Elevated CO₂ with more negative δ¹³C => release of ‘light’ carbon with relatively much 12C => C3 plants or fossil fuels (-24‰ — -30‰) => C4 plants (-12‰) sign of fires ? To be judged through CO…

12 Sep 2019 …Amazon Fires

CO₂: ±0.2 ppm δ¹³C: ± 0.03‰

δ¹³C of CO₂CO₂

8

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

9

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

10

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

11

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

12

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

13

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

*Amazon draft, June 4, 2020*

Figure 3: Annual anomalies of (top) precipitation, soil moisture, and (bottom)NIRv, and SIF integrated over the CTSAM domain. Anomalies are expressedin standard deviation relative to the years 2010-2017. Note that the GRACEsatellites did not complete the year 2017 and we could thus not determine anannual anomaly for that year. The error bar for GLEAM shows the range betweenGLEAM v3.3a and v3.3b. Note that the SIFTER data is not included over thefull record, since the SIFTER data su↵ers from sensor degradation, which is mostpronounced in the later years (described in Sect. 2.2).

21

14