This is a repository copy of Tropical land carbon cycle responses to 2015/16 El Niño as recorded by atmospheric greenhouse gas and remote sensing data. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/135234/ Version: Accepted Version Article: Gloor, E orcid.org/0000-0002-9384-6341, Wilson, C, Chipperfield, MP et al. (12 more authors) (Accepted: 2018) Tropical land carbon cycle responses to 2015/16 El Niño as recorded by atmospheric greenhouse gas and remote sensing data. Philosophical Transactions of the Royal Society B: Biological Sciences. ISSN 0962-8436 (In Press)
Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing
[email protected] including the URL of the record and the reason for the withdrawal request.
[email protected] https://eprints.whiterose.ac.uk/
Submitted to Phil. Trans. R. Soc. B - Issue
rR
Fo !"
#!
% "
&
$
'
iew
ev
( ) *+ , '& # ) . )% *+ , '& # ) % ' 0 1! , % ' #) *+ , '& # ) % , ) 2 3#3 * & %0) " ) . '- -* / ) )4 " *+ , '& ) " 5 7 ) ! *+ , '& ) "7 ) 5 *+ , '& ) " ( )& * /50 % ) % * /50 ( )& * /50 5 ). *. + , ) * /1 $0 " , ) * /% " % " , ) *+ , '& # )
'( '( '(
" " "
#
, #!
0# 1''
'
, '( -
/ /1
9 !: =
>
#
0 ,
"
%
%
!
; )( !
.
1&1(< " -) 2
)
http://mc.manuscriptcentral.com/issue-ptrsb
-
'5
6
'5 '5
6 6
"
" & !
ly
On
% # 8 # !
* /0 % /
2
$( !
-
Page 1 of 44
1 2 3 4
Emanuel Gloor, Chris Wilson, Martyn P. Chipperfield, Frederic Chevallier, Wolfgang
5
Buermann, Hartmut Boesch, Rob Parker, Peter Somkuti, Luciana Gatti, Caio Correia, Lucas
6
Gatti, Wouter Peters, John Miller, Merritt N. Deeter, Martin Sullivan
7 8
Fo
9
The outstanding tropical land climate characteristic over the past two2three decades is rapid
10
warming, while there are no significant large2scale precipitation trends. Warming is expected
11
to continue but effects on tropical vegetation are unknown. El Niño2related heat peaks may
12
provide a test bed of what to expect. Here we analyse tropical land carbon cycle responses to
13
the 2015/16 El Niño climate anomalies using an atmospheric transport inversion. Based on
14
the global atmospheric CO2 record and long2term behaviour of the global carbon cycle, we
15
find no obvious signs of anomalously large carbon release during the 2015/16 El Niño
16
compared to earlier El Niño events. We find roughly equal net carbon release anomalies from
17
Amazonia and tropical Africa, ~0.5 PgC each, and smaller carbon release anomalies from
18
tropical East2Asia and Southern Africa. Atmospheric CO anomalies reveal fire carbon release
19
from tropical East Asia peaking in October 2015 while Amazonian fires played a smaller
20
role. The positive flux anomalies are consistent with down2regulation of primary productivity
21
during peak negative near2surface water anomaly (Oct 2015 2 Mar 2016) as diagnosed by
22
solar2induced fluorescence. Finally, we find an anomalous positive flux to the atmosphere
23
from tropical Africa early in 2016, coincident with substantial CO release.
iew
ev
rR
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
24 25
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
-
0
'
27
%
28
al. 2009) and regulating climate by virtue of their exchanges of carbon, water and energy
29
with the atmosphere. They also play an important role in sustaining endangered fauna and
30
their continued presence is essential for preserving their rich biodiversity. More generally,
31
tropical biomes are home to great cultural diversity and growing economies, and will need to
32
support half the global population by 2050 (State of the Tropics Report, 2017). Thus they
33
have a large impact on livelihoods in these climates. The continued functioning and
34
productivity of vegetation in the tropics is, however, dependent on its response to changing
35
climatic conditions. The dominant climate change signatures across the tropics are rapid
36
warming and an increase of extreme events, severe floods and anomalously dry conditions
37
(Fig. 1, and e.g. Fischer and Knutti 2015).
=
'
?)5@
%
512), storing two2thirds of global plant biomass (e.g. Pan et
ev
rR
Fo
38
iew
39
El Niño events may provide a test bed to examine tropical vegetation responses, likely to be
40
dominated by the response of forests, to these increasingly higher temperatures, paralleled
41
usually by drier than usual conditions. This is because during El Niño events strong positive
42
temperature excursions tend to be spatially correlated with negative precipitation anomalies.
43
These anomalies occur in tropical South2East Asia, tropical South America and to a lesser
44
extent tropical West Africa and Southern Africa (roughly below 10°S, e.g. Dai et al. 2000,
45
Foley et al. 2002).
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 44
46 47
It has been known since at least the 1970s that El Niño events co2occur with periods of
48
anomalously large atmospheric CO2 growth rates (Bacastow et al. 1976). There is not only a
49
strong correlation between the El Niño index (in essence atmospheric sea surface pressure
50
difference between Darwin and Tahiti in the tropical Pacific) and global atmospheric CO2
http://mc.manuscriptcentral.com/issue-ptrsb
Page 3 of 44
)1
1
%
1 &
1
52
temperature anomalies (e.g. Wang et al. 2014). The mechanism causing the correlation with
53
temperature is not entirely clear. One component is increased biomass burning. It has further
54
been argued that water limitation on vegetation performance is important at the local scale,
55
but temperature anomalies are more important at larger scales due to cancelling effects (Jung
56
et al. 2017). The strong correlation between atmospheric growth rate anomalies and El Niño
57
Index suggests that variation of tropical land carbon uptake and release contributes
58
prominently to anomalous atmospheric CO2 growth during positive El Niño phases.
59
Nonetheless, ocean air2sea gas exchange does also play a role. Measurements of this process
60
in the tropical Pacific reveal that during El Niño outgassing in the tropics is reduced, i.e.
61
ocean carbon pool response is in the opposite direction of land carbon pools (Feeley et al.
62
1999). This is because the tropical Pacific thermocline upward tilt towards South America is
63
reduced during El Niño phases of ENSO (El Niño Southern Oscillation) which hinders
64
upwelling of carbon2rich waters along the tropical South American west coast and thus
65
carbon efflux from the sea to the atmosphere is reduced. The decrease over a full El Niño
66
period for the 1997/98 event has been estimated using ocean data to be 0.6±0.1 PgC (Feely et
67
al. 1999, and 2002). Recent estimates of global air2sea gas exchange based on air2sea partial
68
pressure difference measurements and gas exchange parameterization by Feely et al. (2017)
69
suggest a smaller anomaly over the 2015/16 period on the order of 0.120.2 PgC.
iew
ev
rR
Fo
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
70 71
Similar to the other manuscripts in this volume we attempt to analyse whether, and to what
72
extent, the response of tropical land vegetation during the 2015/16 El Niño event is different
73
from responses during previous similar events, and thus may be a harbinger of future
74
responses not just to climate oscillations but climate variation on top of a rapidly increasing
75
temperature background. There have been, for example, reports indicating that increased
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
.-
+6
% %
77
=
517). There have also been reports indicating that dry seasons may get
78
drier across the tropics (Murray2Tortarola et al. 2017). It is not clear what the effect of
79
tropical vegetation productivity may be given both stimulating (rising CO2) and limiting (e.g.
80
the increase in leaf2air water vapour pressure difference) factors. The overarching theme of
81
this manuscript is thus whether the 2015/16 events reveal signs of anomalously increasing
82
vegetation strain.
83
Fo
84
Our analysis is based primarily on a large2scale atmospheric approach which as the main tool
85
uses an inverse model of atmospheric transport (INVICAT, Wilson et al. 2014) to extract
86
information about surface CO2 exchange between land vegetation and atmosphere contained
87
in spatio2temporal variations of atmospheric CO2. Our approach is helped by new data in
88
tropical South America measured by INPE (Instituto Nacional de Pesquisas Espaciais), Sao
89
Jose dos Campos, Brazil. We focus on inter2annual variation of fluxes which should be more
90
robustly estimable than absolute flux magnitudes. To put our results into context we relate the
91
fluxes to climate controls and to distinguish processes, to some extent, we employ solar
92
induced chlorophyll fluorescence and atmospheric carbon monoxide measured from space.
93
We aim to address the following questions: How anomalous is the global CO2 flux anomaly?
94
What are the climate deviations / excesses on land? Where and when do flux anomalies occur
95
and how large are they? How much is due to fire and under what conditions? How much is
96
due to reduction in primary production versus changes in respiration? And finally, are there
97
signs of land vegetation responses outside the usual El Niño patterns given the unprecedented
98
temperatures during the 2015/16 event?
iew
ev
rR
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 44
99 100
!
"
http://mc.manuscriptcentral.com/issue-ptrsb
Page 5 of 44
101
#
102
A possible approach to estimate atmospheric CO2 growth rate anomalies Δ , suggested to our
103
knowledge first by Jones et al. (2001), is as follows
104 Δ (t) =
( )−
∙
( )
105
(1)
106
Here
107
fossil fuel burning and cement manufacture, and
108
the ratio of the annual atmospheric carbon growth rate and fossil fuel emissions. Thus
109
∙
is atmospheric carbon content (in form of CO2), is time,
is global emissions from
is the long2term mean airborne fraction,
Fo
( ) is the expected average increase of atmospheric carbon growth rate for given
rR
110
fossil fuel emissions
in year . Fossil fuel emission estimates used here are from Boden
111
et al. (2017)
112
which are based on energy statistics, and the observed atmospheric CO2 record (the Mauna
113
Loa record)
114
on December 31/January 1 (i.e. mean from July 1 to June 30). As a sensitivity test we have
115
repeated this calculation using
116
conclusions.
iew
ev
~= 0.55. Growth rate is calculated as the difference of annual means centred
118
$
% &
%
$'
ly
117
=0.49 (mean over 190122015) with similar overall
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
#
(
%
119 120
The most robust information provided by atmospheric CO2 concentration records is the
121
global atmospheric inventory and how it changes over time. This reveals, for example, very
122
clearly the well2known rapid increase of atmospheric CO2 over the last decades. In addition
123
to global information the wide2spread surface station observation network, maintained by
124
various groups and in particular NOAA/ESRL (Fig. S1), exhibit spatio2temporal patterns
125
which reflect regional scale variation in CO2 exchange between the land surface and oceans
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
126
1
0
% &
127
distribution and strength of regional surface fluxes, provided the relationship between fluxes
128
and the concentration patterns they cause can be established. This relationship involves
129
representation of the processes of atmospheric advection and dispersion which can be
130
estimated fairly well using numerical fluid dynamics models of the atmospheric flow
131
('atmospheric transport models') (e.g. Chipperfield et al. 2006). The relation to the actual
132
atmospheric flow is established by using wind and cloud convection transport fields derived
133
from regular observations of the state of the atmosphere for purpose of weather prediction.
134
Flux estimation reduces then to a least2squares minimization problem of the difference
135
between a linear combination of concentration fields resulting from localised fluxes in space
136
and time sampled at the same time and location as the observations, and the actual
137
observations. This problem turns out to be poorly constrained by the number of available in
138
situ measured data and thus a possible approach is to instead optimally combine a set of prior
139
flux 'guesses' %p with the flux estimates which replicate concentration data most closely
140
(Enting et al. 1985), i.e. to minimize
iew
ev
rR
Fo
On
141 ( )=( −
) ∙
∙( −
)+( − ) ∙
142
∙ ( − ) ( )
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 44
143
with respect to %. ) is the a priori flux error covariance matrix,
144
observed atmospheric concentrations and * is the transport2model2calculated matrix, which
145
relates surface fluxes to the atmospheric concentration signal they cause at the sampling sites.
146
This approach solves in essence for small deviations from a prescribed flux model. For this
147
problem an explicit expression for the posterior flux error covariance matrix
148
derived (Tarantola and Valette, 1982):
post
is a vector containing the
=[*t ∙ +21∙* ,)21-21 .
149
http://mc.manuscriptcentral.com/issue-ptrsb
post
can be
Page 7 of 44
150
1
%
151
transport model TOMCAT (Chipperfield, 2006). We resolve fluxes monthly and spatially on
152
a grid 5.6o × 5.6o longitude by latitude. The prior flux model includes three components: (i)
153
annually changing fossil fuel emissions, (ii) monthly net land gains or losses which do not
154
change from year to year, based on the CASA (Carnegie Ames Stanford) land biosphere
155
model (average climatology for 200322011). The model estimates primary productivity as the
156
product of solar photo2synthetically active radiation (PAR), land vegetation chlorophyll
157
content (estimated from space NDVI) and a light use efficiency. Respiration is estimated
158
using a carbon cycle model which includes soils (Olsen and Randerson, 2004), and (iii) air2
159
sea fluxes.
ev
rR
160
Fo
161
We have chosen annually repeating and balanced land vegetation 2 atmosphere CO2 flux prior
162
estimates because our interest is in extracting the information on inter2annual variations
163
contained in atmospheric data. For our prior estimates of air2sea fluxes we treat separately the
164
fluxes associated with the pre2industrial carbon cycle (two hemispherical loops with CO2
165
outgassing in the tropics and CO2 uptake at high latitudes) and uptake of carbon induced by
166
the anthropogenic perturbation of atmospheric CO2 (with fluxes steadily increasing and
167
located primarily in the Northern Atlantic and Southern Ocean) (Gloor et al. 2003, Khatiwala
168
et al. 2009). For the former we use the monthly resolved climatology based on air2sea partial
169
pressure differences and an air2sea gas exchange coefficient parameterization, compiled by
170
Takahashi et al. (2009), to which we add a constant and spatially uniform flux such that the
171
fluxes are globally in balance on an annual basis. For the latter we used the spatial air2sea
172
flux pattern of Khatiwala et al. (2009) (their Fig. 1b) which we scale with global net ocean
173
uptake taken from the Global Carbon Project analysis (Global Carbon Project, 2017). This
174
approach leads to improved a posteriori data model fits compared to inversions which use as
iew
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
175
2
4
%
2 2
176
measurement differences alone (specifically Takahashi et al. 2009). Minimisation of (%) is
177
done using a quasi2Newtonian method (L2BFGS implemented in M1QN3 minimiser) with
178
gradients calculated with the adjoint of the TOMCAT atmospheric transport model,
179
ATOMCAT (Wilson et al. 2014). We assumed a prior flux uncertainty of 200% per grid cell
180
and we assumed that there is no flux error correlation given the comparably coarse resolution.
181
Atmospheric data are from 81 sites mainly measured by NOAA/ESRL and include in
182
addition the PBL mean (below 2500 m) and the free troposphere mean (above 2500 m) of
183
vertical profile data in the Amazon measured by INPE, Sao Jose dos Campos, Brazil, (e.g.
184
Gatti et al. 2014) (Fig. S1 and S2). Observational data has an uncertainty of 1 ppm, plus an
185
estimate of representation error derived by averaging the absolute prior concentration
186
variation between the model grid cell containing the measurement location and the
187
surrounding grid cells. This leads to overall observational uncertainties of between 1 and 6
188
ppm, depending on the measurement location. To assess the influence of the Amazon vertical
189
profile data we have also performed an inversion without these data. The effect of including
190
the data is to reduce the magnitude of flux anomalies while the timing and location of
191
anomalies are not affected much (section 3).
iew
ev
rR
Fo
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 44
192 193
./ (
%(
#
194
One cause of plant water stress is negative deviations (or anomalies) of the abundance of soil
195
water (or soil water content) from the climatological mean representative for a region. A
196
proxy for soil water content over large spatial scales can be measured from satellites because
197
large2scale land surface water content anomalies cause Earth gravity anomalies. Such gravity
198
anomalies are being estimated from space by the twin satellite mission GRACE (Gravity
199
Recovery and Climate Experiment, Tapley et al., 2004) with the satellites following each
http://mc.manuscriptcentral.com/issue-ptrsb
Page 9 of 44
55
%
51
B
1
1
202
again once the front satellite has passed the anomaly and the rear satellite is approaching the
203
anomaly. To confirm the realism of the gravity anomaly data measured from space we
204
compared gravity anomaly anomalies with precipitation anomalies measured by TRMM
205
(Tropical Rainfall Measuring Mission, Huffman et al. (2007), Fig. S4). The signatures of the
206
two data types are very consistent (taking into account that gravity anomalies are to first
207
order equal to cumulative precipitation anomalies). To calculate monthly gravity anomaly
208
anomalies we subtracted monthly means calculated using the full 200222016 record from the
209
continuous record of monthly means.
210 211
%% &
(
%
iew
212
0
ev
rR
Fo
213
We use daytime CO air column inventories estimated from MOPITT radiometer data on the
214
TERRA satellite (Deeter et al. 2013) to estimate carbon emissions from fires. To do so we
215
first estimate carbon monoxide fluxes from monthly CO air column anomalies. We then
216
convert the carbon monoxide fluxes to carbon fluxes, assuming they are from fires, by
217
multiplying the carbon monoxide fluxes with a biomass burning emission ratio of (1/74)
218
(ppm CO2)/(ppb CO) (Gatti et al. 2010) although this ratio may vary with the type of fire.
219
The MOPITT CO record we are using is version 6 (version 6 (L3V95.2.3), Deeter 2013)
220
which covers March 2000 to December 2016. This version uses both thermal infrared (TIR)
221
and near infrared (NIR) radiances and so, compared to the other two MOPITT products (TIR2
222
only and NIR2only), it provides the maximum sensitivity to surface2level CO. Nonetheless,
223
because of the non2uniform weighting function of the retrievals, column content estimates
224
may contain a bias (an underestimate of column CO if signals are concentrated to the lower
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
)
> =
')> 0
2
. / 3
0 9
4
8 1
4 =
5 1
8
+
!"
+ #
%$≈ 0.85
1
%
DE − #!
!&
− ∇ ∙ (() ∙ *) 8 0
4
≈ 9 23
≈ 0.1 23 lifetime of CO in the atmosphere, #
%$#!
232
atmosphere, 0
233
to oxidation of non2methane hydrocarbons,
234
to the total air volume above a fixed region to obtain a relation between a CO flux
235
perturbation 5 at the earth surface and the 5() anomalies it causes:
236
5 =
237
normal vector to the vertical walls :; orthogonal to the surface region of interest, and df is
238
an infinitesimal area element of :;. The contribution of in2 and outflow into and out of the
239
air volume above the region is negligible if region boundaries are chosen such that Δ() ≈ 0.
Fo
air flow velocity vector. We apply the equation
+
6
− 7 8 5() ∙ < , > df ≃
6
+
6 !&
where,
is an outwards directed unit
iew
!&
ev
6
1
%
21 34
ly
On
240 241
CO volume source due
rR
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 44
242
Photosynthesis is associated with fluorescence. A small fraction of solar radiation trapped by
243
chlorophyll escapes instead of being used to fix CO2. This fraction is reemitted into the
244
atmosphere from the leaf at larger wavelengths (in the range of 670 and 800 nm, e.g. Meroni
245
et al. 2009) compared to the originally trapped radiation. Fluorescence has been shown to be
246
related to productivity (Yang et al. 2015, van der Tol et al. 2014) and thus we use it here as a
247
proxy for productivity. We specifically use here SIF retrieved from GOSAT (Greenhouse
248
Gases Observation Satellite, Kasuya and Hamazaki, 2005) measurements at 772 nm using the
http://mc.manuscriptcentral.com/issue-ptrsb
Page 11 of 44
249
2%
9
%
& %
= 511). The bias
250
correction2procedure, which is an essential part of the post2retrieval processing, was
251
performed using the European Space Agency Climate Change Initiative land cover maps
252
(Bontemps et al. 2013). GOSAT measurements over permanently non2vegetated areas (where
253
zero fluorescence can be assumed) were identified using these maps in order to derive
254
radiance2dependent calibration curves on a monthly basis. Based on these monthly curves, a
255
2D2spline interpolation was used along time and radiance dimensions to obtain the bias
256
correction term for any given GOSAT sounding. This ensures that the time2dependence of
257
the instrument2related bias is taken into account. The retrievals are available for the period
258
April 2009 to September 2016.
rR
Fo
259
ev
260
Monthly compilations of SIF retrievals exhibit missing values. To obtain sufficient data
261
coverage we therefore calculated quarterly (32monthly) means and similarly calculated
262
anomalies for three2month periods. Despite lumping three months together there are still
263
pixels with no retrievals. To calculate quarterly anomalies we therefore kept track of number
264
of existing retrievals on an individual pixel basis. For comparison of anomalies for specific
265
areas and 32monthly periods we calculated region mean anomalies including only those
266
pixels for which retrievals exist.
iew
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
267 268
. +
269
.
* #
270
( (
5
271
The largest annual global atmospheric CO2 increase rates recorded with modern analytical
272
tools (i.e. since 1959) occurred in 2015 and 2016 with values of 2.94 and 2.85 ppm,
273
respectively
(NOAA/ESRL,
Boulder,
http://mc.manuscriptcentral.com/issue-ptrsb
Colorado,
USA;
Submitted to Phil. Trans. R. Soc. B - Issue
.!
E77
7
7
7
7
P P
4> 1
275
1998 (2.81 ppm). While of concern per se, to detect changes of El Niño land
276
vegetation responses at the global scale, the fossil fuel contribution to the growth rate needs
277
to be separated from other flux contributions, in particular nonlinearly increasing fossil fuel
278
emissions. As explained in Section 2.1 to achieve this we assume a constant fossil fuel
279
airborne fraction which we subtract from the atmospheric carbon inventory growth rate
280
(Figure 2). The anomalies in 2015 and 2016 were positive and when summed up were
281
approximately 1.7 PgC, with the 2016 anomaly being approximately twice as large as the
282
2015 anomaly. Including the reduction in CO2 outgassing from tropical oceans during
283
positive El Niño episodes based on the air2sea flux estimates summarized in Feely et al.
284
(2017), then the total flux anomaly of global land carbon to the atmosphere is ~ 1.92 2.1 PgC
285
over the two years (201522016).
iew
286
ev
rR
Fo
287
A comparison with the 1997/98 El Niño anomaly reveals that the 2015/16 anomaly was not
288
extraordinarily large, certainly of a smaller magnitude than the 1997/98 anomaly. This
289
conclusion does not depend much on the period chosen to estimate the airborne fraction
290
(Section 2.1). From a vegetation process response point of view, the 1997/98 anomaly is,
291
however, somewhat unusual in that it includes a strong direct human2impact large2scale peat
292
drainage component which in 1997/98 led to 'catastrophic' peatland / peat forest fires and
293
carbon release (Page et al., 2002). Thus, part of the 1997/98 positive anomaly is unrelated to
294
climate2induced variation in productivity and respiration of living vegetation or soil
295
respiration in a strict sense. A noticeable indirectly El Niño2related aspect revealed by growth
296
rate anomalies (Fig. 2) is the negative (land carbon uptake) anomalies from roughly 2008
297
onwards.
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
298
http://mc.manuscriptcentral.com/issue-ptrsb
Page 12 of 44
Page 13 of 44
33 55 51
# B
4 =
>
1
+6
7
4 2
2
302
usual conditions, we briefly summarize here measures of vegetation stress related to climate.
303
The first and primary measure is plant water stress caused by negative deviations (or
304
anomalies) of the abundance of soil water (or soil water content) from the climatological
305
mean representative for a region. We use here monthly gravity anomaly anomalies measured
306
by GRACE (Section 2.3) as a proxy for soil water stress. Although these anomalies include
307
both below and aboveground water anomalies our use here as a vegetation water stress
308
indicator is supported by anti2correlation between annual pan2tropical land / tropical South
309
American land gravity anomaly anomalies and global atmospheric CO2 growth rate
310
anomalies g (see Section 2.1) shown in Fig. 3 (Pearson r=20.69 and 20.72, and p=0.0046 and
311
0.0025, respectively).
iew
ev
rR
312
Fo
313
The main features of vegetation water deficits during the 2015/16 El Niño period according
314
to both gravity anomaly as well as precipitation data (Fig. 3, Fig. S4) are (i) in the Amazon
315
basin, an east2to2west spreading and steadily increasing area with large water deficit, with
316
this process starting at the beginning of 2015. The deficit reaches its peak and covers the
317
entire basin by the first quarter of 2016, with water deficit remaining high throughout the
318
basin until the final quarter of 2016; the most pronounced negative precipitation anomaly
319
occurred during the final quarter of 2015 all across the basin; (ii) in Africa one notable signal
320
is a considerable water deficit developing south of roughly 10°S during the first three
321
quarters of 2016. The anomaly is not as strong as for Amazonia. The most pronounced
322
precipitation anomaly for the region southward of 10°S occurred during the final quarter of
323
2015, i.e. a bit earlier than in the other two regions. A second notable feature is excessively
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
!
%
325
>
%
=
9
:
O
:
516, while according to gravity anomalies and precipitation
326
estimated by TRMM there were no very clear indications of drought conditions but this will
327
need further investigation; (iii) tropical South2East Asia experienced strong negative
328
precipitation anomalies and associated water stress during the second half of 2015. These
329
three regions experienced strongly elevated temperatures nearly synchronously with the
330
substantially drier than usual conditions, with peak temperatures all exceeding existing
331
historical records (Figs. 4 & S4).
Fo
332 333
Among the three continents the climate anomalies for the Amazon seem to be the strongest
334
with temperature and precipitation anomalies centred around the last three months of 2015
335
and first three months of 2016, and with the effects of precipitation anomalies on soil
336
moisture lasting over nearly all of 2016, reflecting the time it takes for water deficits to
337
propagate through the catchment (Fig. S3).
339
Overall the observed climate anomalies are similar to the canonical El Niño patterns as
340
described e.g. by Dai and Wigley (2000), with the tropical Asian precipitation anomaly being
341
somewhat weaker. The possible exception is the Congo basin, which was excessively hot
342
during the first quarter of 2016.
ly
On
338
iew
ev
rR
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 44
343 344
$
% &
%
$'
(
%
345 346
What do atmospheric inversion results suggest? Motivated by the relation between inter2
347
annual variation of the global atmospheric CO2 growth rate and gravity anomaly anomalies
348
on tropical land we compare land CO2 flux anomalies with gravity anomaly anomalies for
http://mc.manuscriptcentral.com/issue-ptrsb
Page 15 of 44
!3
'
:
%
)5
$: 4
=
)> 0
4
1
)1
0
352
synchronicity of positive flux anomalies (fluxes to the atmosphere) with negative gravity
353
anomaly anomalies and vice versa (Pearson r=20.42 for monthly means, p < 1023), which is
354
consistent with the global record (Fig. 2). This result demonstrates the inversion’s ability to
355
detect and attribute expected flux anomalies from the atmosphere data. Splitting up the flux
356
estimates by those regions with notable climate anomalies we find the following (Fig. 6).
357
According to our calculations four regions released significant amounts of carbon during the
358
201522016 period: Tropical South America, Tropical Africa, Southern Africa and Tropical
359
East Asia. The losses from tropical South America and tropical Africa are similar in
360
magnitude while losses from Tropical East Asia and Southern Africa are smaller (Table 1).
361
The timing of peak carbon losses differs between the regions, with peak in October 2015 for
362
tropical East Asia, February 2016 in South Africa, February and March 2016 for tropical
363
Africa November2December 2015 and March2April 2016 for tropical South America.
iew
ev
rR
Fo
364 365
Table 1 Tropical land carbon flux anomalies Period
ly
Region
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
Net carbon flux to Biomass atmosphere
carbon flux
(PgC)
(PgC)
Tropical S America
Sep 2015 2 Jun 2016
0.5 ± 0.3
0.0520.1
Tropical Africa
Nov 2015 2 Jul 2016
0.6 ± 0.3
0.0820.16
Southern Africa
Jan 2016 2 May 0.2 ± 0.1 2016
http://mc.manuscriptcentral.com/issue-ptrsb
burning
Submitted to Phil. Trans. R. Soc. B - Issue
0
'
:
'
5152 Dec 2015 0.2 ± 0.1
0.320.4
366 367 368
!
369
CO2 estimates based just on atmospheric CO2 concentration data and inversion of
370
atmospheric transport provide net fluxes but cannot discern between the different underlying
371
processes, such as biomass burning, or changes in vegetation productivity and respiration
372
processes (e.g. by living trees /vegetation and/or dead organic matter in soils). Here in
373
addition we analyse information from atmospheric total column carbon monoxide (CO)
374
retrieved from MOPITT (Measurements of Pollution in the Troposphere) radiometer for the
375
period 200022017 on the NASA TERRA polar orbiting satellite, as an indicator of release of
376
carbon via fire (Fig. 7) (Deeter et al. 2016) and solar2induced fluorescence (SIF) retrieved
377
from GOSAT radiance measurement as an indicator of land vegetation productivity and
378
covering the period April 2009 to September 2016 (Fig. 8).
% &
380
Monthly CO air column anomalies, Δ %
515 and to lesser extent
392
subsequent months. When spatially integrated over tropical South America the decrease
393
during October2December 2015 is ~20%. To obtain a rough estimate of the associated
394
decrease in carbon uptake we use an estimate of tropical South American vegetation annual
395
productivity (gross primary productivity, GPP) estimated by Jung et al. (2011) based on CO2
396
flux measurement between atmosphere and forest canopies. The annual productivity of
397
tropical South American vegetation according to Jung et al. (2011) is ~18 PgC. According to
398
SIF we obtain a 32month reduction in productivity in this region of 20% and thus obtain a
399
reduction of carbon uptake of ~0.9 PgC during this quarterly period. Given limited evaluation
400
of the SIF2GPP relationship in the tropics this estimate needs to be taken with some caution.
401
For the other quarterly periods fluorescence anomalies are lower and signs of change less
402
coherent across large regions.
iew
403
ev
rR
Fo
404
!
405
From a climate perspective the outstanding development in the tropics on land over the past
406
decades is rapid warming. The 2015/16 El Niño adds a positive temperature anomaly on top
407
of
408
experiment of tropical forest vegetation subject to high temperatures in the future. Because of
409
the elevated background temperatures vegetation responses might be more severe compared
410
to responses observed during previous El Niño events. Based on just the global atmospheric
411
CO2 record we do not find any obvious sign of anomalously large carbon release during the
412
2015/16 El Niño compared to El Niño events in the past. This does not exclude compensating
413
effects at continental to regional scales. At these scales there is a strong spatial correlation
414
between positive temperature peaks and negative soil water anomalies diagnosed via gravity
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
this already rapidly warming 'background'. It thus provides a natural sensitivity
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
!15
'
1
4
%
%
416
exchange anomalies which is indeed consistent with correlation of anomalies on land with the
417
global atmospheric CO2 record. By far largest negative near surface water content anomalies
418
occurred in the Amazon Basin during the final quarter of 2015 and the first quarter of 2016.
419
Negative anomalies also occurred during a brief period in tropical East Asia centred on
420
October 2015 and Southern Africa during the first two quarters of 2016. We estimated
421
continental scale CO2 flux estimates based on atmospheric concentration data from surface
422
station networks complemented by vertical profile data in the Amazon. The results of these
423
calculations should be taken with some caution, because uncertainty in model transport can
424
lead to biased flux results (e.g. Gurney et al., 2002). Despite biases, inter2annual variability
425
may still be robust because transport modelling biases will affect all years in a similar way,
426
meaning that correlations with environmental variables can still be reliable. In our study, high
427
correspondence between tropical South American flux anomalies and negative precipitation
428
anomalies gives some confidence in the results, as well as the covariation in time with
429
climate anomalies and atmospheric CO anomalies. We find roughly equal net flux anomalies
430
from Amazon and tropical Africa of around 0.5 PgC each, and somewhat smaller positive
431
flux anomalies from tropical East Asia and Southern Africa. According to atmospheric CO
432
anomalies our analysis attributes anomalous carbon release from tropical East Asia to fires
433
peaking in October 2015, while consistent with fluorescence data from space, biomass
434
burning played a smaller role in the Amazon and with the flux anomaly reasonably consistent
435
with down2regulation of primary productivity during peak negative water anomaly (final
436
quarter of 2015 and first quarter of 2016). The one feature in our results which seems
437
somewhat unexpected, as this is not usually a region considered to be affected significantly
438
by El Niño, is the anomalous flux from tropical Africa coincident with substantial CO release
439
from the Congo Basin, during the first quarter of 2016. Our estimate of CO2 released by fires
iew
ev
rR
Fo
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 44
http://mc.manuscriptcentral.com/issue-ptrsb
Page 19 of 44
!!5
:
!!1
4
4
:
1
%
1 &1
%
$:
1
442
may have caused an anomalous decrease in productivity, SIF data do not give strong support
443
to this mechanism. Thus, in addition to changes in productivity, enhanced heterotrophic
444
respiration may have contributed also to this signal.
445 446
Finally, we examine how our results summarized together with main controls in Table 2
447
compare with the recent analyses of Liu et al. (2017) based primarily on satellite data. For the
448
comparison it is important to realise that the Liu et al. study calculated anomalies with
449
reference to flux estimates from the year 2011, a La Niña year. It is well established that
450
during La Niña years global CO2 growth rate anomalies are strongly negative. Thus Liu et
451
al.'s point of reference is quite different from ours. Taking this into account our results are
452
quite similar with exception of Africa. At pan2tropical level Liu et al. (2017) estimate a
453
difference of flux from land to atmosphere of 2.5 PgC for the period May 2015 2 April 2016
454
compared to Jan 2011 2 Dec 2011. Their specific choice for the 201522016 period is likely
455
motivated by maximum positive anomalies. If we use a similar criteria and thus use the
456
period July 2015 2 June 2016 we find a difference of 2.4 PgC. With regards to tropical Africa,
457
in contrast to Liu et al. (2017), we find a substantial carbon loss from tropical Africa at the
458
same time as the very strong heat peak in the Congo Basin (the beginning of 2016) when a
459
clear CO anomaly also occurred. Our rough biomass burning estimate cannot explain this
460
result on its own 2 thus some down2regulation of tropical forest productivity or enhanced
461
respiration would be needed to explain it. In comparison to Liu et al. (2017) our inverse
462
calculations also attribute less carbon release from Southern Africa during the 201522016 El
463
Niño period.
iew
ev
rR
Fo
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
464
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
!-) !-!-. !-/ !-3 !.5 !.1 472
Fo
473 474
Table 2 Chronology and magnitude of carbon flux anomalies ('Cflx') (sign convention based
475
on a land vegetation perspective, i.e. anomalous carbon loss to the atmosphere has a negative
476
sign while anomalous uptake has a positive sign), climate ('H2O': soil water status, 'T'
477
temperature) and process diagnostics: carbon monoxide ('CO') and solar induced fluorescence
478
('SIF'). Symbols indicate existence of positive (+) and negative anomalies (2) and the number
479
of symbols the strength of the anomalies.
iew
ev
rR
480 481 Tropical S America H2O
T
Cflx
JAS
2
++
OND
222
++
22
JFM
222
+
222
AMJ
22
+
JAS
2
Tropical Africa CO
SIF
H2O
T
Cflx
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 44
Southern Africa CO
SIF
H2O
T
Cflx
CO
SIF
Tropical East Asia H2O
T
Cflx
CO
2015 JFM AMJ
+
222
22
+
22
++
22
2 +
2016
OND
?
2
+++
222
++
22
22
++
22
NA
22
NA
2
482 483
http://mc.manuscriptcentral.com/issue-ptrsb
2
+++
SIF
Page 21 of 44
!/! !/)
6
!/!/.
# &
"
!//
1
%
0
N':
% 1 %
+8::7 '$(
+8::78:$7 '$( ',
E77111 &
1
7
7
8
+ $
489
+ 7 55)/5-71, NE/K01353X/1 and NERC MOYA. MPC, CW, HB, PS acknowledge
490
support from NCEO (NERC National Centre for Earth Observation). LVG acknowledges
491
support from FAPESP and CNPQ grants. WP was partly funded by ERC2CoG grant ASICA
492
(649087).
496
iew
ev
494 495
rR
493
Fo
497
EG, CW and MPC conceived the study, CW did the atmospheric transport inverse
498
calculations with contributions from MPC, CF and EG. PS, RP, HB provided solar
499
fluorescence retrievals from GOSAT, LG, CC, WP, CC led / contributed to the Amazon
500
greenhouse data collection and laboratory analysis, MS contributed to the climate analysis
501
and MND provided CO air column retrievals from MOPITT. EG wrote the manuscript. All
502
co2authors commented on and contributed to the science and writing of the manuscript.
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
503 504 505 506 507 508 509
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
)10
'
511 512
Table S1 Domains used for spatial integration of air column CO anomalies. Region
Latitude
Longitude
Tropical South America
20.5 S,...,19.5 N
80.5 W,..., 20.5W
Tropical Africa
20.5 S,...,19.5 N
Tropical Asia
20.5S,...,19.5 N
20.5 W,...39.5E 49.5 E, ..., 149.5 E
513
Fo
514 515 516
rR
+ %
ev
517 518
Bacastow, R. B. (1976) Modulation of atmospheric carbon dioxide by the Southern
519
Oscillation, Nature, 1162118.
521
Boden, T.A., G. Marland, and R.J. Andres (2017) Global, Regional, and National Fossil2Fuel
522
CO2 Emissions. Carbon Dioxide Information Analysis Center, Oak Ridge National
523
Laboratory,
524
10.3334/CDIAC/00001_V2017.
U.S.
Department
of
Energy,
Oak
ly
On
520
iew
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 44
Ridge,
Tenn.,
U.S.A.
doi
525 526
Bontemps, S., P. Defourny, J. Radoux, E. Van Bogaert, C. Lamarche, F. Achard, P. Mayaux,
527
M. Boettcher, C. Brockmann, and G. Kirches (2013) Consistent global land cover maps for
528
climate modelling communities: current achievements of the ESA’s land cover CCI, in:
529
Proceedings of the ESA Living Planet Symposium, 9–13 September 2013, Edinburgh, UK,
530
9–13.
http://mc.manuscriptcentral.com/issue-ptrsb
Page 23 of 44
) 1 532
Chipperfield, M. P. (2006) New version of the TOMCAT/SLIMCAT offline chemical
533
transport model: Intercomparison of stratospheric tracer experiments, Q. J. Roy. Meteor.
534
Soc., 132, 1179–1203, doi:10.1256/qj.05.51
535 536
Dai A. and T. M. L. Wigley (2000) Global Patterns of ENSO2induced Precipitation,
537
Geophys. Res. Lett., 27(9), 128321286.
538
Fo
539
Deeter, M. N., S. Martínez2Alonso, L. V. Gatti, M. Gloor, J. B. Miller, L. G. Domingues, and
540
C. S. C. Correia (2016) Validation and analysis of MOPITT CO observations of the Amazon
541
Basin, Atmos. Meas. Tech., 9, 3999–4012.
ev
rR
542 543
Deeter, M. N. (2013) MOPITT (Measurements of Pollution in the Troposphere) Version 6
544
Product User's Guide.
546
Fan, Y., and H. van den Dool (2008) A global monthly land surface air temperature analysis
547
for 1948–present, J. Geophys. Res., 113, D01103, doi:10.1029/2007JD008470.
ly
On
545
iew
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
548 549
Feely, R. A., R. Wanninkhof, T. Takahashi & P. Tans (1999) Influence of El Niño on the
550
equatorial Pacific contribution to atmospheric CO2 accumulation, Nature, 398, 5972601.
551 552
Feely, R. A., et al. (2002) Seasonal and interannual variability of CO2 in the equatorial
553
Pacific. Deep2Sea Res., 49, 2443–2469.
554
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
555
$ : $
&
(
W;
" $
* : 0 6
= 517)
556
BAMS state of climate Chapter 3, Global Oceans, Global ocean carbon cycle, Bulletin
557
Americ. Met. Soc., S892S92.
558 559
Fernandes, K., L. Verchot, W. Baethgen, V. Gutierrez2Velez, M. Pinedo2Vasquez, and C.
560
Martius (2017) Heightened fire probability in Indonesia in non2drought conditions: the effect
561
of increasing temperatures, Environ. Res. Lett. 12, 054002.
562
Fo
563
Fischer, E. M., and R. Knutti (2015) Anthropogenic contribution to global occurrence of
564
heavy2precipitation and high2temperature extremes, Nature Climate Change, 5, 5602564,
565
doi:10.1038/nclimate2617.
ev
rR
566 567
Foley, J. A., A. Botta, and M. T. Coe (2002) El Niño–Southern oscillation and the climate,
568
ecosystems and rivers of Amazonia, Glob. Biogeochem. Cycl., 16 (4), 1132,
569
doi:10.1029/2002GB001872.
iew
On
570 571
Frankenberg, C., J. B. Fisher, J. Worden, G. Badgley, S. S. Saatchi, J.2E. Lee, G. C. Toon, A.
572
Butz, M. Jung, A. Kuze, and T. Yokot (2011) New global observations of the terrestrial
573
carbon cycle from GOSAT: Patterns of plant fluorescence with gross primary
574
productivity,
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 44
, 38, L17706, doi:10.1029/2011GL048738.
575 576
Gatti, L. V., M. Gloor, J. B. Miller, C. E. Doughty, Y. Malhi, L. G. Domingues, L.S. Basso,
577
A. Martinewski, C. S. C. Correia, V. F. Borges, S. Freitas, R. Braz, L. O. Anderson, H.
578
Rocha, J. Grace, O. L. Phillips, and J. Lloyd (2014) Drought sensitivity of Amazonian carbon
http://mc.manuscriptcentral.com/issue-ptrsb
Page 25 of 44
).3 580
%
%
+
)5-
.-2/5
E10.1038/nature12957
581 582
Gatti, L. V., J. B. Miller, M.T.S. D’Amelio, E. Gloor, A. Martinewski, L. S. Basso, S. Wofsy,
583
and P. Tans (2010) Vertical profiles of CO2 above eastern Amazonia suggest a near neutral
584
carbon balance between 2000 and 2009, Tellus, 62, 5812594, DOI: 10.1111/j.16002
585
0889.2010.00484.x
586
Fo
587
Gloor M., N. Gruber, J. L. Sarmiento, C. Sabine, R. Feely, and C. Rödenbeck (2003) A first
588
estimate of present and preindustrial air2sea CO2 flux patterns based on ocean interior carbon
589
measurements and models, Geophys. Res. Lett., 30, No. 1 10.1029/2002GL015594.
ev
rR
590
iew
591
Gurney, K. R., R. M. Law, A. S. Denning, P. J. Rayner, D. Baker, P. Bousquet, L. Bruhwiler,
592
Y.2H. Chen, P. Ciais, S. Fan, I. Y. Fung, M. Gloor, M. Heimann, K. Higuchi, J. John, T.
593
Maki, S. Maksyutov, K. Masarie, P. Peylin, M. Prather, B. C. Pak, J. Randerson, J.
594
Sarmiento, S. Taguchi, T. Takahashi, and C.2W. Yuen (2002) Towards more robust estimates
595
of CO2 fluxes: control results from the TransCom3 inversion intercomparison, Nature, 415,
596
6262630.
ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
597 598
Huffman, G.J., R.F. Adler, D.T. Bolvin, G. Gu, E.J. Nelkin, K.P. Bowman, Y. Hong, E.F.
599
Stocker, and B. Wolff (2007) The TRMM Multi2satellite Precipitation Analysis: Quasi2
600
Global, Multi2Year, Combined2Sensor Precipitation Estimates at Fine Scale. J. Hydrometeor.,
601
8, 38255.
602
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
603
*
,
4
' : '
= 551) The Carbon Cycle Response to
604
ENSO: A Coupled Climate–Carbon Cycle Model Study, J. Clim., 411324129.
605 606
Jung, M., M. Reichstein, H. A. Margolis, A. Cescatti, A. D. Richardson, M. Altaf Arain, A.
607
Arneth, C. Bernhofer, D. Bonal, J. Chen, D. Gianelle, N. Gobron, G. Kiely, W. Kutsch, G.
608
Lasslop,1 B. E. Law, A. Lindroth, L. Merbold,L. Montagnani, E. J. Moors, D. Papale, M.
609
Sottocornola, F. Vaccari and C. Williams (2011) Global patterns of land‐atmosphere fluxes
610
of carbon dioxide, latent heat, and sensible heat derived from eddy covariance, satellite, and
611
meteorological observations, J. Geophys. Res., 116, G00J07, doi:10.1029/2010JG001566
Fo
612
rR
613
Jung, M., M. Reichstein, R. Schwalm, and C. Huntingford (2017) Compensatory water
614
effects link yearly global land CO2 sink changes to temperature, Nature, 541, 516 – 520.
ev
615
iew
616
Khatiwala, S., F. Primeau, and T. Hall (2009) Reconstruction of the history of anthropogenic
617
CO2 concentrations in the ocean, Nature, 462, 3462349.
On
618 619
Kasuya, M., and T. Hamazaki (2005) Overview of the Greenhouse Gases Observing Satellite
620
(GOSAT), Proceedings of WSANE 2005 (Workshop for Space, Aeronautical and
621
Navigational Electronics 2005), KARI, Daejeon, Korea, March 325, 2005, 1452149.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 44
622 623
Global
Carbon
Project
(2017)
624
[www.globalcarbonproject.org/carbonbudget]
Carbon
budget
and
trends
2017.
625 626
Liu, J., K. W. Bowman, D. S. Schimel, N. C. Parazoo, Z. Jiang, M. Lee, A. A. Bloom, D.
627
Wunch, C. Frankenberg, Y. Sun, C. W. O’Dell, K. R. Gurney, D. Menemenlis, M. Gierach,
http://mc.manuscriptcentral.com/issue-ptrsb
Page 27 of 44
628
,
:
= 517) Contrasting carbon cycle responses of the tropical
629
continents to the 2015–2016 El Niño. Science, 358, 191
630 631
Meroni, M. Rossini, L. Guanter, L. Alonso, U. Rascher, R. Colombo, and J. Moreno (2009)
632
Remote sensing of solar2induced chlorophyll fluorescence: Review of methods and
633
applications, Rem. Sens. Env., 113, 2037–2051.
634 635
Mitchell, T. D., and P. D. Jones (2005) An improved method of constructing a database of
636
monthly climate observations and associated high2resolution grids, Int. J. Climatology, 25,
637
6932712, Doi: 10.1002/joc.1181.
ev
rR
638
Fo
639
Murray2Tortarolo, G., P. Friedlingstein, S. Sitch, S. I. Seneviratne, I. Fletcher, B. Mueller, P.
640
Greve, A. Anav, Y. Liu, A. Ahlström, C. Huntingford, S. Levis, P. Levy, M. Lomas, B.
641
Poulter, N. Viovy, S. Zaehle, and N. Zeng (2016) The dry season intensity as a key driver of
642
NPP trends, Geophys. Res. Lett., 43, 2632–2639, doi:10.1002/2016GL068240
iew
On
643 644
Olsen, S., and J. Randerson (2004) Differences between surface and column atmospheric
645
CO2 and implications for carbon cycle research, J. Geophys. Res., 109, D02301,
646
doi:10.1029/2003JD003968.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
647 648
Page, S. E., F. Siegert, J. O. Rieley, H.2D. V. Boehm, A. Jayak and S. Limink (2002) The
649
amount of carbon released from peat and forest fires in Indonesia during 1997, Nature, 61265.
650 651
Pan, Y., Birdsey, R. A., Fang, J., Houghton, R., Kauppi, P. E., Kurz, W. A., Phillips, O. L.,
652
Shvidenko, A., Lewis, S. L., Canadell, J. G., Ciais, P., Jackson, R. B., Pacala, S. W.,
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
653 654
: ,
' $
: '
'
#
, : = 511) A large and
persistent carbon sink in the world’s forests, Science, 333, 988–993.
655 656
Scheffers B. R., Joppa L. N., Pimm S. L., and W.F. Laurance (2012) What we know and
657
don't know about Earth's missing biodiversity, Trends Ecol. Evol. 27, 501–10.
658 659
State of the Tropics (2017) Sustainable Infrastructure for the Tropics. James Cook
660
University, Townsville, Australia.
661
Fo
662
Takahashi, T., S. C. Sutherland, R. Wanninkhof, et al. (2009) Climatological mean and
663
decadal change in surface ocean pCO2, and net sea–air CO2 flux over the global oceans.
664
Deep2Sea Res, Pt II 56, 554–77.
iew
665
ev
rR
666
Tarantola A. and B. Valette (1982) Generalized nonlinear inverse problems solved using the
667
least squares criterion, Rev. Geophys. Space Phys., 20, 2192232.
On
668 669
Tapley, B., Flechtner, F., Bettadpur, S., and M. Watkins (2004) The Gravity Recovery and
670
Climate Experiment Mision overview and early results, Geophys. Res. Lett., 31, doi:
671
10.1029/2004GL019920.
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 44
672 673
Van der Tol, C., J. A. Berry, P. K. E. Campbell, and U. Rascher (2014) Models of
674
fluorescence and photosynthesis for interpreting measurements of solar2induced chlorophyll
675
fluorescence, J. Geophys. Res. Biogeosci., 119, 2312–2327, doi:10.1002/2014JG002713.
676 677
Wang, X., S. Piao, P. Ciais, P. Friedlingstein, R. B. Myneni, P. Cox, M. Heimann, J. Miller,
http://mc.manuscriptcentral.com/issue-ptrsb
Page 29 of 44
678
'
0
# M
679
sensitivity
680
doi:10.1038/nature12915.
to
tropical
:
= 514) A two2fold increase of carbon cycle
temperature
variations,
Nature,
506,
2122215,
681 682
Wilson, C., M. P. Chipperfield, M. Gloor, and F. Chevallier (2014) Development of a
683
variational flux inversion system (INVICAT v1.0) using the TOMCAT chemical transport
684
model (2014) Geosci. Model Dev., 7, 2485–2500.
685
Fo
686
Yang, X., J. Tang, J. F. Mustard, J.2E. Lee, M. Rossini, J. Joiner, J. W. Munger, A. Kornfeld,
687
and A. D. Richardson (2015) Solar2induced chlorophyll fluorescence that correlates with
688
canopy photosynthesis on diurnal and seasonal scales in a temperate deciduous forest,
689
Geophys. Res. Lett., 42, 2977–2987, doi:10.1002/ 2015GL063201.
ev
rR
690
iew
691 692
On
693 694
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
695 696 697 698 699 700 701 702
3
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
.5 704
1 Climate trends for tropical and subtropical forest biome based on CRU (Climate
705
Research Unit) TS 3.24 climatology (Mitchell and Jones, 2005). Stippling denotes
706
statistically significant trends.
707 708
Fig. 2 Global atmospheric CO2 growth rate anomalies ∆g estimated using Eq. 2 and El Niño
709
3.4 Index (obtained from http://www.cpc.ncep.noaa.gov).
710
Fo
711
Fig. 3 Tropical land gravity anomaly anomalies measured by the GRACE satellites and
712
global CO2 growth rate anomalies ∆g estimated using Eq. 2.
713
ev
rR
714
Fig. 4 Land gravity anomaly anomalies and monthly means of daily minimum and maximum
715
temperatures, respectively, for tropical land regions. Temperature data are from Climate
716
Prediction Center (CPC), Global Land Surface Air Temperature Analysis (Fan et al. 2008).
718
Fig. 5 Tropical South America to atmosphere CO2 flux anomalies estimated with inverse
719
modelling of atmospheric transport and atmospheric CO2 concentration observations, and
720
tropical South American gravity anomaly anomalies estimated by GRACE satellite mission.
ly
On
717
iew
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
721 722
Fig. 6 Time2series of land2to2atmosphere carbon flux estimates (low2pass filtered) for tropical
723
land regions. The portion for which Amazon vertical profile data have been available and
724
included in the atmospheric transport inversion calculations is coloured in dark black.
725
Stippled red are estimates which do not include tropical South American data.
726
http://mc.manuscriptcentral.com/issue-ptrsb
Page 30 of 44
Page 31 of 44
727 728
.0
%
4
515/16 of total air column carbon
monoxide measured from space (MOPITT, Deeter et al. 2013).
729 730
Fig. 8 Solar induced fluorescence anomalies (measured from GOSAT satellite (Kasuya and
731
Hamazaki 2005) and based on retrievals at 772 nm).
732 733
Fig. S1 Locations where CO2 mixing ratio data are being measured which are used to
734
estimate CO2 surface fluxes using atmospheric transport inversions.
Fo
735 736
Fig. S2 Availability of atmospheric CO2 data used for atmospheric transport inversions.
rR
737 738
ev
Fig. S3 Gravity anomaly anomalies as measured by GRACE satellites for 2015 and 2016.
739
iew
740
Fig. S4 Precipitation anomalies calculated from TRMM (Tropical Rainfall Measuring
741
Mission) (Huffman et al. 2007) version 7 and monthly daily maximum and minimum
742
temperature from NOAA Climate Prediction Center (CPC) climatology (Fan et al. 2008).
On
743
ly
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
744
Fig. S5 Example (October 1, 2015) of CO retrieval weighting kernels for total column CO
745
estimation using MOPITT data for 32by23 degree latitude by longitude regions in Amazonia,
746
Africa, Indonesia, and the Indian Ocean respectively. The weighting kernels are normalized
747
such that they represent the response of the retrieved total column to a perturbation in CO
748
partial column applied at each level in the retrieved profile. The ideal value would be 1 at all
749
levels.
http://mc.manuscriptcentral.com/issue-ptrsb
Submittedtrend to Phil.1981 Trans. R. B - Issue Mean annual temperature − Soc. 2016 ( oC/ decade)
.... .. .. .. .. .... ..... .. . .... ............. .. . 1 ....... ........... ............ ............. 2 ................ . . ................... ................. .............. .. 3 .............. .............. ... .... .. ..... .. ... 4 .. ... .. ...... .. ........ . ....... 5
Fo
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
.... ........ . ... .. .. ..... ..... ...... . ..... .. ...... ........... . ............ .. .... ...... ..... .... ...... .. ....... ...... . .. ..... ... .. . ... ..... ... . ... . ... .. .. ..
rR −0.3
evi −0.1
ew 0.1
Page 32 of 44
....... .............. ...... ...... ...... . .. . ... ... ..... .... ......... .. .... .... .. ... ... .... . . .. ...... ... ...... ... .. ... . .. ..... . ... ..... ...... ..... ..... ... .. ... ..... ...... ....... ..... ..... .. ..... . .. ..... ... ..... . .. . 0.3
>0.5
On
Mean annual precipitation trend 1981 − 2016 .(mm/ decade) ..
.
. .. ... .. .... .. .. . .. . .. ..... ... ... ... .. .. . .. . .. .. . ..
..... ...
. .. .. .. . .... . ... ... ... . ... . ... . .. . ... ... ... ..
ly
.... . .. .... . . .. .. .. . ... .. .. ... .. ... ...
http://mc.manuscriptcentral.com/issue-ptrsb −275 −175
−75
25
125
225
..... .. ... .. . ..
Rev
iew
−2
−1
0
1
For
On
ly
Atm. carbon growth rate − AF*(Fossil Fuel emissions) ElNino 3.4 Index (scaled)
−3
(PgC yr−1)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Submitted to Phil. Trans. R. Soc. B - Issue
2
Page 33 of 44
1960
1970
1980
1990
Year
http://mc.manuscriptcentral.com/issue-ptrsb
2000
2010
Submitted to Phil. Trans. R. Soc. B - Issue
Page 34 of 44
2 −2 −1 0 1 (PgC yr−1)
Fo
rR
2005
ev
2010
2015
ie
w
On
Tropical South America Anomalies 2
ly
−2 −1 0 1 (PgC yr−1)
Gravity anom anomaly (cm) −10 −5 0 5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
Gravity anom anomaly (cm) −4 −2 0 1 2
Pantropical Anomalies
Grav Anom Anom AGR − AF*FF, scaled 2005
2010 Year http://mc.manuscriptcentral.com/issue-ptrsb
2015
Submitted to Phil. Trans. R. Soc. B - Issue
Tropical S America 24
Min Day T (C)
23
32
22
31
21
30
20
29 28
−10
0
5 10
Max Day T (C)
33
Tropical South America (20S − 10N)
2010
2015
20
28
21
29
22
30
Min Day T (C)
23
Central Africa (5S,5N)
19
27
Max Day T (C)
1990
On
2005
2010
2015
2010
2015
2010
2015
20
Min Day T (C)
Southern Africa
1995
15
2000
2005
16
26
18
28
20
30
Min Day T (C)
22
32
Tropical East Asia
24
Max Day T (C)
2000
ly
1990
Tropical South East Asia (10S − 10N, 90W − 150W)
1995
10
5 10 0
30
iew Max Day T (C)
Southern Africa (40S − 10S)
2005
26
ev
25
−10
0
rR
2000
20
5 10
Fo
1995
31
North and Central Africa (10S − 30N)
−10
Grav anom anom (cm)
Grav anom anom (cm)
1990
Grav anom anom (cm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Grav anom anom (cm)
Page 35 of 44
http://mc.manuscriptcentral.com/issue-ptrsb 2002
2004
2006
2008
2010
2012
2014
2016
2018
1990
1995
2000
2005
Submitted to Phil. Trans. R. Soc. B - Issue
Page 36 of 44
5
ly
1997
1999
2001
2003
−15
Land to Atmosphere C flux Gravity Anomaly Anomaly 1995
(cm)
0
2 −1 −2
On
−5
iew
−10
Rev
0
1
For
−3
(PgC yr−1)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
10
Tropical South America
2005
2007
2009
http://mc.manuscriptcentral.com/issue-ptrsb
2011
2013
2015
2017
Tropical Land
2 0 1 −1 −3 1 −3
−1
iew ly On
1
Southern Africa
−1
(PgC yr−1)
−3
Tropical Africa
ev
(PgC yr−1)
rR
Tropical South−East Asia
(PgC yr−1)
(PgC yr−1)
Tropical South America
Fo
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Submitted to Phil. Trans. R. Soc. B - Issue
−3
(PgC yr−1)
Page 37 of 44
http://mc.manuscriptcentral.com/issue-ptrsb 1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
view
nuscriptcentral.co toPage Phil. 38 Trans. of 44 R. S
0
0
50
−10
−10
0
−100
−30
−30
−50
10
20
JAS
250
300
350
100
Fo
150
200
250
300
ie
w
250
300
100
350
150
200
100
20 10 0 −10
Longitude
250
300
350
250
300
350
ly
http://mc.manuscriptcentral.com/issue-ptrsb 200
300
On
150
200 Longitude
150
250
AMJ
JAS
100
350
OND
−30
−10
0
10
20 10 0 −10
200
300
0
350
−30
150
250
−10
ev
JFM 2016 anomalies
100
200
−30
−30
−10
0
rR
100
150
20
200
10
150
20
100
−30
Latitude Latitude
Latitude
Latitude
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
AMJ
20
(%)
Submitted to100 Phil. Trans. R. Soc. B - Issue 10
20
JFM 2015 anomalies
10
Page 39 of 44
350
Submitted to Phil. Trans. R. Soc. B - Issue
iew
ev
rR
Fo ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
http://mc.manuscriptcentral.com/issue-ptrsb
Page 40 of 44
Page 41 of 44
abp alt aoc asc ask azr bal bhd bkt bme bmw brw bsc cba cgo chr cib cmo cpt crz drp dsi eic gmi goz hba hpb hsu hun ice izo kco key kum kzd kzm lef llb lln lmp mbc mex mhd mid mkn mlo nat nmb nwr oxk pal pao poc psa pta rpb scs sdz sey sgp shm smo spo stm sum syo tap thd tik ush uta uum wis wlg wpc zep alf rba tef tab
iew
ev
rR
Fo ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Submitted to Phil. Trans. R. Soc. B - Issue
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
number of observations per month 0
1
2
3
4
5
6
7
8
http://mc.manuscriptcentral.com/issue-ptrsb
9
10
11
2016
Submitted to Phil. Trans. R. Soc. B - Issue
iew
ev
rR
Fo ly
On
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
http://mc.manuscriptcentral.com/issue-ptrsb
Page 42 of 44
Page 43 of 44
Submitted to Phil. Trans. R. Soc. B - Issue
2016
2015 JFM
cm
JFM
cm
10
0
0
40 20
0
10
20
0
−20 −40 0
50
100
150
200
250
300
−20
−20
−20
Latitude
40
350
−40 0
50
100
150
250
300
350
200
250
300
350
200
250
300
350
200
250
300
350
10
10
100
150
200
250
−20
0
300
0
350
ev
−20 0
50
100
150
200
250
300
350
iew 0
150
On 50
100
150
OND
10 0
−20
−20
0
100
ly
OND 10
50
JAS
−20
0
10
JAS
rR
10
50
Fo
0
0 −20
Latitude
0
Latitude
200
AMJ
AMJ
Latitude
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
0
50
100
150
200
250
300
350
0
50
100
150
Longitude
Longitude
http://mc.manuscriptcentral.com/issue-ptrsb
Submitted to Phil. Trans. R. Soc. B - Issue
30.0
31.0
31.0
32.0
32.0
S E Asia S E Asia
30.0
2000
2005
2010
29.0
Fo
rR 2015
1990
ev
iew
19.5
20.5
15 16 17 18 19 20
29.0
1995
21.5
1990
1990
1995
2000
2005
2010
2015
1995
2000
2005
2010
2015
1995
2000
2005
2010
2015
1995
2000
2005
2010
2015
On
ly
0 −10
−5
−2
cm/mo
0
2
5
4
6
1990
−6
Day max. Temp Day(C) Max Temp (C)
Tropical S America Tropical S America
Precip. Anomaly Day min. Temp cm/mo (cm/mo) (C)Temp (C) Day Min
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Page 44 of 44
1990
1995
2000 Year
2005
2010
2015
1990
http://mc.manuscriptcentral.com/issue-ptrsb
Year