NASA Technical Reports Server (NTRS) 20170012506: An 8-Year

An 8-year, high-resolution reanalysis of atmospheric CO2 mixing ratios based on GMAO

OCO-2 and GOSAT-ACOS retrievals B. Weirt:2, A. Chatterjee’2, L.E. Ott2, and S. Pawson? ee ee

1) USRA/GESTAR, 2) NASA GSFC/GMAO A33G-2451

The NASA GMAO reanalysis blends OCO-2 and — aerate ails
GOSAT-ACOS retrievals (top) with GEOS model i
predictions (bottom) to estimate the full 3D state of _* _
COz every 3 hours (middle). = :
— ee Se Sates This poster describes monthly atmospheric 7
D2 eens growth rates derived from the reanalysis and an a zo
ee idee ee SS application to aircraft data with the potential to Fig 1. Global average mixing ratios from Fig 2. Global total fluxes computed as the
Ee a —s . (blue) GEOS with no assimilation and change in monthly means from Fig 1. Note
—— aid bias correction. (green) reanalysis of GOSAT-ACOS data. the change in Jan. thru Mar.
GT are ob 6/877 w 2048/72 Atmospheric growth rate

since OCO-2 and GOSAT-ACOS observe the column integral
of CO2 and have a 16-day repeat cycle, they have the
potential to better constrain atmospheric growth rates on sub-
annual time scales than data restricted to the PBL. Here, we
compute monthly growth rates and show that the satellites
suggest the model has too much outgassing in boreal
oe winter.

Latitude (°N)

Fig 1. OCO-2 overpass and ACT flight path Fig 2. Measured CO2 from ACT aircraft.
for 27 July 2016. Global average mixing ratios Global total fluxes

GEOS
OCO-2
ACOS

|

—_-

=
x
_—

c)
se)

=
=
a=
<x

°o
“\/\
-
e
=
°o oO

Ww Ww Ww
© © wo wo
u a ~ oe ©
Se pam :
A yA

Aircraft-based analysis

systematic model and satellite retrieval errors with coherent
™ regional and seasonal patterns limit our ability to infer surface
fluxes from satellite retrievals.

ie

i

——
18 Dec 2014 15:00 |

CO, (dry-air ppm)

oO
oS
Total flux (Pg C/yr)

= : Wy Here, we assimilate aircraft data from AC T-America into the
oe ge se a GEOS model and compare the results to an OCO-2 overpass.

a> 5 Fit: a Ni pai ¥ Be The assimilation indicates that the model PBL was too Fig 3. Same as Fig 1, but now with (red) Fig 4. Same as Fig 2, but now with (red)
7 a " ay. i reanalysis of OCO-2 data. reanalysis of OCO-2 data.

395 -25
2014/07 2014/10 2015/01 2015/04 2015/07 2015/10 2016/01 2016/04 2014/07 2014/10 2015/01 2015/04 2015/07 2015/10 2016/01 2016/04

Ay. ae ao high. Overall, the mean difference with OCO-2 v7b is
_ Ne . zs reduced from 0.59 ppm to 0.28 ppm.

NOBM ocean exchange
CASA-GFED biospher

QFED biomass burning ; 10 ee 10 aa) Backscatter and ABL Attribution Jul 27, 2016 405 ee
; _ 2.7
9 9 g4 404
8 8 Ze | 403
VIIRS nighttime ligh - - eT 402
. : E . oF 5 401
7 5 7 5 . 400
24 = 4 = og 399
3 3 <4 < 398
2 Z z 397
P ———— of 395
40 40 36 38 40 42 44 46
Latitude (°N) Latitude (°N) Latitude ( °N)
Fig 3. GEOS forecasted values before Fig 4. Curtain for aircraft data assimilated Fig 5. PBL height determined from Cloud Fig 6. Comparisons to column CO2 (XCOz)
assimilation. into the GEOS model. Note the decrease in Physics Lidar measurements. These heights from OCO-2 v7b. The analysis decreases
Acknowledgements: the OCO-2 project at JPL, CalTech, the ACT-America campaign, NASA GSFC PBL height over 39° to 42°N from Fig 3. are closer to those from Fig 4. the bias from 0.59 ppm to 0.28 ppm.

SVS, the NASA CMS project, Sandip Pal, Chris O’Dell, Emily Bell, John Miller, & everyone | forgot

National Aeronautics and
E-mail: brad.weir@nasa.gov | Web: gmao.gsfc.nasa.gov Space Administration