Towards a Long-Term Record of the Northern Hemisphere Growing Season Net Flux

Quantifying large-scale land-atmosphere fluxes, such as the Northern Hemisphere Growing Season Net Flux (GSNF), is necessary to understanding terrestrial carbon uptake and its response to anthropogenic climate change. One approach to estimating the GSNF involves using observations of atmospheric carbon dioxide (CO2) taken on aircraft campaigns that attain vertical profiles of atmospheric CO2. However, these campaigns are infrequent, making detection of trends in the GSNF a challenge. The goal of this study is to construct a time series of the Northern Hemisphere GSNF using observations of atmospheric CO2concentrations from the Orbiting Carbon Observatory - 2 (OCO-2) satellite. We apply the same method, with the exception of finding the partial column average, used previously on aircraft observations to the column-average CO2from OCO-2 to estimate the GSNF in the Northern extratropics. We do this for each year in the OCO-2 time-period to find a time series of GSNF. Preliminary results indicate that the GSNF has increased by 0.01 PgC yr-1 over the last decade. Continued study will be needed to confidently establish a trend. Results also show that March climate and summer wildfire activity are major drivers of interannual variability in the GSNF. Our research shows that the flux inferred from these satellite observations provide a constraint on the GSNF that allows continuous monitoring over time, providing a target for Earth system models and improving our understanding of carbon-climate feedbacks. •OCO– 2 Column Average CO2 Dry Air Mole Fractions (Level 2 Lite product) [2] •Global Fire Emissions Database 5 monthly emissions estimates (GFED) [3] •Multivariate ENSO index version 2 (MEI) [4] •Terrestrial water storage anomaly reconstructions using Bidirectional Long Short-Term Memory (GRAiCE BiLSTM TWSA) [5] •Palmer Drought Severity Index (PDSI), soil moisture, precipitation, maximum air temperature, and minimum air temperature from TerraClimate [6] •3 inverse models (MIROC4, CAMS, and CT2022) used for the calibration process 1. Loechli, M., et al. (2023). Evaluating northern hemisphere growing season net carbon flux in climate models using aircraft observations. Global Biogeochemical Cycles, 37, e2022GB007520. https://doi.org/10.1029/2022GB007520 2. O'Dell, C. W., et al. (2018). Improved retrievals of carbon dioxide from Orbiting Carbon Observatory-2 with the version 8 ACOS algorithm, Atmos. Meas. Tech., 11, 6539–6576, https://doi.org/10.5194/amt-11-6539-2018 3. van der Werf, G. R., et al. (2017). Global fire emissions estimates during 1997–2016, Earth Syst. Sci. Data, 9, 697–720, https://doi.org/10.5194/essd-9-697-2017 4. Multivariate ENSO Index Version 2 retrieved from https://psl.noaa.gov/enso/mei/ 5. Palazzoli, I., Ceola, S. & Gentine, P. (2025). GRAiCE: reconstructing terrestrial water storage anomalies with recurrent neural networks. Sci Data 12, 146 https://doi.org/10.1038/s41597-025-04403-3 6. Abatzoglou, J. T., S. Z. Dobrowski, S. A. Parks, and K. C. Hegewisch, 2018: TerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958– 2015. Scientific Data, 5, https://doi.org/10.1038/sdata.2017.191. 7. H. D. Graven et al. (2013) Enhanced Seasonal Exchange of CO2 by Northern •Total burned area during the growing season, March/April MEI, and minimum March temperature best explain the interannual variability in GSNF, indicating that summer wildfires and March climate may be the dominate drivers of the observed variability in the GSNF. •Water availability does not seem to be a driver of observed variability in the GSNF. •The estimate from OCO-2 agrees with Loechli et al. 2023 within the error bounds but is slightly lower (5.5 PgC yr-1 vs. 5.7 PgC yr-1), likely due to the inclusion of the stratosphere. •The 0.01 ± 0.03 PgC yr-1 increasing trend is likely less than the 0.03 PgC yr-1 – 0.05 PgC yr-1 trend expected [7] based on the observed increase in the northern hemisphere seasonal cycle amplitude, potentially indicating a slowdown in the growth rate of the northern hemisphere land sink. •This work is limited by the number of years of observations available from OCO-2. A longer record would allow for verification of the apparent trend. ABSTRACT DATA AND METHODS CONCLUSIONS REFERENCES March climate and summer wildfires are major drivers of year-to-year variation in Northern Hemisphere carbon exchange during the growing season. Towards a Long-Term Record of the Northern Hemisphere Growing Season Net Flux Using Satellite Observations Morgan Loechli1, Naveen Chandra2, Gretchen Keppel-Aleks3, Matthäus Kiel4 1 K a l a m a z o o C o l l e g e , 2 J a p a n A g e n c y f o r M a r i n e - E a r t h S c i e n c e a n d T e c h n o l o g y , 3 U n i v e r s i t y o f M i c h i g a n , 4 J e t P r o p u l s i o n L a b o r a t o ry co nt a c t : mo rg a n . L o e c hli@kzo o.edu Carbon Seasonal Cycle Finite Difference Calibrate The r2 value vs. p-value is plotted for each correlation with GSNF. Inset: r2 values for correlations of all statistically significant correlations with GSNF and each other. a) Flux plotted for the warmest March (2016) and the coldest March (2018). The warmer year is shallower. b) Flux plotted for the year with the largest burned area (2015) and the year with the least (2022). The year with more burned area is narrower. c) GSNF for years plotted in a) and b) with the average for 2015-2024. The colder March (2018) and less burned area (2022) years have larger GSNF the typical. GFED growing season total burned area vs. OCO-2 GSNF. As burned area increases, GSNF decreases as expected. Minimum March temperature vs. OCO-2 GSNF. As minimum temperature increases, GSNF decreases. a) OCO-2 GSNF over time. b) OCO-2 GSNF compared to predicted GSNF using multiple linear regression model of minimum March temperature and total growing season burned area.