The CEOS carbon strategy was published in 2014 (1), in response to a call from GEO to address:
‘the dearth of global observations [that are a key reason for our lack of understanding of the global carbon cycle] and the development of an increased, improved and coordinated observing system for observing the carbon cycle’
It provided 42 recommendations for action across all the Space Agencies specific to the atmosphere, land and ocean domains, integration across those domains and a series of generic coordination needs. A typical example is:
‘Carbon-Action-36: CEOS will strengthen linkages with relevant carbon communities and organizations to facilitate the communications and coordination necessary to ensure that the satellite data products provided by CEOS Agencies meet needs and are utilized appropriately.’
These actions were reviewed, converted into specific activities by CEOS, and allocated to individual component parts of CEOS (Working Groups and Virtual Constellations) with the objective to review progress every 3-5 years.
As part of the Space Agency response, ESA organised the 3rd Carbon from Space meeting in 2016 which brought together the EO science and carbon cycle communities to discuss their response and make recommendations for activities to be undertaken in the form of potential specific activities across ocean, land and atmosphere domains. This paper reports on the recommendations (2) and progress in addressing them by highlighting individual projects funded through ESA Programmes and planned future activities.
Examples of ongoing and planned projects presented will include the RECCAP-2 pilot study dealing with comparison of methods for determination of the regional terrestrial carbon budgets and the preparation for development of the RECCAP-2 initiative, the combination of satellite products for permafrost assessment (GlobPermafrost, Permafrost_cci), efforts to estimate forest biomass (GlobBiomass, Biomass_cci) in anticipation of the launch of the BIOMASS Earth Explorer and other key missions (GEDI, IceSat-2, NISAR), the importance of disturbance (Fire_cci, S1-Amazonas), exploration of the methane cycle (S5P+ Methane, METHEO) and investigation of approaches for assessing ocean acidification (OceanSODA) and understanding the ocean's role in uptake and redistribution of natural and anthropogenic carbon (Carbon+ Ocean).
In addition, the paper will describe the ESA Terrestrial Carbon Constellation Initiative, an ESA effort to prepare for the joint exploitation of BIOMASS, FLEX, SMOS and the Copernicus Sentinel missions to address the terrestrial carbon cycle with focus on the mission synergies and complementarity with non-ESA satellites (e.g., NISAR, GEDI, ALOS, SAOCOM). Finally, the status of the relevant missions, including BIOMASS, FLEX and the Copernicus candidate Sentinel CO2 mission will be reported.


1. http://ceos.org/home-2/the-ceos-carbon-strategy-space-satellites/
2. http://www.carbonfromspace.info

Differences among Earth System Model forecasts of the carbon cycle are currently significant, and lead to major uncertainties in predictions of the land sink. These differences have been linked to variations in the internal processing of carbon, particularly in the large pools in biomass and soil organic matter. We show here how new remote sensing products, and novel model-data fusion approaches, advance fundamental knowledge of forest ecology and biogeochemistry, and thereby constrain ESMs. Biomass mapping linked to estimates of net primary production provides a constraint on the turnover time of the biomass pool, a critical model parameter. Using the CARDAMOM model-data fusion system, we show how biomass mapping supports improved parameterisation of carbon cycling for ESMs. We identify regional variations in C turnover currently missing from tropical plant functional types. A sensitivity analysis performed using the CARDAMOM system shows a tropical mean 49% reduction of uncertainty of vegetation C turnover time achieved when assimilating the current state-of-the-art pantropical biomass map. This error reduction has clear spatial variability across latitude and between continents. Further, we use CARDAMOM to show how multiple repeated biomass measurements, for instance from ESA’s Biomass Mission, reduce bias on estimates of C increment in aggrading ecosystems, and also on internal process parameters like carbon use efficiency. Finally, we show how CARDAMOM analyses can be used to reliably generate an ensemble average of ESM land sink forecasts.

We have studied the growth, disturbance and carbon (C) storage of boreal forests to improve understanding of Arctic/Boreal terrestrial ecosystems that may be approaching a potential tipping point of C release. At continental scales, climate change is altering vegetation productivity, dynamics and C sequestration. These shifts are reflected in vegetation structure (cover and height) which across the landscape and we seek to understand environmental constraints to predict impacts of environmental change on vegetation cover and C-stock/flux. Dynamic Global Vegetation Models (DGVMs) thus far have found large increases in productivity and C-flux. However, turnover rates in these models have a large source of divergence between them. Site Index (SI) is a parameter widely used in forestry to describe the potential height-growth of trees in a particular location or 'site.' SI knowledge will reduce uncertainty of live C turnover into soil C pools. We have successfully estimated boreal forest SI by pairing Landsat estimates of forest disturbance through the MSS record with GLAS and WorldView sub-meter resolution stereo imagery. In this talk, explicit patterns of SI will be presented along with estimated rates of carbon accumulation. Boreal forest management questions will also be discussed in light of these new approaches.

Major uncertainties persist in the current understanding of the global methane (CH4) budget: in particular, the primary sources of ambiguity in attributing the atmospheric CH4 growth rate to underlying source and sink processes include the confounding roles of atmospheric transport, prior budget and biogeochemical process assumptions, sector-specific isotopic CH4 fraction (δ13CH4) uncertainties and atmospheric chemistry errors. In an effort to comprehensively account for the interaction between the various uncertainty terms, we simultaneously optimize continental-scale sector-specific emissions in a Bayesian formulation of the global CH4 source-sink attribution problem; constraints on the global CH4 budget include satellite measurements of CH4 and CO, surface-based measurements of the total and isotopic CH4 budgets, ancillary constraints on the variability of atmospheric CH4 lifetime, and comprehensive prior uncertainty characteristics on the magnitude and isotopic fraction for each principal CH4 budget term. Based on annual sector-specific CH4 emissions estimates and their associated uncertainties, we specifically examine the emergent co-variability of natural CH4 fluxes and climate variability, and diagnose the role of anthropogenic and natural CH4 emissions in the recent evolution of the atmospheric CH4 budget. Finally, we identify the missing constraints required to improve statistical confidence in hypothesized CH4 budget trends, we quantify the information content of each assimilated observational constraint and we investigate the potential role of recent multi-satellite lower-tropospheric CH4 retrievals on the global CH4 budget.

Space-borne microwave radiometer observations of nearly 40 years (since 1979) are applied to investigate the changes in boreal carbon cycle, in particular the trend of carbon uptake and its significance to the global climate system. The developed methodology combines radiometer data with ecosystem level hemispheric flux observations in order to assess the critical moments of annual carbon cycle, such as the start of the growing season at the evergreen coniferous boreal forests. Additionally, tower-based microwave radiometer observations from the Sodankylä site, northern Finland are used for the detailed analysis of the relation between microwave observations and boreal carbon exchange processes - related to the characteristics and melt of seasonal snow as well as freezing and thawing of soil. The results show that the snow clearance from the landscape is strongly correlated to the start of the carbon uptake, i.e. the increase of gross primary production due to the photosynthesis. Thus, the snow melt information derived from satellite observations can be used to derive a proxy for the start of the growing season, and thereby, to assess the trend of carbon uptake and its effect to the annual carbon balance.

The results obtained indicate that the trend of the advance in spring recovery of carbon uptake of northern hemisphere boreal forests is 2.3 days per decade. Through the comparison with flux observation data this can be estimated to result to a 3.7% decadal increase in gross primary production during the first six months of the year. In addition to to the CO2 exchange at boreal forests, the studies are concerned with the CH4 processes at the high latitude wetlands. In that case the role of soil freezing during autumn is highly important. In investigating CH4 processes we focus on the use of L-band radiometer data including tower-based Elbara-II observations from the Sodankylä site accompanied with the space-borne ESA SMOS data. Multi-frequency SMMR and SSM/I observations are applied in the global assessment of seasonal snow cover characteristics for the conducted trend analyses (the GlobSnow data record).

First results of the investigations have been published in the paper: Pulliainen et al. (2017), Early snowmelt significantly enhances boreal springtime carbon uptake. PNAS, 114(42): 11081-11086.

SMOS-IC L-VOD: A NEW INDEX TO MONITOR YEARLY CHANGES IN ABOVEGROUND CARBON STOCKS - AN OVERVIEW OF RECENT RESULTS

Jean-Pierre Wigneron(1), Martin Brandt(2), Lei Fan(1), Philippe Ciais(3), Amen Al-Yaari(1), Ana Bastos(4), Jérome Chave(5), Jean-Christophe Domec(1), Frédéric Frappart(1), Koen Hufkens(1), Xiaojun Li(1), Arnaud Mialon(6), Jérome Ogée(1), Josep Penuelas(7), Jennifer Swenson(8,1), Torbern Tagesson(2), Feng Tian(2,9), Compton Tucker(10), Lisa Wingate(1), Rasmus Fensholt(2)

(1) INRA, UR1391 ISPA, Centre INRA Bordeaux Aquitaine, France
(2) Dep. of Geosc. & Natural Resource Management, University of Copenhagen, Denmark.
(3) LSCE, CEA-CNRS-UVSQ, UMR8212, Gif-sur-Yvette 91191, France
(4) Dep. of Geography, Ludwig Maximilians University Munich, Munich D-80333, Germany
(5) UMR 5174 EDB, Université Paul Sabatier, CNRS, Toulouse, France
(6) CESBIO, Université de Toulouse, CNES/CNRS/IRD/UPS, 31400 Toulouse, France
(7) CSIC, Global Ecology Unit CREAF-CSIC-UAB, Bellaterra, Spain
(8) Duke University, 90328 Durham, NC 27708, USA
(9) Dep. of Physical Geography and Ecosystem Science, Lund University, Sweden

(10) NASA, Goddard Space Flight Center, Greenbelt, MD, USA

Abstract
The global terrestrial CO2 sink exhibits strong year-to-year variations that dominate the inter-annual variability (IAV) in the atmospheric CO2 growth rate [1]. This land carbon sink has been mainly attributed to tropical and boreal forests but semi-arid ecosystems appear to play a key role in its inter-annual variability [2]. However, results based on land surface models, atmospheric inversions, in situ or remotely sensed observations do not always converge. New remote-sensing tools that monitor routinely year-to-year changes in aboveground vegetation carbon stocks (AGC) at the global scale would be key to quantify and understand the drivers of inter-annual variability in the terrestrial CO2 sink.
Here, we present a new remote-sensing product (referred to as the SMOS-IC L-VOD) that has strong capabilities for monitoring annual and seasonal changes in aboveground vegetation carbon stocks. The VOD (Vegetation Optical Depth) parameter is associated to the transmissivity of microwave radiation within the vegetation layer and is retrieved from passive or active microwave space-borne sensors. L-VOD (i.e. L-band VOD) presented in this review is retrieved from the first ever space-borne passive microwave sensor at L-Band onboard the ESA SMOS satellite, launched at the end of 2009. Even though SMOS initially focused on Soil Moisture (SM) mapping, the dual-polarization and multi-angular capabilities of SMOS can allow accurate simultaneous retrievals of both VOD and SM [3]. VOD is an attractive alternative tool to monitor changes in vegetation carbon, because microwaves at frequencies lower than 15 GHz are almost insensitive to atmospheric and cloud effects. As opposed to higher-frequency (i.e. 5-15 GHz) VOD products that saturate rapidly over dense forested areas, the low frequency (1.4GHz) of L-VOD senses structural components deeper in the canopy with less influence of green, non-woody plant components [3]. L-VOD is thus less sensitive to canopy saturation effects marking a major step forward in the monitoring of aboveground vegetation carbon stocks [4].
Here, we used the SMOS L-VOD product in the SMOS-IC version. SMOS-IC is a recent product based on the original SMOS retrieval algorithm (Wigneron et al., 1995, 2000, 2007) and is independent of auxiliary data (modelled SM, optical indices), a key feature for developing robust applications. In recent years, SMOS-IC SM was found to be highly performant in soil moisture inter-comparisons [5]. In parallel, SMOS-IC L-VOD has been used in several studies that aim to monitor changes in aboveground carbon stocks (AGC) and vegetation water content at continental (Africa, China, Tropics, etc.) and global scales over the period 2010-2017. In this communication, we will review recent promising applications of this product, including the evaluation of 1) the impact of climate on AGC dynamics over the tropics [6] and the USA, 2) the role of semi-arid, boreal and tropical ecosystems on the inter-annual variability of the terrestrial carbon sink and 3) the buffering effect of plant water storage on the seasonal dynamics of water supply in dry tropical woodlands [7]. Collectively, this review demonstrates the strong potential of the L-VOD product to serve as an important tool for policy makers and stakeholders in the context of on-going climate change.

References

[1] Le Quéré CL et al. Global carbon budget 2017. Earth Syst. Sci. Data 10, 405–448. (doi:10.5194/essd-10-405-2018) (2018)
[2] Ahlström, A. et al. The dominant role of semi-arid ecosystems in the trend and variability of the land CO2 sink. Science 348, 895-899, doi:10.1126/science.aaa1668 (2015)
[3] Wigneron, J.-P. et al., Modelling the passive microwave signature from land surfaces: a review of recent results and application to the L-band SMOS & SMAP soil moisture retrieval algorithms. Remote Sens. Environ. 192, 238–262. (doi:10.1016/j.rse. 2017.01.024) (2017)
[4] Brandt, M. et al. Satellite passive microwaves reveal recent climate-induced carbon losses in African drylands. Nature Ecology & Evolution, 2, 827–835. (doi:10.1038/s41559-018-0530-6) (2018)
[5] Colliander, A. et al. Intercomparison of SMOS and SMAP Soil Moisture Products over Core Validation Sites, this issue
[6] Bastos, A. et al., Impact of the 2015-16 El Nino on the terrestrial carbon cycle constrained by bottom-up and top-down approaches. Phil. Trans. R. Soc. B, 20170304, http://dx.doi.org/10.1098/rstb.2017.0304, (2018)
[7] Tian, F. et al., Coupling of ecosystem-scale plant water storage and leaf phenology observed by satellite. Nature Ecology & Evolution, Vol 2, Spt. 2018, 1428–1435, https://doi.org/10.1038/s41559-018-0630-3 (2018.)