The Open-source Data Inventory for Anthropogenic CO
Carbon dioxide (CO
Similarly, FFCO2 estimates serve as a reference in atmospheric CO
Global FFCO2 data are available in a gridded form from different
institutions and research groups (e.g., CDIAC/ORNL and Europe's Joint
Research Centre, JRC) and those gridded emissions data are often based on
disaggregation of national (or sectoral) emissions (e.g., Andres et al.,
1996; Rayner et al., 2010; Oda and Maksyutov, 2011; Janssens-Maenhout et al.,
2012; Kurokawa et al., 2013; Asefi-Najafabady et al., 2014). The emissions
spatial distributions are often estimated using spatial proxy data that
approximate the location and intensity of human activities (hence, CO
Satellite-observed nighttime light data have been identified as an excellent
spatial indicator for human settlements and intensities of some specific
human activities (e.g., Elvidge et al., 1999, 2009) and have been used to
infer the associated CO
In response to increasing needs from the CO
Figure 1 illustrates our current ODIAC emissions modeling framework (we defined it as “ODIAC 3.0 model”, in contrast to the original version). Major changes and differences from Oda and Maksyutov (2011, ODIAC v1.7) are (1) the use of emissions estimates made by the CDIAC/ORNL (rather than our own emissions estimates), (2) the use of multiple spatial emissions proxies in order to distribute CDIAC national emissions estimates made by fuel type, and (3) the inclusion of emissions temporal variations (version 1.7 only indicates annual emissions fields). Given that CDIAC emissions estimates have been well-respected and widely used in the carbon research community (e.g., Ballantyne et al., 2012; Le Quéré et al., 2016), our mission in our emissions data development is to develop and deliver an extended, comprehensive global gridded emissions data product, fully utilizing CDIAC emissions data (e.g., emissions estimates in both tabular and gridded forms). We also extend CDIAC emissions data where possible. Our emissions modeling framework was also designed to produce an annually updated emissions data product in a timely manner. Given the discontinuity of future updated CDIAC emissions data, we believe that our capability of producing an extended product of the CDIAC emissions data is significant.
Starting with national emissions estimates as an input, our model framework
achieves monthly, global FFCO2 gridded fields via preprocessing and spatial
and temporal disaggregation. CDIAC national estimates made by fuel type
(liquid, gas, solid, cement production, gas flare, and international bunker
emissions) are further divided into an extended set of ODIAC emissions
categories (point source, nonpoint source, cement production, gas flare,
international aviation, and marine bunkers; further described in Sect. 3).
It is important to note that ODIAC2016 carries emissions from international
bunkers (international marine bunkers and aviation often account for a few
percent of the global total emissions), which are not included in the CDIAC
gridded emissions data products (CDIAC gridded emissions data only indicate
national emissions and international bunker emissions are often not
considered to be a part of national emissions in an international
convention). With the inclusion of international bunker emissions, we
provide a more comprehensive global gridded emissions field. We extended the
CDIAC national estimates over the recent years that were not yet covered
in
the previous version of CDIAC gridded data (2014–2016) in order to support near-real
time CO
In the following sections (Sects. 3–5), we describe how ODIAC2016 was
developed. It is important to note that ODIAC2016 is based on the best
available data at the time of the development (ODIAC2016 was released in
September 2016). Thus, some of the emissions estimates and underlying data
used in ODIAC2016 might now be outdated. For traceability purposes, data
used in this development, their versions or editions, and data sources are
summarized in Table A1 in the Appendix. Following the results and evaluation section
(Sect. 6), we discuss caveats and current limitations in our modeling
framework and emissions data product (Sect. 7), and then we describe how we will
update the ODIAC emissions data product with updated fuel statistics and/or
emissions information (Sect. 8). Recently published atmospheric CO
A schematic figure of the ODIAC emissions modeling framework (defined as “ODIAC 3.0 FFCO2 model”). Starting with CDIAC national emissions estimates made by fuel type (emissions estimates), the CDIAC national emissions estimates are first divided into extended ODIAC emissions categories (input data processing; see Sect. 3). The ODIAC 3.0 FFCO2 model then distributes the emissions in space and time, using point source geolocation information and spatial data depending on emissions categories such as nighttime light (NTL) and aircraft and ship fleet tracks (spatial disaggregation; see Sect. 4). The emissions seasonality for emissions over land and international aviation were adopted from existing emissions inventories (temporal disaggregation; see Sect. 5).
CDIAC FFCO2 emissions estimates are based on fuel statistic data published
as the United Nation Energy Statistics Database (Boden et al., 2017). Emissions
estimates are calculated on a global, national, and regional basis and by fuel
type in the method described in Marland and Rotty (1984). CDIAC also
provides their own gridded emissions data products that indicate annual and
monthly FFCO2 fields at a 1
In CDIAC emissions estimates, the global total emissions and national total
emissions are obtained using different calculation methods (global fuel
production vs. apparent national fuel consumption; see Andres et al., 2012)
and the CDIAC national totals do not sum to the CDIAC global total due to the
difference in calculation method and inconsistencies in the underlying
statistical data (e.g., import–export totals) (e.g., Andres et al., 2012). We
thus calculate the difference between the global total and the sum of
national totals and scaled up national totals to account for the difference.
Andres et al. (2014) reported that global total emissions estimates calculated with
production data (as opposed to apparent consumption data) have the smallest
uncertainty (approximately 8 %; 2
The 2016 version of the CDIAC emissions estimates only covers years to 2013 (Boden et al., 2016). We thus extrapolated the 2013 CDIAC emissions to years 2014 and 2015 using the 2016 version of the BP global fuel statistical data (BP, 2017). Our emissions extrapolation approach is the same as Myhre et al. (2009) and Le Quéré et al. (2016). Emissions from cement production and gas flaring (approximately 5.7 and 0.6 % of the 2013 global total; Boden et al., 2016) were assumed to be the same as those in 2013. International bunker emissions were scaled using changes in national total emissions.
CDIAC national emissions estimates (prepared by fuel type) were recategorized into our own ODIAC emissions categories (point source, nonpoint source, cement production, gas flare, and international aviation and marine bunkers). Following Oda and Maksyutov (2011), the sum of emissions from liquid, gas, and solid fuels was further divided into point source emissions and nonpoint source emissions. The total emissions from point sources were estimated using national total power plant emissions calculated using Carbon Monitoring for Action (CARMA; Wheeler and Ummel, 2008) (Oda and Maksyutov, 2011). As mentioned earlier, CDIAC gridded emissions data products only indicate national emissions and do not include international bunker emissions (Andres et al., 1996, 2011). In contrast, EDGAR provides bunker emissions in their gridded data product (JRC, 2017). Peylin et al. (2013) show some models include international bunker emissions and some do not, although the difference due to the inclusion–exclusion of the international bunker emissions in the prescribed emissions could be corrected afterwards (Peylin et al., 2013). In ODIAC2016, we carry CDIAC international bunker emissions reported on a country basis to achieve the complete picture of the global fossil fuel emissions. Country total bunker emissions (aviation plus marine bunkers) were distributed using spatial proxy data adopted from other emissions inventories described later (see Sect. 4.3). Although the CDIAC/ORNL does not report emissions from international aviation and marine bunkers separately, we loosely estimated those two emissions using UN statistics. We estimated the fraction of aircraft emissions using jet fuel and aviation gasoline consumption and then the international bunker emissions were divided into aircraft and marine bunker emissions.
We define the sum of the emissions from solid, liquid, and gas fuels as land
emissions (see Fig. 1). Land emissions are further divided into two emissions
categories (point source emissions and nonpoint source emissions) and then
distributed at a 1
In ODIAC v1.7, emissions from gas flaring were not considered (Oda and
Maksyutov, 2011). Nighttime light pixels corresponding to gas flares often
appear very bright and would result in strong point sources in
emissions data (Oda and Maksyutov, 2011). We thus identified and excluded
those bright gas flare pixels before distributing land emissions using
another global nighttime light data product that was specifically developed
for gas flares by NOAA,
National Centers for Environmental Information (NCEI, formerly National
Geophysical Data Center, NGDC) (Oda and Maksyutov, 2011). In ODIAC2016 we
separately distributed CDIAC gas flare emissions using the 1
Emissions from international aviation and marine bunkers were distributed
using aircraft and ship fleet tracks. International aviation emissions were
distributed using the AERO2k inventory (Eyers et al., 2005). The AERO2k
inventory was developed by a team at the Manchester Metropolitan University
and indicates the fuel use and NO
The inclusion of the emissions temporal variations is often a key in transport model
simulation. For CO
Although ODIAC2016 only provides monthly emissions fields, users can derive
hourly emissions by applying scaling factors developed by Nassar et al. (2013). The Temporal Improvements for Modeling Emissions by Scaling (TIMES)
is a set of scaling factors that one can derive weekly emissions and
diurnal emissions from with any monthly emissions data. Temporal
profiles are collected from Vulcan, EDGAR, and the best available information
and are
gridded on a 0.25
Global emissions time series from four gridded emissions data: CDIAC (red, 2000–2013) plus projected emissions (dashed maroon, 2014–2015)
(values taken from ODIAC2016), CDIAC 1
In Fig. 2, global emissions time series from different emissions data were
compared to give an idea of agreement among them. We calculated the global
total for each year from four gridded emissions data for the period of
2000–2016: CDIAC global total
All four global total values obtained from four gridded emissions data agree well within 8 % uncertainty. The difference between ODIAC and CDIAC gridded data (3.3–5.7 %) was largely attributable to the international bunker emissions and global correction. ODIAC (where the total was scaled by CDIAC global total) and the two versions of EDGAR showed minor differences in magnitude (0.3–2.7 %) and trend, which are largely attributable to the differences in the underlying statistical data (e.g., UN Statistics Division vs. the US Energy Information Administration from different inventory years) and the emissions calculation method (fuel basis vs. sector basis). Global total estimates at 5-year increments are shown in Table 1. For the years 2014 and 2015, we estimated the global total emissions at 9.836 and 9.844 PgC. Boden et al. (2017) reported the latest estimate for 2014 global total emissions as 9.855 PgC. Our projected 2014 emissions estimate was lower than the latest estimate by approximately 0.02 PgC (0.2 %).
Global total emissions estimates for 2000, 2005, and 2010 from
four gridded emissions data estimates (ODIAC2016, CDIAC, EDGAR v4.2, and EDGAR
FastTrack). Values for two versions of EDGAR emissions data were calculated by
subtracting emissions from agriculture (IPCC code: 4C and 4D), land use
change and forestry (5A, C, D, F, and 4E), and waste (6C) from the total EDGAR
CO
NA
National emissions time series for top 10 emitting countries (China,
US, India, Russian Federation, Japan, Germany, Islamic Republic of Iran,
Republic of Korea (South Korea), Saudi Arabia, and Brazil). The values are
given in the unit of petagrams (equal to a gigaton) of carbon per year. The values
are calculated using gridded emissions data, not tabular emissions data. The
national total values in the plots might thus be different from values
indicated in the tabular form due to the emissions disaggregation. The shaded
area in grey indicates the 2
Figure 3 shows the same type of comparison as Fig. 2, but for the top 10
emitting countries (China, US, India, Russian Federation, Japan, Germany,
Islamic Republic of Iran, Republic of Korea (South Korea), Saudi Arabia, and
Brazil, according to the 2013 ranking reported by CDIAC). We aggregated
all four gridded emissions fields to a common 1
Annual uncertainty estimates associated with CDIAC national emissions estimates. The uncertainty estimates were made following the method described by Andres et al. (2014). The national total emissions for the year 2013 were taken from Boden et al. (2016).
The 2013 global fossil fuel CO
The global total emissions fields of CDIAC gridded emissions data and
ODIAC2016 for the year 2013 (the most recent year CDIAC indicates) are shown
in Fig. 4. Emissions fields are shown at a common 1
The 2013 global distributions of ODIAC fossil fuel emissions by
emissions type. The panels show emissions from (from top to the right, then
down) point source, nonpoint source, cement production, gas flaring,
international aviation, and international shipping. The values in the figures
are given in the unit of log of thousand tons of
carbon per year per cell
(1
In Fig. 6, we compared the four global gridded products over land and also calculated differences from ODIAC2016 (shown in Fig. 7; histograms are presented in Fig. A1). It is often very challenging to evaluate the accuracy and uncertainty of gridded emissions data because of the lack of direct physical measurements on grid scales (Andres et al., 2016). Recent studies have attempted to evaluate the uncertainty of gridded emissions data by comparing emissions data to each other (e.g., Oda et al., 2015; Hutchins et al., 2016). The differences among emissions were used as a proxy for uncertainty. However, it is important to note that such evaluation does not give us an objective measure of which one is closer to truth, beyond characterizing the differences in emissions spatial patterns and magnitudes from methodological viewpoints (e.g., emissions estimation and disaggregation). Some of the gridded emissions data are partially disaggregated using commercial information, and users are often not authorized to fully disclose the information used. This thus makes the comparison even less meaningful and/or significant. Oda et al. (2015) also discussed that emissions inter-comparison approaches often do not allow us to evaluate two distinct uncertainty sources (emissions and disaggregation) separately. In addition, because of the use of emissions proxy for emissions disaggregation (rather than mechanistic modeling), such comparison can be only implemented at an aggregated, coarse spatial resolution. These issues will be further discussed in Sect. 7.
Because of the limitation mentioned above, here we compared emissions data
only to characterize the differences that can be explained by the
differences in emissions disaggregation methods. We implemented this
comparison exercise using the 2008 emissions field aggregated at a 1
Land emissions from ODIAC
ODIAC minus other emissions data differences. CDIAC
Figure 8 shows time series of regional fossil fuel emissions aggregated over 11 land regions defined in the TransCom transport model intercomparison experiment (e.g., Gurney et al., 2002). The global seasonal variation and the associated uncertainty have been presented and discussed in Andres et al. (2011). Here monthly total emissions values were calculated for eleven TransCom land regions and presented with the associated uncertainty values (see Table 3). The monthly total values were calculated both excluding international bunker emissions (hence, land emissions only) and including the emissions. The uncertainty range was calculated with mass weighted uncertainty estimates of countries that fall into the TransCom regions. The uncertainty ranges shown in Fig. 8 show annual uncertainty plus the monthly profile uncertainty (12.8 %; reported by Andres et al., 2011). Monthly time series are presented for land-only emissions and land and international bunker emissions (here, largely aviation emissions). As described earlier, the emissions seasonality was adopted from Andres et al. (2011). The patterns in the emissions seasonality are often largely characterized by the large emitting countries within the regions (e.g., US for region 2 and China for region 8). Since Andres et al. (2011) used geographical closeness (also, type of economic systems) to define proxy countries, the countries in the same TransCom regions can have similar or the same seasonal patterns in their emissions.
As we can see in Fig. 4 (panel plot for aviation emissions), aviation
emissions are intense over North America, Europe, and Asia. Global total
aviation emissions was approximately 0.12 PgC yr
As the ODIAC emissions data product is now used for a wide variety of carbon cycle research (e.g., global, regional inversions, urban emissions studies), it would be useful for the users of the ODIAC emissions data product to note and discuss issues, limitations, and caveats in our emissions data. Some of the issues and limitations are specific to our study; however, the majority of them are often shared by other existing gridded emissions data and emissions models.
Emissions time series over inversion analysis land regions defined by
the Transport Model Intercomparison Project (TransCom) (Gurney et al.,
2002). The TransCom region map (bottom right) is available from
Annual uncertainty estimates over the TransCom land regions.
The uncertainty estimates were mass weighted values of uncertainty estimates of countries that fall in the regions.
Country uncertainty estimates were estimated using the method
described (Andres et al., 2014). The values were reported as the 2
In the production of ODIAC2016, we used several versions or editions of CDIAC estimates (e.g., global estimates, national estimates, and monthly gridded data). This could often happen in emissions data production, as some of the underlying data are not updated ro upgraded at the time of emissions data production (we often start updating emissions data after new fuel statistical data are released). We sometimes accept the inconsistency and try to use the most up-to-date information available. For example, we could use GCP's emissions estimates (e.g., Le Quéré et al., 2016) to constrain the global totals, if CDIAC global total emissions estimates are not available. The way we obtained emissions estimates for each version is often described in the NetCDF header information of the emissions data product. The use of the CARMA power plant estimates for estimating the magnitude of the point source portion of emissions is hard to eliminate, although ideally this is done using emissions estimates that are fully compatible with CDIAC estimates. We are currently examining UN statistical data (which CDIAC emissions estimates are based on) to assess the ability of explaining power plant emissions.
Although the use of the power plant geolocation allowed us to achieve
improved high-resolution emissions spatial distributions over land (Oda and
Maksyutov, 2011), the availability of power plant data is often very
limited. For example, CARMA does not provide power plant emissions and their
status (e.g., commission–decommission) every year. Furthermore,
updates and upgrades after their version 3.0 database (which is dated to 2012) are also not provided. The
error in their power plant geolocation is another issue that has been
identified (e.g., Oda and Maksytuov, 2011; Woodard et al., 2015). In ODIAC,
the base year emissions (2007) were projected and all the power plants were
assumed to be active over the period (Oda and Maksyutov, 2011). There are
only a few global projects such as
the Global Energy Observatory (GEO,
Emissions from cement production (which are currently distributed by Ziskin et al., 2010, using nighttime light) and gas flare (which is distributed by Elvidge et al., 2009, using gas flare nighttime light data) should be distributed as point sources. For gas flare emissions, we examine the use of Nightfire (Elvidge at al., 2013a) to pinpoint active gas flares in a timely manner and improve their emissions spatial disaggregation over recent years. Currently, the point source emissions in ODIAC do not have an injection height due to the lack of global information. This limitation is shared with other existing global emissions data products.
Nighttime light data have been an excellent proxy for human settlements
(hence, CO
In ODIAC, the disaggregation of nonpoint emissions is solely performed using nighttime light data for estimating subnational emissions spatial distributions, and no additional subnational emissions constraints were applied. Rayner et al. (2010) proposed to better constrain subnational emissions spatial distribution by combining population data, nighttime lights, and GDP in their Fossil Fuel Data Assimilation System (FFDAS) framework. Asefi-Najafabady et al. (2014) further introduced the use of point source information in their disaggregation; the optimization in their current framework is however under-constrained by the lack of GDP information. Without having such optimization, the state level per capita emissions estimates can provide subnational constraints. Nassar et al. (2013) evaluated the per capita emissions in CDIAC and ODIAC emissions data over Canada using the national inventory and found that ODIAC outperformed. However, as the nighttime light–population relationship might have a bias for developing and the least developed countries (Raupach et al., 2010), we would expect to see significant biases over those countries and the per capita estimates can provide a useful constraint.
As seen in the comparison to other emissions data, the major difference from EDGAR emissions spatial distribution was due to the lack of line sources in ODIAC. We do not believe the result from the emissions data comparison can falsify the emissions distribution in ODIAC, as discussed earlier. However, we do expect an inclusion of the line sources would improve the spatial distributions and emissions representations in both cities and rural areas. We are currently examining the inclusion of transportation network data (e.g., OpenStreetMap) as a proxy for line source emissions to explore the better spatial emissions aggregation method. Oda et al. (2017) recently implemented the idea of adding a spatial proxy for line sources and improved emissions estimates for a US city.
We estimated emissions from international aviation from CDIAC using UN
statistical data. The emissions are currently provided as a single layer
emissions field, although this is not appropriate given the nature of the
aviation emissions. Nassar et al. (2010) discussed the importance of the
three-dimensional (e.g.,
The emissions seasonality in ODIAC2016 is based on Andres et al. (2011) and
it can be further extended to an hourly
scale using the TIMES scaling parameter. We note that the emissions seasonality was based on the top 10 emitting
countries' fuel statistics and Monte Carlo simulation (Andres et al., 2011).
The emissions seasonality for countries other than the top 10 could be less
robust. Also, because of the use of Monte Carlo, the seasonality is
different over different editions of monthly emissions data. It is also
important to note that the repeated use of climatological (mean) seasonality
for recent years (described in Sect. 5) could be a source of
uncertainty and bias. Andres et al. (2011) estimated the monthly
uncertainty as 12.8 % (2
As mentioned earlier, the evaluation of gridded emissions data is often very
challenging and most of the emissions data studies share this difficulty.
Although the emissions estimates are made on global and national scales with
small uncertainties (e.g., 8 % for the global scale by Andres et al., 2014), considerable
errors seem to be introduced when the emissions are disaggregated (e.g.,
Hogue et al., 2016; Andres et al., 2016). Andres et al. (2016), for example,
estimated the uncertainty associated with CDIAC gridded emissions data on a
per grid cell basis with an average of 120 % and a range of 4.0 to 190 %
(2
While the quality (i.e., bias and uncertainty) of the gridded emissions
estimates remains unquantified for most of the emissions data mentioned in
this paper, the emissions data are still used because sufficient
measurements in space and time are not presently available to offer a better
alternative. At the very least, we presented the uncertainty estimates over the
aggregated TransCom land regions. We believe that the regional uncertainty
estimates are highly useful for atmospheric CO
The ODIAC2016 data product is available from a website hosted by the Center
for Global Environmental Research (CGER), Japanese National Institute for
Environmental Studies (NIES) (
We update the emissions data on an annual basis, following the release of an updated global fuel statistical data. Future versions of the emissions data are in principle based on an updated version or edition of the underlying statistical data with the same name convention (ODIACYYYY, YYYY is the release year; the end year is YYYY minus 1). In October 2017, we started distributing the updated 2017 version of ODIAC data (ODIAC2017, 2000–2016). We primarily focus on years after 2000. Future versions of ODIAC data, however, might have a longer, extended time coverage.
For detailed information about data availability, please refer to Sect. 8 in this paper.
This paper describes the 2016 version of ODIAC emissions data
(ODIAC2016) and how the emissions data product was developed within our
upgraded emissions modeling framework. Based on the CDIAC emissions data,
ODIAC2016 can be viewed as an extended version of the CDIAC gridded data
with improved emissions spatial distribution representations. Utilizing the
best available data (emissions estimates and proxy), we achieved a
comprehensive, global fossil fuel CO
A list of components in ODIAC2016 and data used in the development.
A table for the global scaling factor for 2000–2013.
A histogram of the inter-emissions data differences from
ODIAC. Values are given in the unit of million tons carbon per year
(MTC yr
The authors declare that they have no conflict of interest.
Tomohiro Oda is supported by the NASA Carbon Cycle Science program (grant no. NNX14AM76G). RJA is now retired but this work was sponsored by US Department of Energy, Office of Science, Biological and Environmental Research (BER) programs and performed at the Oak Ridge National Laboratory (ORNL) under the US Department of Energy contract DE-AC05-00OR22725. The authors would like to thank Chris Elvidge and Kim Baugh at NOAA/NGDC for providing the nighttime light data. The authors also thank Yasuhiro Tsukada and Tomoko Shirai for hosting the ODIAC emissions data on the data server at NIES.Edited by: David Carlson Reviewed by: two anonymous referees