On the performance of satellite-based observations of CO2 in capturing the NOAA Carbon Tracker model and ground-based flask observations over Africa land mass

Africa is one of the most data-scarce regions as satellite observation at the equator is limited by cloud cover and there are a very limited number of ground-based measurements. As a result, the use of simulations from models are mandatory to fill this data gap. A comparison of satellite observation with model and available in-situ observations will be useful to estimate the performance of satellites in the region. In this study, GOSAT XCO2 is compared with the NOAA CT2016 and six flask observations over Africa using five years of data covering the period from May 2009 to April 2014. Ditto for OCO-2 XCO2 5 against NOAA CT16NRT17 and eight flask observations over Africa using two years of data covering the period from January 2015 to December 2016. The analysis shows that theXCO2 from GOSAT is higher thanXCO2 simulated by CT2016 by 0.28 ppm whereas OCO-2 XCO2 is lower than CT16NRT17 by 0.34 ppm on African landmass on average. The mean correlations of 0.83 and 0.60 and average RMSD of 2.30 and 2.57 ppm are found between the model and the respective datasets from GOSAT and OCO-2 implying the existence of a reasonably good agreement between CT and the two satellites over Africa’s 10 land region. However, significant variations were observed in some regions. For example, OCO-2XCO2 are lower than that of CT16NRT17 by up to 3 ppm over some regions in North Africa (e.g., Egypt, Libya, and Mali ) whereas it exceeds CT16NRT17 XCO2 by 2 ppm over Equatorial Africa (10 S 10 N ). This regional difference is also noted in the comparison of model simulations and satellite observations with flask observations over the continent. For example, CT shows a better sensitivity in capturing flask observations over sites located in Northern Africa. In contrast, satellite observations have better sensitivity 15 in capturing flask observations in lower altitude island sites. CT2016 shows a high spatial mean of seasonal mean RMSD of 1.91 ppm during DJF with respect to GOSAT while CT16NRT17 shows 1.75 ppm during MAM with respect to OCO-2. On the other hand, low RMSD of 1.00 and 1.07 ppm during SON in the model XCO2 with respect to GOSAT and OCO-2 are determined respectively indicating better agreement during autumn. The model simulation and satellite observations exhibit similar seasonal cycles of XCO2 with a small discrepancy over Southern Africa and during wet seasons over all regions. 20


Introduction
Changes in atmospheric temperature, hydrology, sea ice, and sea levels are attributed to climate forcing agents dominated by CO 2 (Santer et al., 2013;Stocker et al., 2013). However, understanding the climate response to anthropogenic forcing in a more traceable manner is still difficult due to a major uncertainty in carbon-climate feedbacks (Friedlingstein et al., 2006). Part of this uncertainty is due to a lack of sufficient data on the regional and global carbon cycle. This is compounded 5 with inappropriate modeling practices to capture spatiotemporal variability of the carbon cycle. These problems can be solved through strengthening carbon monitoring networks, setting up proper modelling and reducing uncertainties in satellite retrieval.
Models with appropriate physical and mathematical formulations and sufficiently constrained by observations, can be used to understand the spatio-temporal nature of atmospheric CO 2 .
Towards this, a number of national and international efforts have been initiated in the recent past by different government 10 and non-government agencies across the globe. Among these efforts, ground-based observations of greenhouse gas using Total Carbon Column Observing Network (TCCON) is a notable one since it provides accurate and high-frequency measurements of column-integrated CO 2 mixing ratio. For example, it has been established that TCCON has a precision of 0.25% for measurements taken under clear sky conditions (Wunch et al., 2011). However, the number of TCCON sites is limited and can not establish an accurate CO 2 amount and flux on a subcontinental or regional scale. Moreover, some studies show that the 15 large uncertainty is amplified due to the uneven global distribution of TCCON sites (Velazco et al., 2017). In addition, none of these ground-based observation networks were found in Africa land mass. However, there are few TCCON sites around the continent plus some flask observations in and around Africa. For example, the TCCON station on Ascension Island records direct solar absorption spectra of the atmosphere in the near-infrared and retrieved accurate and precise column-averaged abundances of atmospheric constituents including CO 2 , CH 4 , N 2 O, HF, CO, H 2 O, and HDO . 20 On the other hand, the CO 2 concentration retrieved from the satellite-based CO 2 absorption spectra have the advantages of being unified, long-term, and global observations as compared to ground-based measurements. It has been established from theoretical studies that accurate and precise satellite-derived atmospheric CO 2 can appreciably minimize the uncertainties in estimated CO 2 surface flux (Rayner and O'Brien, 2001;Chevallier, 2007). Other studies have revealed that significant improvement in the estimation of weekly and monthly CO 2 fluxes can be achieved subject to CO 2 retrieval error of less 25 than 4 ppm from satellite and modeling scheme whereby CO 2 concentration is an independent parameter of the carbon cycle model (Houweling et al., 2004;Hungershoefer et al., 2010). However, XCO 2 shows temporal variability on different time scales: diurnal, synoptic, seasonal, inter-annual, and long term (Olsen and Randerson, 2004;Keppel-Aleks et al., 2011). More recent missions such as the Greenhouse gases Observing SATellite (GOSAT) (Hamazaki et al., 2005), the Orbiting Carbon Observatory-2 (OCO-2) (Boesch et al., 2011) and planned missions such as the Active Sensing of CO 2 Emissions over Nights, 30 Days, and Seasons (ASCENDS) (Dobbs et al., 2008) have been and are being developed specifically to resolve surface sources and sinks of CO 2 and provide information on these different scales of temporal variability. For example, GOSAT observations started in 2009 and provide XCO 2 based on spectra in the Short-Wavelength InfraRed (SWIR) region with a standard deviation of about 2 ppm with respect to ground-based and in-situ air-borne observations (Yokota et al., 2009;NIES GOSAT Project, 2012). The bias and performance of column averaged carbon dioxide (XCO 2 ) retrievals from an algorithm could change in different regions with differing land surfaces and anthropogenic emissions (Bie et al., 2018).
Moreover, the NOAA Carbon Tracker (CT) is an integrated modeling system that assimilates CO 2 from other observations in order to complement satellite observations in understanding CO 2 surface sources and sinks as well as its spatiotemporal variabilities. However, both satellite and model data should be validated against other independent satellite observations and/or 5 in-situ observations before using them to answer scientific questions. As a result, a lot of validation and intercomparison have been conducted in previous studies. For example, Kulawik et al. (2016) found root mean square deviation of 1.7, and 0.9 ppm in GOSAT and CT2013b XCO 2 relative to 17 TCCON sites acroos the globe respectively. Other authors have undertaken validation exercises and found the bias of -8.85 ± 4.75 ppm in retrieving XCO 2 from the GOSAT observed spectrum by Japans the National Institute for Environmental Studies (NIES) level 2 V02.xx XCO 2  with respect 10 to TCCON (Morino et al., 2010). In addition, Chevallier (2015) shows retrieved XCO 2 from GOSAT observed spectrum by NASA Atmospheric CO 2 Observations from Space (ACOS) (O'Dell et al., 2012) suffers a systematic error over African Savanna. Lei et al. (2014) also showed a regional difference of XCO 2 between the ACOS and NIES datasets. For example, a larger regional difference from 0.6 to 5.6 ppm was obtained over China land region, while it is from 1.6 to 3.7 ppm over the global land region and from 1.4 to 2.7 ppm over US land region. These findings suggest that it is important to assess the accuracy 15 and uncertainty of XCO 2 from satellite observations with respect to more accurate models (e.g., NOAA Carbon Tracker) and ground-based observations over other regions as well. As satellite retrievals are strongly constrained by cloud cover, aerosol lodgings, land use change and Africa is a continent with wide extremes in surface type (which ranges from desert, rainforest and Savannah) and aerosol loading. In addition, seasonal variation of biomass burning in Africa: agricultural residues burned in the field, savanna burning, and forest wild fires results in a very seasonal aerosol loading in African. Africa is under the 20 influence of semi-permanent high-pressure cells which led to the Sahara Desert in the North and the Kalahari in the South. The equatorial low-pressure cell which allows the formation of the seasonally migrating Inter-Tropical Convergence Zone (ITCZ) is part of the major large scale atmospheric circulation systems. These large scale pressure systems, Oceanic circulations and their interaction with the atmosphere coupled with diverse topographies of the region allow for the formation of different climates (e.g., equatorial, tropical wet, tropical dry, monsoon, semi desert (semi arid), desert (hyper arid), subtropical high 25 climates). Geographically, the Sahel, a narrow steppe, is located just south of Sahara; the central part of the content constitutes the largest rainforest next to Amazon whereas most southern areas contain savanna plains. The continent gets rainfall from migrating ITCZ, west Africa monsoon, the intrusion of mid-latitude frontal systems, travelling low pressure systems (Mitchell, 2001, and references therein). Since CO 2 fluxes exhibit seasonal variability and Africa experiences different seasons as noted above, it is important to divide Africa into three major regions, namely North Africa (10 to 35 0 N ), Equatorial Africa (10 0 S to 30 10 0 N ), and Southern Africa (35 to 10 0 S) and conduct the comparison of the two XCO 2 datasets. Assessing the performance of satellites over the region can tell much about how these systematic errors vary geographically over the continent.
Therefore, this paper aims to assess the performance of observed XCO 2 from GOSAT and OCO-2 satellite in capturing simulated XCO 2 from NOAA Carbon Tracker model over Africa. These satellite observations and Carbon Tracker mixing ratios near the surface are also compared to available in suit CO 2 flask data from Assekrem, Algeria; Mt. Kenya; Gobabeb, 35 3 Namibia; and Cape Town; as well as to data off the coast at Seychelles, Ascension Island, and at Izana, Tenerife. Moreover, the consistency between the model and satellite observations in capturing the amplitudes and phases of observed seasonal cycles over different parts of the continent are evaluated. The agreement of modeled spatiotemporal variability with the known seasonal climatology of the regions, that determines carbon source and sink levels, is also assessed.
2 Data and Methodology 5

Carbon Tracker Model and Data
Carbon Tracker provides an analysis of atmospheric carbon dioxide distributions and their surface fluxes (Peters et al., 2007). It is a data assimilation system that combines observed in situ carbon dioxide concentrations from 81 sites around the world with model predictions of what concentrations would be based on a preliminary set of assumptions ("the first guess") about sources and sinks for carbon dioxide. Carbon Tracker compares the model predictions with reality and then systematically tweaks and 10 evaluates the preliminary assumptions until it finds the combination that best matches the real world data. It has modules for atmospheric transport of carbon dioxide by weather systems, for photosynthesis and respiration, air-sea exchange, fossil fuel combustion, and fires. Transport of atmospheric CO 2 is simulated by using the global two-way nested transport model (TM5).
TM5 is an offline atmospheric tracer transport model (Krol et al., 2005) driven by meteorology from the European Centre for Medium-Range Weather Forecasts (ECM W F ) operational forecast model and from the ERA-Interim reanalysis (Dee et al.,15 2011) to propagate surface emissions. TM5 is based on a global 3 0 × 2 0 and at a 1 0 × 1 0 spatial grids over North America.
The model can be used in a wide range of applications, which includes aerosol modeling, stratospheric chemistry simulations, hydroxyl-radical trend estimates. Detailed description of the TM5 model can be found in the works of ? Krol et al. (2005) CT date from the CT2015 release and on wards uses aircraft profiles from the stratosphere to the top of the atmosphere (Inoue et al., 2013;Frankenberg et al., 2016) and also co-location error are quantified . The older data 20 versions have been used and also compared with different data sets over other parts of the globe in previous studies (Nayak et al., 2014;Kulawik et al., 2016). Most of the studies confirm that CT XCO 2 captures observations reasonably well. In this study, we use Carbon Tracker release version CT2016 (Peters et al., 2007), hereafter (CT2016) and near real-time version (CT-NRT.v2017). Both versions of NOAA CT provides 3 hourly CO 2 mole-fractions data for global atmosphere at 25 pressure levels in a 3 0 × 2 0 spatial resolution for a period covering 2000 to 2016. The data can be accessed freely at the public domain 25 (ftp://aftp.cmdl.noaa.gov/products/carbontracker).

GOSAT measurements
GOSAT is the world's first spacecraft partiality designed to measure the concentrations of carbon dioxide and methane, the two major greenhouse gases, from space. The spacecraft was launched successfully on January 23, 2009, and has been operating properly since then. GOSAT records reflected sunlight using three near-infrared band sensors. The field of view at nadir allows 30 a circular footprint of about 10.5 km diameter (Kuze et al., 2009;Yokota et al., 2009;Crisp et al., 2012). GOSAT consists of two instruments. The sensors for the two instruments can be broadly labeled as thermal, near infrared and imager. The first two sensors are used as part of Fourier Transform Spectrometer for carbon monitoring which is referred to as TANSO-FTS while the imager for cloud and aerosol observations is referred to as TANSO-CAI. The details on spectral coverage, resolution, field of view, and different products of TANSO-FTS in the three SWIR bands can be found in a number of previous studies (Kuze et al., 2009;Saitoh et al., 2009;Yokota et al., 2009Yokota et al., , 2011Crisp et al., 2012;Nayak et al., 2014;Deng et al., 2016a, 5 and references therein). In this study ACOS B3.5 Lite XCO 2 from GOSAT Level 2 (L2) retrieval based on the SWIR spectra of FTS observations and made available by Atmospheric CO 2 Observations from Space (ACOS) of NASA is used. ACOS B3.5 Lite XCO 2 has lower bias and better consistency than NIES GOSAT SWIR L2 CO 2 globally (Deng et al., 2016a).
However, this version of ACOS XCO 2 found to suffer systematic retrieval error over the dark surfaces of high latitude lands and and over African savanna (Chevallier, 2015). Chevallier (2015) shows systematic error in the African savanna associated 10 to underestimating the intensity of fire during March at the end of the savanna burning season. Therefore, our choice of the ACOS B3.5 Lite, hereafter (GOSAT) XCO 2 is motivated by these differences.

OCO-2 measurements
OCO-2, the world's second full-time dedicated CO 2 measurement satellite. It was successfully launched by the National Aeronautics and Space Administration (NASA) on 2 July 2014 (Crisp et al., 2012). OCO-2 measures atmospheric carbon 15 dioxide with the accuracy, resolution, and coverage required to detect CO 2 source and sink on global and regional scale. OCO-2 has three-band spectrometer, which measures reflected sunlight in three separate bands. The O 2 A-band measures molecular absorption of oxygen from reflected sunlight near 0.76 µm while the CO 2 bands are located near 1.61 µm and 2.06 µm (Liang et al., 2017). In this study, both the nadir and glint-mode measurements of OCO-2 XCO 2 V7 lite level 2 covering the period from January 2015 to December 2016, hereafter referred to as OCO-2 XCO 2 are used. Due to the scarcity of data, CT values 20 from the two releases CT2016 for the year 2015 and CT-NRT.v2017 for the year 2016, hereafter (CT16NRT17) are employed in this study. The OCO-2 project team at Jet Propulsion Laboratory, California Institute of Technology, produced the OCO-2 XCO 2 data used in this study. The data can be accessed from NASA Goddard Earth Science Data and Information Service Center. 25 Measurements of CO 2 from nine ground-based flask observations near and within Africa land mass were accessed from the NOAA/ESRL/GMD CCGG cooperative air sampling network https://www.esrl.noaa.gov/gmd/ccgg/flask.php. Sites description is given in Table 1.

Methods
The GOSAT and CT model XCO 2 time series used in this investigation span five years, ranging from May 2009 to April 2014. 30 Atmospheric CO 2 concentrations of NOAA Carbon-Tracker have global coverage with a 3 0 ×2 0 Longitude/Latitude resolution which covers 426 grid boxes in our study area. Satellite observations, however, are different from model assimilation and have gaps because of various reasons (e.g., cloud and the observational mode of the satellite). As a result, there is no one to one spatiotemporal match between the two data sets. For example, CO 2 products from the two datasets are not directly comparable since CT is a 3 hourly smooth and regular grid dataset whereas GOSAT XCO 2 is irregularly distributed in space and time.
Thus, the CT CO 2 is extracted on the time and location of GOSAT-XCO 2 data. Using the grid point of CT as a reference bin, 5 the corresponding GOSAT XCO 2 found within a rectangle of 3 0 × 3 0 with center at the reference bin and with a temporal mismatch of a maximum of 3 hrs is extracted. Moreover, CT has higher vertical resolutions than GOSAT. As a result, the two can not be directly compared. It is customary to smooth the high-resolution data (in this case CT) with averaging kernels and a priori profiles of the low-resolution satellite measurements (in this case GOSAT). Besides, due to a difference between CT and GOSAT on the number vertical levels, CT CO 2 is interpolated to vertical levels of GOSAT. The CT XCO 2 (XCO model (2018) and given as: where i is the index of the satellite retrieval vertical level and T is the matrix transpose. To compare the CT simulations and the Satellites observation with the flask observations, the vertical profile of the satellite and CT were extracted at the corresponding pressure level and location within a box of 1.5 0 .
Correlation coefficients (R), bias and root mean square deviation (RMSD) are used to assess the level of agreement between the two data sets. The mean bias determines the average deviations in XCO 2 between Carbon Tracker simulation and satellite 20 observations. In this work the bias at the j th grid point is computed as: where S i and O i are CT and GOSAT XCO 2 values over the j th pixel at the i th time respectively. To quantify the extent to which XCO 2 of CT and GOSAT agree, the pattern correlations at the j th grid point are computed as: whereS andŌ are the mean values of S i and O i over the j th pixel. The root mean square deviation (RMSD) which shows the standard error of the model with respect the observation at the j th grid point is computed as : this is the centered pattern root mean squared (RMS) difference which is obtained from the RMS error after the difference in the mean has removed (Taylor, 2001). 10 Comparison with in situ flask observation is achieved in a way that the Carbon Tracker and satellite observations are taken at a corresponding pressure level of the in-situ flask observation (as mentioned in Table 1) in order to correspond to flux-towers surface observation. Further the datasets are re sampled to fit the flask observations in a 3 0 X3 0 window centered the flux-towers and to the available months were averaged.
3 Results and discussions  anomalous high XCO 2 . The Southern Africa region is characterized by weak anthropogenic CO 2 emission and high CO 2 uptake by the vegetation than Northern Africa (Ciais et al., 2011). This contributed to the observed dipole distribution. Another 25 important pattern is anomalous peak over the annual average location of the Inter-tropical convergence zone (ITCZ) (Fig. 1b) which appears to fade over Eastern Africa. This is in agreement with the fact that carbon stocks and net primary production per unit land area is higher over Equatorial Africa and decreases towards northward and southward of the equator over arid environments (Williams et al., 2007). However, Fig. 1b shows that GOSAT observations has some limitations in simulating this spatial pattern in comparison to CT. Fig. 1c shows the mean difference (CT2016-GOSAT) XCO 2 which ranges from -4 to 2 ppm. The highest difference between the CT2016 and GOSAT XCO 2 (as high as -4 ppm) is observed over Northern part of Equatorial Africa (e.g., southern Guinea, 5 southern Ghana, southern Nigeria, southeast of Central Africa, western Ethiopia and South Sudan, .etc.) which are also known for near-year-round rainfall and relatively dense vegetation. The regions are known for their rain forest (Malhi et al., 2013).
The likely explanation could be XCO 2 the mean (over five years) climatology may be slightly positively biased due to fewer GOSAT observations as shown in Fig.1d. The satellite retrievals has noise which can be smoothed out when large number of datasets are averaged. The strategy and methods for cloud screening in GOSAT retrievals could lead to a smaller number  Fig. 2a shows differences between CT2016 and GOSAT XCO 2 ranges from -4 to 3 ppm. Out of 100% occurrence, more than 90% of observed differences are within ± 2 ppm. The mean difference between CT2016 and GOSAT means is about -0.27 ppm with the standard deviation of 0.98 ppm indicating better regional consistency and low potential outliers. Moreover, a negative mean of the difference implies that XCO 2 simulated from CT2016 is lower than that of GOSAT retrievals over Africa land mass. 5 Because of selection criteria which permits a difference of 3 degrees long and wide, the two datasets are not exactly at the same point. The impact of the relative distance between them should be assessed before performing any statistical comparison. Fig. 2b depicted color-coded scatter plot of CT2016 model simulation verses GOSAT to determine if the discrepancy between the datasets arise from the spatial mismatch. The color code indicates the relative distance between the model and observation datasets. For these datasets the 50 th percentile has a relative distance of 1.19 0 which means 50% of the data has a relative 10 distance of shorter than 1.19 0 . The maximum relative distance between them is 2.12 0 . However, there is no indication that this has been the case since the scatter is not a function of the relative distance between the data sets. For example, data points with blue color with the lowest location difference is scattered everywhere instead of along the 1:1 line. Furthermore, we found the bias of -0.26 ppm, correlation coefficient of 0.86 and RMSD of 2.19 ppm for datasets which has a relative distance shorter than 1.19 0 . On the other hand, the bias, correlation coefficient, and RMSD are -0.33 ppm, 0.86 and 2.22 ppm for those which are 15 above 1.19 0 . These statistics provide information there will be no strong discrepancy due to our selection criteria. The above statistics was performed merely to test the influence of location mismatch.  Fig. 3 shows a statistical comparison of XCO 2 from the CT2016 and GOSAT over Africa. The number of data used in this comparison is shown in Fig. 1d. As it is depicted in Fig. 3a, the bias ranges from -4 to 2 ppm with a mean bias of -0.28 ± 1.05 ppm (see Table 2). A larger negative bias of about -2 ppm was found along with the annual mean position of ITCZ, the main climatic mechanisms controlling rainfall in Africa. Systematic errors duo to ITCZ and the East African Monsoon needs to be addressed well in satellite retrievals and modeling works. The correlation varies from 0.4 over some isolated pockets in Congo, 5 Tanzania, Mozambique, Uganda, and western Ethiopia to 0.9 over the northern part of Africa above 13 0 N , Eastern Ethiopia and the Kalahari Desert. Fig. 3b depicts correlation coefficient between GOSAT and Carbon Tracker XCO 2 . The region with poor correlation also exhibits high RMSD as shown in Fig. 3c. To understand whether this discrepancy originates from model weakness alone or terrible satellite visibility when the ITCZ is present and clouds are extremely thick and widely present, we have looked at the GOSAT posterior estimate of XCO 2 error (Fig. 3d), which are high over regions where the bias and 10 RMSD between GOSAT and Carbon Tracker XCO 2 is high. GOSAT's posterior estimate of XCO 2 error is a combination of instrument noise, smoothing error and interference errors (Connor et al., 2008;O'Dell et al., 2012). This posterior estimate of XCO 2 error does not include forward model error which may lead to underestimation of the true error of satellite XCO 2 by a factor of two (O'Dell et al., 2012). Therefore, part of the discrepancy is clearly linked to satellite retrieval uncertainty, which might have been amplified due to the small number of data points used to calculate the mean error of GOSAT XCO 2 15 measurements (see Fig. 1d). In general, the two data sets are characterized by a high spatial mean correlation of 0.83 ± 1.20, a global offset of -0.28 ± 1.05 ppm, which is the average bias, a regional precision of 2.30 ± 1.46 ppm, which is average RMSD and relative accuracy of 1.05 ppm which is the standard deviation in the bias as depicted in Table 2.    Table 4). Moreover, the two dataset shows a monthly mean regional mean bias of -0.36 ppm with a correlation of 1.0 and small root mean square deviation of 0.36 ppm (see Table 3).  (Table 4). Moreover, both datasets show that concentration of CO 2 increases from October to March while it decreases from June to October. This similarity in the seasonal variability of the two datasets shows that they are in good agreement in terms of amplitude and phase. In addition, the two datasets show a monthly average regional average bias of -0.17 ppm, correlation of 0.98 and a small root mean square deviation of 0.71 ppm over Equatorial Africa (see Table 3). Fig. 6a  This is most likely CT2016 simulation is more sensitive to the growing size of sink following the rainy season. Moreover, the two datasets show a monthly mean regional mean bias of 0.07 ppm, correlation of 0.97 and RMSD of 0.87 ppm over southern Africa (see Table 3).

10
Figs. 4b -6b show regional averaged bias in the monthly mean XCO 2 from CT2016 and GOSAT.    and a standard deviation of 0.85 ppm which indicates that XCO 2 from CT2016 was slightly higher than that of GOSAT over Southern Africa on average. In addition, the low standard deviation of monthly mean difference over North Africa typically indicates good regional consistency between CT2016 and GOSAT. This is mainly because Northern Africa is dominated by the Sahara desert, which is a vegetation free area, and the systematic bias due to the local atmosphere biosphere interaction is minimum. However, the spatial mean of monthly mean bias is slightly higher (-0.36 ppm) over North Africa than over 10 Equatorial Africa (-0.17 ppm ) and Southern Africa (0.01 ppm). This is possibly due to the presence of strong local emissions from Egept, Algeri and Libya as well due to long-range transport from the Northern Hemisphere as reported in other studies (Williams et al., 2007;Carré et al., 2010 (Kulawik et al., 2015). The growth rate may not be conclusive due to the short length of the datasets used. However, it reflects how the CT and GOSAT observations perform with respect to each other.   The right panels in Fig. 7 show that the seasonal mean difference (CT2016 -GOSAT) ranges from -4 to 6 ppm. A maximum difference of 6 ppm over the Gulf of Guinea and Congo during JJA. However, such maximum difference was also observed over Southern Africa during DJF. A minimum of -4 ppm over annual mean ITCZ region was observed during DJF and MAM.

Comparison of seasonal climatology
Moreover, the difference is above 1 ppm over Southern Africa regions during DJF and MAM (wet season of the region). This implies high spatial variability of the seasonal mean difference during different seasons (see also Table 5). It also suggests that 5 the discrepancy between the CT2016 and GOSAT becomes significant when vegetation cover is weak during DJF and MAM (dry seasons) over North Africa.
During SON the seasonal difference in most Africa's land region ranges from -2 to 1 ppm. The result implies CT2016 simulates lower values of XCO 2 than that of GOSAT observation indicating that there is a better spatial consistency during this season. Furthermore, during these seasons both the Northern and Southern Africa have a moderate vegetation cover following 10 their respective summer seasons. The two datasets show lower regional variation (i.e., only from -2 to 2 ppm) over most of Africa land mass. However, Equatorial Africa exhibits the mean difference lower than -2 ppm during DJF and MAM. This indicates the model tends to simulate lower than GOSAT retrievals XCO 2 over the region. Fig. 7 Mengistu Tsidu et al. (2015) have also shown that ERA-Interim precipitable water is higher than measurements from radio-sonde, FTIR and GPS observations. Therefore, such wet bias in the driving ERA-Interim GCM might have forced NOAA CT2016 to generate dense vegetation which serves as CO 2 sink.  Fig. 8 shows the mean difference between CT2016 and GOSAT XCO 2 seasonal means which ranges from -0.37 to 0.04 ppm with a standard deviation within a range of 1.00 to 1.91 ppm over the continent. The highest mean difference of XCO 2 (-0.37 ppm) occurs during SON and the lowest (0.04 ppm) occurs during MAM. Table 5 presents the summary of statistical values for the spatial mean of each season means. The comparison between the two data sets also shows there is a strong correlation (>0.5) during each season over the continent. However, there are moderate correlations (0.3 to 0.5) during DJF and 5 MAM over North Africa and during DJF over Southern Africa. The low correlation over Northern Africa may be linked to a weak absorption by vegetation and a strong emission from human activities during winter as reported elsewhere (Liu et al., 2009;Kong et al., 2010). Moreover, Table 5 shows that the seasonal biases are negative over North Africa while they are mostly positive over Equatorial and Southern Africa. Negative biases are observed during DJF and SON over Equatorial and Southern Africa respectively implying that XCO 2 from CT2016 are lower than GOSAT during dry seasons. For the comparison, the volume mixing ratio of CO 2 from GOSAT and CT2016 at the pressure level that corresponds to surface observation of flask (see Table 1 Figure 9. CO2 time series for the coincident period for CT2016 (red), GOSAT (green) and flask (black). The standard deviation in computing the monthly mean is indicated by the vertical error bar.
Monthly mean CO 2 from flask observations at IZO and ASK in northern Africa shows an excellent agreement with both 5 CT2016 and GOSAT CO 2 . Moreover, CT2016 has a better sensitivity in capturing the amplitudes than GOSAT where observations from GOSAT mostly under estimates higher values of flask CO 2 (Fig. 9). However, this agreement has deteriorated over sites in Equatorial Africa (ASC and MKN) and Southern Africa (MNB). Over MKN, CT2016 shows better correlation (0.43) than GOSAT observation (0.08). In addition, monthly amplitudes from CT2016 was closer to the flask observations suggesting that satellite retrievals need much attention over the region. On the other hand, GOSAT observations were found to be in better 10 agreement with flask observations over ASC. Zhang et al. (2015) also show that GOSAT data was correlated well with ground observation and found to be more centralized, having high system stability, especially over the ocean.
CT2016 has a better sensitivity over IZO, ASK and NMB. Moreover, CT2016 compared well with flask observations than GOSAT over these sites, almost all flask observations are within the standard deviations of the monthly mean of CT2016.
However, GOSAT observations were found in better agreement with flask observations than CT2016 was over WIS and ASC.

15
On the other hand, both CT2016 and GOSAT have low sensitivity to flask observation over MKN (see Fig. 10). Similar to our previous discussion over sites in the Northern Africa (IZO, ASK and WIS), CT2016 underestimates XCO 2 during August,  September, and October (wet season) compered to GOSAT observation and overestimates during January to June. However, the CT2016 and the flask observations exhibit better agreement indicating a bias in GOSAT observation during the wet season.

Comparison of mean XCO 2 from NOAA CT16NRT17 and OCO-2
The strong El Niño event occurred during 2015-2016 provides an opportunity to compare the performance of CT16NRT17 during strong El Niño events. Because of the decline in terrestrial productivity and enhancement of soil respiration, the con-5 centration of CO 2 increases during El Niño events (Jones et al., 2001). In this section we compare mean XCO 2 of NOAA CT16NRT17 and NASA's OCO-2 covering the period from January 2015 to December 2016. The comparison was done based on the selection criteria discussed in Section 2.5. Fig. 11 shows mean distribution of XCO 2 from CT16NRT17 (Fig. 11a) and OCO-2 (Fig. 11b)  to 400 ppm over Southern Africa. The XCO 2 distribution from OCO-2 is consistent with the maximum CO 2 concentration reported in past study by Williams et al. (2007) implying that the CT16NRT17 likely underestimates XCO 2 values over Equatorial Africa. It is also possible that the discrepancy is a compounded effect of OCO-2 XCO 2 positive bias over the region (O'Dell et al., 2012;Chevallier, 2015). Fig. 11c shows the mean difference between two years mean of XCO 2 from CT16NRT17 and OCO-2, which is in the range from -2 to 2 ppm. However, high (<-2 ppm) negative mean difference between 10 the two data sets over rain forest regions (Gulf of Guinea and Congo basin) and ITCZ zone over Eastern Africa (South Sudan and southeastern Sudan) is observed implying that CT16NRT17 simulates lower XCO 2 values than that of OCO-2 observation over regions where vegetation uptake is strong. Conversely, high (>1) positive mean difference over the Sahara desert, Somalia and Tanzania implies CT16NRT17 simulates higher XCO 2 values than OCO-2 observation where the vegetation uptake is weak. Moreover, a positive (>2) mean difference over Egypt, Libya, Sudan, Chad, Niger, Mali and Mauritania is likely due 15 to overestimates of XCO 2 emission from local sources by CT16NRT17. Overall, the two datasets show a fairly reasonable agreement with a correlation of 0.60 and offset of 0.36 ppm, a regional precision of 2.51 ppm and a regional accuracy of 1.21 ppm. Table 7. Summary of statistical relation between CT16NRT17 and OCO-2 observation. The statistical tools shown are the mean correlation coefficient (R), the average of bias (Bias), the average root mean square deviation (RMSD), the standard deviation in bias (std of Bias), mean posteriori estimate of XCO2 error from OCO-2 (OCO-2 err), the standard deviation in CT16NRT17 XCO2 (CT16NRT17 std) and the standard deviation in OCO-2 XCO2 (OCO-2 std). Positive Bias indicates that CT16NRT17 is higher than OCO-2. The number of data used in the statistics is 1,659,411 over 426 pixels covering the study period, distribution at each grid point is shown in Fig 11d. Statistical tool R Bias (ppm) RMSD (ppm) std of Bias (ppm) OCO-2 err (ppm) CT16NRT17 std (ppm) OCO-2 std (ppm) Values 0.6 0.34 2.57 1.21 0.55 0.55 1.28 Fig. 12a shows the histogram of two years mean difference, which is characterized by a positive mean of 0.34 ppm and a standard deviation of 1.21 ppm. This suggests that CT16NRT17 simulates high XCO 2 as compared to observations from OCO-2 over Africa's land mass.
5 Figure 12. Histogram of the difference of CT16NRT17 relative to OCO-2 (left panel) and color code scatter diagram of XCO2 concentration as derived from CT16NRT17 and OCO-2 (right panel). Color indicates the relative distance in unit of degrees as shown in colorbar between datasets.
Because of presence of spatial and temporal mismatch of some level between CT16NRT17 and OCO-2 datasets, it is important to assess the effect of relative distance between the datasets. Fig. 12b shows a color coded distribution of the two datasets.
In the figure color codes indicate the relative distance. The random scatter of blue dots implies that the statistical discrepancies do not arise from the relative distance between the two datasets. More specifically, a statistical comparison of datasets lower and higher the 50 th percentile (1.2 0 ) shows bias of 0.58 and 0.57 ppm, correlation of 0.57 and 0.57 and RMSD of 2.65 and 2.67 ppm respectively.  December 2016. The number of data used are displayed in Fig. 11d. Fig. 13a depicts the bias which ranges from -2 to 2 ppm with a mean bias of 0.34 ppm. However higher biases (<-2 ppm) are observed over Equatorial Africa along the annual average location of ITCZ. Fig. 13b shows the correlation map with values from 0.2 to 0.8 over Africa's land mass. A good correlation of above 0.6 are seen over many regions of the continent while weak correlation of less than 0.2 and higher root mean square error (> 3 ppm ) are observed over small pockets of Equatorial and Eastern Africa regions (see Fig. 13c). These regions also 10 show a higher (> 0.65 ppm) error in satellite retrieval (see Fig. 13d). In addition, Fig. 11d  into the atmosphere (Chatterjee et al., 2017). Fig. 14a also shows that XCO 2 from CT16NRT17 simulation are higher than OCO-2 observation over North Africa.  the difference between the two datasets is minimum; On the other hand, a maximum difference of exceeding 1.5 ppm was 15 observed during MAM which can be mentioned as a burning season of Northern Africa, as area north of the equator was burned mostly from March to June (Hao and Liu, 1994). The observed lower XCO 2 values from OCO-2 observations than that of CT16NRT17 simulation will be a consequence of much respiration which exceeded photosynthesis when vegetation uptake is weak following the strong El Niño and dry season over North Africa. Further more, intense burning of during this season may cause more aerosol loading which will further intensified by of strong El Niño may not sufficiently estimated.
Moreover, Fig. 14c displays a monthly mean regional mean bias of 0.87 ppm, correlation of 0.95 and a root mean square 5 deviation of 0.72 ppm between CT16NRT17 and OCO-2 XCO 2 . This implies that CT16NRT17 is in a good agreement with OCO-2. However, a small discrepancies arose most likely due to a strong anthropogenic emission from Nigeria, Egypt and Algeria. a global mean of more than 3 gigatone of CO 2 added to the atmosphere due to the strong El Niño event that occurred during 2015-2016. In agreement with this, both CT16NRT17 and OCO-2 shows an annual growth rate that ranges from 3.10 to 3.42 ppm year −1 of XCO 2 over Africa's land mass (see also shows lower XCO 2 annual growth rate than those of OCO-2. their difference (right panels) covering period of January 2015 to December 2016. The white space seen over some regions (e.g., Mali during JJA) is due to insufficient coincident satellite data according to the selection criteria during these seasons. 25 XCO 2 increases from winter to spring and then decreases from spring peak to summer minimum over the whole continent. The decrease from spring maximum to summer continued into autumn over northern half of Africa in contrast to southern half of Africa which exhibits an increase in XCO 2 . The decrease from spring to autumn (northward of equator) and until summer (southward of equator) is likely to be a consequence of the land vegetation awakening from dormancy of winter and partly spring. Conversely, the decomposition of died and decayed vegetation which began in autumn and continued throughout winter adds extra CO 2 leading to a maximum concentration during spring (Idso et al., 1999). In agreement with this, both CT16NRT17 and OCO-2 show maximum XCO 2 during MAM over North Africa and during SON over Southern Africa.
Conversely, minimum concentrations are observed during SON over North Africa and during DJF over South Africa. Figure 17. Seasonal mean of CO2 for NOAA CT16NRT17 (left panels) and OCO-2 (middle panels) and their difference (right panels). Fig. 17 (right panels) shows the seasonal mean difference of CT16NRT17 and OCO-2. A higher mean difference greater 5 than 1 ppm is observed over North Africa during DJF and MAM when the vegetation cover over the region decreases and also an intensive burning of the northern savanna during this season (Hao and Liu, 1994). This indicates that XCO 2 values from CT16NRT17 are higher than that of OCO-2 when vegetation uptake is weak and more fire. On the other hand, higher negative mean difference of less than -2 ppm are observed over Equatorial Africa during DJF during SON over Southern Africa. This difference between the CT and OCO-2 arises likely due to grass fires from the dry savanna. Consistent with report by Liang 10 et al. (2017), low seasonal variability is observed between CT16NRT17 and OCO-2 in the range from -4 to 4 ppm with greater amplitude over North and Equatorial Africa than over Southern Africa (see Fig. 17 (right panels)). During dry seasons OCO-2 over estimates values over the Northern Africa but it underestimates for the Southern Africa.   Figure 19. CO2 from CT16NRT17, OCO-2 and flask observations. Monthly CT16NRT17 XCO 2 has a better sensitivity over IZO and ASK both in terms of temporal pattern (phase) and amplitude than OCO-2 (see Fig. 19) where observations from OCO-2 mostly underestimates XCO 2 at the two flask sites. Over LMP and WIS, both CT16NRT17 and OCO-2 have moderate sensitivity in capturing the seasonal cycle. On the other hand, OCO-2 has a better sensitivity over ASC and SEY. In addition, XCO 2 from both CT16NRT17 and OCO-2 is found to have poor correlations with flask observations over NMB and CPT. However, OCO-2 has closer sensitivity in capturing amplitudes than CT16NRT where CT16NRT17 overestimates XCO 2 at these flask sites. In general, CT has a better performance over sites located at high altitude (IZO, ASK) where satellite observations underestimates XCO 2 . Conversely, satellite observations 5 have better performance over low altitude island sites (ASC and SEY) as revealed by better agreement with flask XCO 2 observations.

Conclusions
In this study, the tow GOSAT and OCO-2 XCO 2 observations values are compared with NOAA CT XCO 2 and available ground based flask observations over Africa land mass. Comparison between GOSAT and CT2016 were done using a five The monthly average time series of CT2016 over North Africa, Equatorial Africa and Southern Africa are separately compared with XCO 2 from the two satellites. CT2016 agrees well with measurements from the two instruments in terms of pattern and amplitude. However, this agreement deteriorates over Equatorial and Southern Africa in terms of amplitude. It is also found that there is a seasonal dependent bias between them which is negative during dry seasons while it is positive during wet sea- 20 sons. This indicates results of CT2016 are mostly lower than the GOSAT observation during dry seasons. High spatial mean of seasonal mean RMSD of 1.91 during DJF and 1.75 ppm during MAM and low RMSD of 1.00 and 1.07 ppm during SON in the model XCO 2 with respect to GOSAT and OCO-2 are observed respectively thereby indicating better agreement between CT and the satellites during autumn. CT2016 has the ability to capture monthly time series and seasonal cycles. However, XCO 2 from CT2016 is lower than GOSAT observations over North Africa during all seasons whereas XCO 2 from CT2016 is higher than that of GOSAT over Equatorial and Southern Africa with the exceptions of DJF over Equatorial Africa and SON over 5 Southern Africa. In addition, CT2016 simulates lower XCO 2 than the observations over some regions (e.g., Congo, South Sudan and southwestern Ethiopia) and during summer season over the whole continent following large vegetation uptake. In contrast, XCO 2 from CT16NRT17 is higher than that of OCO-2 over North Africa whereas it is lower than that of OCO-2 during DJF and SON over Equatorial and Southern Africa respectively. Comparison of satellite and CT with ground-based flask observation shows CT has a better performance over sites located at high altitude (IZO, ASK) as determined from good agree-10 ment with flask XCO 2 observations where satellite observations underestimates XCO 2 . Conversely, satellite observations have better performance over low altitude sites (ASC and SEY).
In general, XCO 2 from NOAA CT shows a very small bias with respect to GOSAT and OCO-2 observation over Africa's land mass. Moreover, there is a good agreement between CT simulation and observations in terms spatial distribution, monthly average time series and seasonal climatology. However, there are some discrepancies between the model and the two XCO 2 15 datasets from GOSAT and OCO-2 implying that the accuracy of the model data needs further improvements for the rain forest regions (e.g., Congo) through assimilation of in-situ observations and tuning of the model through process studies.