Evaluation of atmosphere-biosphere exchange estimations with TCCON measurements

Introduction Conclusions References


Introduction
Understanding, quantifying and predicting the atmospheric carbon cycle is a challenging task, since global transport, carbon fluxes due to fossil fuel emissions, oceanatmosphere exchange, and biosphere-atmosphere exchange must all be known.Thus, an accurate estimation of carbon fluxes is a central goal of the carbon cycle community.Such estimates have been derived on small spatial scales using direct measures of the fluxes via eddy-covariance (e.g., http://fluxnet.ornl.gov/)and by carbon stock analysis (e.g., Gaudinski et al., 2000;Goodale et al., 2002).On larger spatial scales, inverse or "top down" methods have been attempted using measurements of the spatial and temporal variations in atmospheric CO 2 concentrations.Typically, these inverse studies use "Bayesian" methods where "a priori" estimates are combined with atmospheric observations and an atmospheric transport model.The "a priori" estimates represent the best knowledge of the global flux distribution (e.g., Baker et al., 2006;Peters et al., Figures Back Close Full flux estimates are the optimal estimates as determined by the assigned error covariance based on the assumed a priori distribution, the observations and the atmospheric transport model.In inverse methods, unless the design of the inverse machinery is carefully constructed, errors in the a priori distribution at one spatial scale can alias into errors in the inferred distribution at other spatial scales.In particular, developers must make choices about the spatial and temporal scale at which they retrieve fluxes.In theory, such choices are determined by the scales that drive the variance in the observations, but in practice, computational considerations may limit the resolution of the model. The substantial impact of synoptic scale weather systems (e.g., 3-10 days) in driving local variability in atmospheric CO 2 has been illustrated in recent studies (e.g., Keppel-Aleks et al., 2011;Parazoo et al., 2008).Meridional advection produces significant local variability in atmospheric CO 2 during the Northern Hemisphere summertime, when there are strong north-south gradients in CO 2 .Such variance is driven at hemispheric scales, and so the inverse method must extend to global scales.Traditionally, atmospheric inverse modeling has been based on a global network of in situ boundary layer measurement stations.Hence, large-scale errors in the a priori distribution, like an incorrect north-south CO 2 gradient, can alias into errors at local scale around the in situ boundary layer measurement stations in the optimization, generally yielding local/regional flux variability that is too large.Thus, accurate large-scale fluxes are critical for estimating accurate local fluxes, because errors in the description of large-scale flux patterns will alias into the retrieved regional scale fluxes.
Total column measurements are expected to improve the constraint on carbon cycle processes (Rayner and O'Brien, 2001;Yang et al., 2007).These data are particularly helpful for this evaluation because variations in total column measurements are dominated by hemispheric flux distributions, and local and regional fluxes have only a minor impact (Keppel-Aleks et al., 2012).Hence, total column measurements provide a largely independent piece of information to in situ boundary layer measurements, which are mostly driven by local influences.2012) estimated the north-south CO 2 gradient using total column measurements from the Total Carbon Column Observing Network (TCCON) by correlating the CO 2 abundances to the potential temperature, which serves as a dynamical tracer for synoptic-scale dynamics.Additionally, the authors showed that the seasonal CO 2 amplitude seen in total column measurements is dominated by the net ecosystem exchange (NEE) in the boreal forest and the temporal phase of the uptake.The CO 2 fields were simulated with a general circulation model (GCM) and the NEE was estimated by the CASA model.In a sensitivity study, the NEE was enhanced by 40 % in the boreal forest and the onset of the growing season was shifted earlier.These changes significantly improved the comparison of the simulation with the total column measurements.
Here, we evaluate three a priori NEE flux distributions using the GEOS-Chem global three-dimensional (3-D) chemical transport model (CTM) driven by year-specific meteorological input data.The net ecosystem exchange (NEE) is defined in this work as follows: a net CO 2 flux from the ecosystem to the atmosphere is positive and referred to as net CO 2 release.A net CO 2 flux from the atmosphere to the ecosystem is negative and referred to as net CO 2 uptake.We analyze the following distinct atmospherebiosphere exchange inventories: the Carnegie-Ames-Stanford Approach (CASA, Olsen and Randerson, 2004, described in Sect. 3), the Simple Biosphere model (SiB, Baker et al., 2003, described in Sect. 4) and the GBiome-BGC model (Trusilova and Churkina, 2008, described in Sect. 5).The CO 2 total column abundances from these model runs are compared with measured columns from the TCCON (Sect.7).The GEOS-Chem model and TCCON observations are described in Sect. 2 and 6, respectively.Additionally, we investigate a simulation with CASA, but with uptake enhanced in the boreal forest and with the onset of the growing season shifted according to Keppel-  (Bey et al., 2001).Estimates of CO 2 fluxes due to fossil fuel emissions, ocean-atmosphere exchange and biosphere-atmosphere exchange are provided by inventories and atmospheric inverse models.In the standard version of the GEOS-Chem CO 2 simulation, described by Nassar et al. (2010), CASA is used to estimate the balanced atmosphere-biosphere exchange (Olsen and Randerson, 2004).
In this study, we use GEOS-Chem version v9-01-01 with the GEOS-5 fields, and a spatial resolution of 2 In GEOS-Chem the NEE consists of two components: the first component is the net yearly uptake, based on the TransCom climatology and approximated by −5.29 PgC per year (Baker et al., 2006).The second component is the NEE disregarding the net yearly CO 2 uptake, the balanced NEE, driving the seasonal CO 2 cycle.This balanced NEE is the focus of this study and in the following sections referred to as NEE.It will be approximated with three different biosphere models, described in the following sections.

CASA
The Carnegie-Ames-Stanford Approach (CASA) model is the standard biosphere model input in GEOS-Chem CO 2 simulations.Three-hourly net ecosystem production (NEP) fields are computed from the difference between the gross primary production (GPP) and the respiration R e .Monthly GPP data with a 1 • × 1 • (latitude × longitude) spatial resolution are defined as two times the net primary production (NPP) derived Introduction

Conclusions References
Tables Figures

Back Close
Full with the CASA model and scaled to 5.5 • × 5.5 • grid boxes.The monthly GPP values are distributed with shortwave radiation flux data from the National Center for Environmental Prediction (NCEP, Kalnay et al., 1996) data assimilation model for the year 2000 to 3-hourly values.Monthly respiration R e data are calculated with NCEP temperature data for the year 2000 at 5.5 • × 5.5 • grid boxes and also interpolated to 3-h intervals (Olsen and Randerson, 2004;Potter et al., 1993).
The GEOS-Chem CO 2 simulation uses the CASA NEE interpolated to the 2 • × 2.5 • (latitude × longitude) GEOS-Chem grid.Hence, the standard NEE is based on data derived for the year 2000, and GEOS-Chem does not account for any interannual variability, for instance due to droughts or fire.

SiB
The Simple Biosphere model (SiB) parameterizes land surface biophysical processes and ecosystem metabolism (Sellers et al., 1986(Sellers et al., , 1996;;Denning et al., 1996).We use 3-hourly reanalysis data of air temperature, pressure, humidity, wind speed, radiation and precipitation from the Modern-Era Retrospective analysis for Research Applications (MERRA) (Rienecker et al., 2011) to drive the model for years 2006 through 2010.Model parameters are determined using a combination of satellite data, literature values and standard SiB parameters (Sellers et al., 1996).The SiB surface fluxes are calculated at 1 • × 1.25 • (latitude × longitude) spatial resolution, saved as three-hour averages and scaled to the 2 • × 2.5 • (latitude × longitude) GEOS-Chem grid.By using MERRA data, SiB accounts for interannual variability, which is in contrast to the CASA NEE estimations.Further details on the SiB NEE simulations are given in Parazoo et al. (2008) 2).Therefore, the GEOS-Chem CO 2 simulations using the GBiome-BGC NEE estimations are detrended to compensate for the net yearly uptake.
In order to compare the GEOS-Chem CO 2 profile data with the TCCON data, they have to be integrated to column-averaged CO 2 dry-air mole fractions.We do this by applying the TCCON averaging kernels and a priori profiles to the model, employing the method developed by Rodgers and Connor (2003).For each TCCON measurement, the daily averaged GEOS-Chem CO 2 simulation profile for the same day was smoothed with the averaging kernel and a priori profile from the TCCON measurement and integrated to column averaged X CO 2 ,model .For the integration we use the GFIT a priori pressure, altitude, temperature and H 2 O profile, which are the NCEP data interpolated to the location of the TCCON station and to local noon (Wunch et al., 2011a) • and 75 • N) is evident in all three models.Nevertheless, the largest difference between the models can also be found in this region.Both SiB and GBiome-BGC exhibit a sink larger than CASA by up to 40 %.As GBiome-BGC is not balanced and a large portion of the sink can be attributed to the net yearly uptake of −0.705 Pg yr −1 , the sink in SiB and GBiome-BGC is dissimilar.The seasonal CO 2 cycle is mostly dominated by the NEE in the boreal forest and the biggest differences (up to 40 %) between the models are found in this region as well.Thus, our analyses focus on this region.
In the bottom panel, the time series of the monthly NEE integrated over all grid points between 30 • N and 90 • N (<NEE> 30−90,model ) are compared.The time series reflect the differences already seen in the latitudinal NEE distributions: the pronounced summer sink in both SiB and GBiome-BGC leads to a larger seasonal cycle amplitude than in CASA.The winter NEE peak is unique for all three models: In January, CASA shows a dip, in contrast to the maximum in GBiome-GBC and the slightly earlier maximum in SiB.The CO 2 drawdown starts in April in GBiome-BGC and SiB and is shifted one month later in CASA.The autumn release occurs simultaneously in SiB and CASA, and about a month earlier in GBiome-BGC.These differences lead to the widest seasonal cycle minimum in SiB and narrower widths in CASA and GBiome-BGC.CASA lags GBiome-BGC by about a month.
To evaluate the differences in the GEOS-Chem CO 2 simulations using these different NEE inputs, the simulated monthly mean CO 2 between 30 ferences in the NEE estimations (Fig. 1).The drawdown starts about a month earlier using SiB or GBiome-BGC, in contrast to the CASA input.The largest drawdown is found for the SiB NEE and the smallest using the CASA NEE.The growing season is longest for SiB and shortest for CASA.
In the bottom panel, the monthly mean CO 2 at 700 hPa is shown averaged over four TCCON sites: Białystok (Poland), Bremen (Germany), Lamont (Oklahoma), and Park Falls (Wisconsin) (<CO 2 > TCCON,model ).All four TCCON sites lie between 30 • and 90 • N (Table 3).The same differences as described for the CO 2 simulations integrated over nearly the entire Northern Hemisphere (upper panel) can be seen.This implies that studying these differences at the four TCCON sites gives information about the GEOS-Chem CO 2 simulation for nearly the entire Northern Hemisphere.

Evaluating GEOS-Chem CO 2 simulations with TCCON measurements
The differences between the CO 2 simulations using CASA and SiB at the four TC-CON sites are as large as 0.8 % and between CASA and detrended simulations using GBiome-BGC are as large as 0.9 % (Table 4).With a precision better than 0.25 % (∼ 1 ppm), TCCON total CO 2 column measurements are suitable to validate these differences.
In Fig. 3, the monthly averages of the X CO 2 ,model are compared to the mean of the monthly averages of the X CO 2 time series at the four TCCON sites.Comparing the X CO 2 ,model values reveals the same yearly pattern as seen for the GEOS-Chem CO 2 simulations at 700 hPa (Fig. 2).Comparing the X CO 2 ,model with the TCCON measurements reveals an underestimation of the seasonal amplitude for the simulations using GBiome-BGC and CASA and an overestimation by SiB.Using CASA, the start of the growing season is delayed for all years.The start of the growing season in SiB and GBiome-BGC is in relatively good agreement with the TCCON data.
In order to analyze these findings in more detail, the monthly means of the five years were averaged to give a mean seasonal cycle for each NEE input as well as for the TCCON data (Fig. 4).The start of the CO 2 drawdown in spring and the start of the 12768 Introduction

Conclusions References
Tables Figures

Back Close
Full CO 2 release in autumn are estimated by the turning points of the seasonal CO 2 cycle, indicated by dots and dashed lines in in Fig. 4. The delay of the onset and the ending of the growing season are calculated by the time lag between the turning points of the simulated seasonal CO 2 cycle and the turning points of the TCCON time series.For the GEOS-Chem CO 2 simulation using SiB or GBiome-BGC, the CO 2 drawdown starts too early (with a lag of −6 days ± 1 day and −16 days ± 1 day, respectively), whereas the standard CASA NEE inventory leads to a delay in the CO 2 drawdown (by +10 days ± 1 day).In contrast, the CO 2 release is estimated to be too early using the CASA inventory (by −3 days ± 1 day), but is delayed using SiB or GBiome-BGC NEE inputs (by +9 days ± 1 day for both models).The time lags in days are given in Table 6 as well.This estimation of the CO 2 drawdown and release relies only on the turning points.The entire seasonal cycle shape can be evaluated in a cross-correlation of the modeled X CO 2 and the measured TCCON X CO 2 (Fig. 5).The cross-correlation is a measure of the similarity of two waveform patterns as a function of a time shift applied to one waveform.The cross-correlation of the GEOS-Chem CO 2 simulation infers a time shift of −4 days ± 1 day for CASA and +4 days ± 1 day for GBiome-BGC.The simulation using SiB is optimized without shifting.The time shifts in days are also listed in Table 6.
The seasonal amplitude differences are estimated by taking the ratio of the amplitude from the GEOS-Chem CO 2 simulations and the amplitude measured by the TCCON instruments.The amplitude is calculated by the difference between the maximum and the minimum X CO 2 in the seasonal cycle curve.Both CASA and GBiome-BGC simulate amplitudes that are too small by 15 % and 12 %, respectively and the SiB simulation has a seasonal cycle that is too large by 9 % (Table 6).
The GEOS-Chem CO 2 simulation using the SiB model provides the best match to the measured seasonal cycle.The time delay in the CO 2 drawdown is the shortest, at −6 ± 1 days, and the cross-correlation is maximized for the unshifted simulated seasonal cycle.The time delay of the CO 2 release reveals a seasonal cycle minimum that is slightly too wide, but overall the seasonal amplitude matches the measurements well.Introduction

Conclusions References
Tables Figures

Back Close
Full Figure 6 shows the monthly averages of the GEOS-Chem CO 2 simulation using the original CASA NEE estimations, as already depicted in Fig. 3, and the GEOS-Chem CO 2 simulation using CASA NEE estimations, manipulated as described above.The seasonal amplitude increased significantly and even overestimates the seasonal CO 2 cycle amplitude measured by the TCCON sites.The onset of the growing season seems to be in agreement with the TCCON measurements.In order to quantify the changes, the data are analyzed in an analogous fashion to the analysis in Sect.7.1.Figure 7, like Fig. 4, shows the averages of the modeled data and the TCCON data for 2006 through 2010.The seasonal amplitude is overestimated by a factor of 1.20 (Table 6).The onset and the time period of the growing season are estimated accurately.The values of −1 ± 1 day and +2 ± 1 days for the CO 2 drawdown and release delays are a significant improvement, and the cross-correlation optimization yields an unchanged CO 2 seasonal cycle (Table 6).The calculation of the correlation coefficient between the X CO 2 ,CASA and the TCCON data improved from 0.954 to 0.963 for the manipulated X CO 2 ,CASA .changes in these quantities significantly influence the seasonal CO 2 cycle measured at single locations in the Northern Hemisphere.Hence, the CO 2 distribution on synoptic scales drives local variability in atmospheric total column CO 2 .

Impact of year-specific NEE fluxes on the GEOS-Chem CO 2 simulation
The GEOS-Chem CO 2 simulations with CASA and GBiome-BGC were performed with the same NEE estimates for each year.SiB, however, has year-specific meteorology, and so the SiB NEE changes every year.In order to quantify the difference between the year-specific NEE and the static NEE, a simulation for 2006 through 2010 was calculated using the SiB NEE estimation for the year 2009.This approach gives a measure for the difference between the climatology and year-specific fluxes.
Figure 8 shows the monthly averages of the GEOS-Chem CO 2 simulation using year-specific NEE estimations, as already depicted in Fig. 3, and the GEOS-Chem CO 2 simulation using SiB 2009 NEE estimations for the entire time period.Both simulations show only slight differences.The scatter plot of X CO 2 ,SiB2009NEE , X CO 2 ,SiB and X CO 2 ,CASA against the TCCON data is shown in Fig. 9 and the correlation coefficients are given in Table 7 with 0.971 for year-specific SiB NEE estimates and 0.970 for SiB 2009 NEE estimations.These findings show that the year-specific NEE only slightly improves the agreement with the measured seasonal cycle, suggesting that the CO 2 seasonal cycle is mainly driven by the spatial flux distribution and the atmospheric dynamics.
In summary, the analyses highlight good performance for all three models, with the best fit given by the SiB NEE estimates calculated with the specific yearly meteorology.
8 Improved NEE inventory for the GEOS-Chem CO 2 simulation GEOS-Chem CO 2 simulations using year-specific SiB NEE estimations are a significant improvement compared to simulations with the standard CASA climatology.To illustrate the differences between the standard CASA climatology and the SiB yearly val-Introduction

Conclusions References
Tables Figures

Back Close
Full ues, we follow the method described by Nassar et al. (2010) in comparing the GEOS-Chem CO 2 simulations using SiB with the GLOBALVIEW measurements of the surface CO 2 concentrations and to the individual TCCON time series used in the analyses in Sect.7.

Comparison with TCCON measurements
In Fig. 10 the TCCON X CO 2 time series at the four TCCON sites used in this study, Park Falls (Wisconsin), Lamont (Oklahoma), Bremen (Germany), and Białystok (Poland) are compared with GEOS-Chem CO 2 simulations using CASA and SiB NEE.
The findings from Sect. 7 are evident in the comparisons of the individual time series.The seasonal cycle of the GEOS-Chem CO 2 simulations using SiB estimations fits the data best when comparing the measured and modeled seasonal cycle amplitude and phase.The GEOS-Chem CO 2 simulations using CASA inputs tend to underestimate the CO 2 abundance, especially in the seasonal cycle minimum, and the seasonal cycle phase is often delayed compared to the TCCON measurements.

Comparison with GLOBALVIEW-CO2 data
The GLOBALVIEW-CO2 data (GLOBALVIEW-CO2, 2011) are maintained by the Carbon Cycle Greenhouse Gases Group of the National Oceanic and Atmospheric Administration, Earth System Research Laboratory (NOAA ESRL) within the Cooperative Atmospheric Data Integration Project.They are derived (as described by Masarie and Tans, 1995) from highly precise atmospheric CO 2 measurements and widely used for atmospheric model validation.The GEOS-Chem CO 2 simulations using the standard CASA input and the SiB fluxes are compared to GLOBALVIEW-CO2 data at 30 sampling sites.The sampling sites were chosen so as to cover the whole latitude range (82 • N to 90 • S) and to be comparable to the study by Nassar et al. (2010) In Fig. 11, the GLOBALVIEW-CO2 data are shown in addition to the weekly averages of the simulated model concentrations at the sampling site altitude for the years 2006 to 2010.The seasonal amplitude is always larger using SiB fluxes than using CASA fluxes, and tends to over-and underestimate the seasonal maximum and minimum in the Northern Hemisphere measurements.In the Southern Hemisphere, the use of SiB fluxes leads to simulation of lower CO 2 in contrast to the use of CASA estimations.Figure 12 shows the differences between the model and the GLOBALVIEW-CO2 data for all 30 measurement sites.The mean differences for the CO 2 simulations are 0.92 ppm and 0.61 ppm using CASA and SiB, respectively.

Conclusions
We evaluated three estimations of biosphere fluxes within the chemical transport model GEOS-Chem.Errors in the global CO 2 distribution could be analyzed through comparison with TCCON measurements.The standard GEOS-Chem CO 2 simulation (Nassar et al., 2010) uses CASA NEE to estimate the balanced atmosphere-biosphere exchange.However, we show that the estimate of the CO 2 uptake in the growing season in the boreal forest is underestimated and that the onset of the growing season is delayed using this estimate of biospheric fluxes.By enhancing the CO 2 uptake in the boreal forest and shifting the onset of the growing season earlier, the comparison with TCCON data is significantly improved.Similar to CASA, GBiome-BGC also underestimates the CO 2 uptake in the growing season.SiB shows reasonably good agreement in comparison with the TCCON data.
The accurate estimation of carbon fluxes is crucial for the correct simulation of the carbon cycle.The inconsistency of some atmospheric inverse model results with vertical aircraft profiles and total column measurements shown in recent studies reveal a general problem in inverse estimates of carbon fluxes (e.g., Stephens et al., 2007).local CO 2 concentrations is affected even by variations in the CO 2 distribution on hemispheric scale.This means that atmospheric inverse modeling must extend globally to retrieve fluxes.Additionally, errors on a synoptic scale must be carefully evaluated before retrieving local fluxes.Variations in the total column are a good validation resource for diagnosing errors in the hemispheric scale in the estimates of these fluxes, because they provide information on the largest scales.We suggest that an inverse model designed to retrieve the north-south distribution of the fluxes from total column measurements and local fluxes from in situ surface sampling would be helpful.Introduction

Conclusions References
Tables Figures

Conclusions References
Tables Figures

Back Close
Full    Full   The emerging patterns reveal the characterics already seen in the NEE inputs, as well as in the GEOS-Chem CO 2 simulations at 700 hPa and in the smoothed X CO2 .The simulated CO 2 drawdown using SiB or GBiome-BGC starts starts too early compared to the TCCON measurements and too late using CASA inputs.The seasonal amplitude is slightly overestimated with SiB and underestimated using GBiome-BGC and CASA.The simulated CO 2 release starts too early using CASA and too late using SiB or GBiome-BGC.The crossings with the dashed lines indicate the turning points of the seasonal cycles and give an estimate of the delays in the CO 2 drawdown and release (Table 6).20 Fig. 4. Averaged seasonal cycles, derived with the averages of the monthly means, shown in Fig. 3.The emerging patterns reveal the characterics already seen in the NEE inputs, as well as in the GEOS-Chem CO 2 simulations at 700 hPa and in the smoothed X CO 2 .The simulated CO 2 drawdown using SiB or GBiome-BGC starts starts too early compared to the TCCON measurements and too late using CASA inputs.The seasonal amplitude is slightly overestimated with SiB and underestimated using GBiome-BGC and CASA.The simulated CO 2 release starts too early using CASA and too late using SiB or GBiome-BGC.The crossings with the dashed lines indicate the turning points of the seasonal cycles and give an estimate of the delays in the CO 2 drawdown and release (Table 6).The averaged seasonal cycle using SiB is optimized without a time shift.The time shifts in days are given in Table 6.
21 Fig. 5.The cross-correlations between the averaged seasonal cycles for the three NEE estimation models and the averaged TCCON X CO 2 seasonal cycle (averaged seasonal cycles shown in Fig. 4).The cross-correlation optimizes for a negative time shift for CASA and for a positive time shift for GBiome-BGC.The averaged seasonal cycle using SiB is optimized without a time shift.The time shifts in days are given in Table 6.Introduction

Conclusions References
Tables Figures
22  5). 24 Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Keppel-Aleks et al. ( Aleks et al. (2011) (Sect.7).An improved and year-specific NEE flux inventory for the years 2006-2010 is presented in Sect.8. GEOS-Chem CO 2 simulations using this inventory are compared with GLOBALVIEW-CO2 data (GLOBALVIEW-CO2, 2011) and total column CO 2 measurements at four Northern Hemisphere TCCON sites (Sect.8). is a global 3-D chemical transport model for atmospheric composition driven by meteorological input data from the Goddard Earth Observing System (GEOS) of the NASA Global Modeling and Assimilation Office to simulate global atmospheric composition, including CO 2 Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

(
Oklahoma) and Park Falls (Wisconsin) are used.The Park Falls and Bremen sites have the longest data records, covering the whole time period from 2006 to 2010.Measurements at Lamont started in July 2008 and in Białystok in March 2009.The data density is dependent on whether the TCCON instrument performs measurements automatically (Park Falls, Lamont and Białystok) and on the weather conditions at the site.Larger time periods without measurements indicate major instrumental failures.
Discussion Paper | Discussion Paper | Discussion Paper | 7 GEOS-Chem CO 2 simulations with different NEE estimations The NEE estimations of the three models, CASA, SiB (2009 only) and GBiome-BGC, are shown in Fig. 1.In the upper panel, the latitudinal NEE distributions, integrated for the months May to August, are depicted.The large CO 2 sink in boreal forests (between 30 • and 90 • N at the vertical layer of 700 hPa (<CO 2 > 30−90,model ) is shown in the upper panel of Fig. 2. The CO 2 abundance at 700 hPa represents the free troposphere abundance and is less sensitive to local influences (Keppel-Aleks et al., 2011).The most obvious feature is that the amplitude and phase of the simulated seasonal CO 2 cycle is dominated by the dif-Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | 7.2 GEOS-Chem CO 2 simulation using manipulated CASA NEE estimationsThe comparison of the GEOS-Chem CO 2 simulation using CASA NEE estimations with the TCCON measurements revealed a delay in the start of the growing season and a seasonal amplitude, that was too small (Fig.4).Keppel-Aleks et al. (2012)  demonstrated that a GCM simulation could be significantly improved by enhancing NEE in the boreal forest by 40 % and an earlier onset of the growing season.Here, the CASA NEE was amplified by 40 % between 45• N and 65• N and the onset of the growing season was shifted earlier by adding the NEE in July to the NEE in May between 50 • N and 60 • N, analogous to Keppel-Aleks et al. (2012).The resulting GEOS-Chem CO 2 simulation was detrended by 1.081 Pg yr −1 to account for the increased NEE uptake.
These results are consistent with the findings of Keppel-Aleks et al. (2011): the NEE in the boreal forest dominates the amplitude of the seasonal CO 2 cycle and the onset time of the growing season determines the phase of the seasonal CO 2 cycle.Small Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , which introduces the current version of the GEOS-Chem CO 2 simulation.Discussion Paper | Discussion Paper | Discussion Paper | The inverse machinery must span hemispheric scales, otherwise errors in the inferred distribution at one spatial scale can alias into errors at other spatial scales.Variability in Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Keppel-Aleks, G., Wennberg, P. O., Washenfelder, R. A., Wunch, D., Schneider, T., Toon, G. C., Andres, R. J., Blavier, J.-F., Connor, B., Davis, K. J., Desai, A. R., Messerschmidt, J., Notholt, J., Roehl, C. M., Sherlock, V., Stephens, B. B., Vay, S. A., and Wofsy, S. C.: The imprint of surface fluxes and transport on variations in total column carbon dioxide, Biogeosciences, 9, 875-891, doi:10.5194/bg-9-875-2012,2012.12761, 12766Discussion Paper | Discussion Paper | Discussion Paper | Wunch, D., Toon, G. C., Blavier, J.-F.L., Washenfelder, R. A., Notholt, J., Connor, B. J., Griffith, D. W. T., Sherlock, V., and Wennberg, P. O.: The total carbon column observing network, Philos.T. R. Soc.A, 369, 2087-2112, doi:10.1098/rsta.2010.0240,2011a.12765, 12766 Wunch, D., Wennberg, P. O., Toon, G. C., Connor, B. J., Fisher, B., Osterman, G. B., Frankenberg, C., Mandrake, L., O'Dell, C., Ahonen, P., Biraud, S. C., Castano, R., Cressie, N.Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Fig. 1.Upper panel: Latitudinal NEE distributions integrated for May until August.The sink between 30 • and 75• N reflects the CO 2 uptake in the boreal forest in all three models and the biggest differences between the models are found here as well.CASA has a less distinct sink than SiB and GBiome-BGC, but the large sink in the GBiome-BGC is partially due to the net yearly uptake.Bottom panel: The time series of monthly NEE integrated between 30 • and 90 • N. The pronounced summer sink in both, Sib and GBiome-BGC, can be seen in the seasonal cycle amplitude.The CASA amplitude is smaller and later than both SiB and GBiome-BGC.The SiB and CASA autumn releases occur simultaneously and about a month later than GBiome-BGC.This leads to the widest seasonal cycle minimum in SiB and similar widths for CASA and GBiome-BGC, shifted about a month against each other. 17

Fig. 1 .Fig. 4 .
Fig. 1.Upper panel: Latitudinal NEE distributions integrated for May until August.The sink between 30 • and 75• N reflects the CO 2 uptake in the boreal forest in all three models and the biggest differences between the models are found here as well.CASA has a less distinct sink than SiB and GBiome-BGC, but the large sink in the GBiome-BGC is partially due to the net yearly uptake.Bottom panel: The time series of monthly NEE integrated between 30 • and 90 • N. The pronounced summer sink in both, Sib and GBiome-BGC, can be seen in the seasonal cycle amplitude.The CASA amplitude is smaller and later than both SiB and GBiome-BGC.The SiB and CASA autumn releases occur simultaneously and about a month later than GBiome-BGC.This leads to the widest seasonal cycle minimum in SiB and similar widths for CASA and GBiome-BGC, shifted about a month against each other.

Fig. 5 .
Fig.5.The cross-correlations between the averaged seasonal cycles for the three NEE estimation models and the averaged TCCON X CO2 seasonal cycle (averaged seasonal cycles shown in Figure4).The crosscorrelation optimizes for a negative time shift for CASA and for a positive time shift for GBiome-BGC.The averaged seasonal cycle using SiB is optimized without a time shift.The time shifts in days are given in Table6.

Fig. 6 .
Fig. 6.The time series of the monthly averages of column averaged X CO2 .The NEE was enhanced by 40 % in the boreal forest (45 • N and 65 • N) and the onset of the growing season shifted earlier by adding the July NEE to the May NEE between 50 • N and 60 • N. In comparison with the TCCON measurements, the GEOS-Chem CO 2 simulation improves significantly with these changes.The variability of the TCCON timeseries in the winter of 2007-2008 is due to the few measurements averaged (Table5).

Fig. 6 .Fig. 8 .
Fig. 6.The time series of the monthly averages of column averaged X CO 2 .The NEE was enhanced by 40 % in the boreal forest (45 • N and 65 • N) and the onset of the growing season shifted earlier by adding the July NEE to the May NEE between 50• N and 60• N. In comparison with the TCCON measurements, the GEOS-Chem CO 2 simulation improves significantly with these changes.The variability of the TCCON timeseries in the winter of 2007-2008 is due to the few measurements averaged (Table5).

Fig. 8 .Fig. 9 .Fig. 9 .Fig. 11 .
Fig.8.The same as in Fig.3, showing the monthly mean X CO 2 for the GEOS-Chem CO 2 simulation using year-specific SiB fluxes and using only SiB 2009 NEE estimations for the whole time period.The differences between these GEOS-Chem CO 2 simulations give a measure of the impact of year-specific NEE fluxes in contrast to the climatology, showing only slight differences.The variability of the TCCON timeseries in the winter of 2007-2008 is due to the few measurements averaged (Table5).

Table 4 .
Differences in the NEE estimations of SiB and GBiome-BGC in contrast to the standard CASA inventory (first row) and the differences in the GEOS-Chem CO 2 simulations, integrated between 30 • N and 90 • N (second row) and averaged at four TCCON sites (third row).

Table 5 .
Days of TCCON measurements averaged in the monthly means shown in Figs. 3, 6 and 8.

Table 7 .
Correlation coefficients for GEOS-Chem CO 2 simulations with TCCON measurements for the standard CASA inventory, year-specific SiB fluxes and SiB climatology.