Bernard Pak

TransCom 3 inversion intercomparison: Impact of transport model errors on the interannual variability of regional CO2 fluxes, 1988–2003

Monthly CO2 fluxes are estimated across 1988–2003 for 22 emission regions using data from 78 CO2 measurement sites. The same inversion (method, priors, data) is performed with 13 different atmospheric transport models, and the spread in the results is taken as a measure of transport model error. Interannual variability (IAV) in the winds is not modeled, so any IAV in the measurements is attributed to IAV in the fluxes. When both this transport error and the random estimation errors are considered, the flux IAV obtained is statistically significant at P ≤ 0.05 when the fluxes are grouped into land and ocean components for three broad latitude bands, but is much less so when grouped into continents and basins. The transport errors have the largest impact in the extratropical northern latitudes. A third of the 22 emission regions have significant IAV, including the Tropical East Pacific (with physically plausible uptake/release across the 1997–2000 El Nino/La Nina) and Tropical Asia (with strong release in 1997/1998 coinciding with large-scale fires there). Most of the global IAV is attributed robustly to the tropical/southern land biosphere, including both the large release during the 1997/1998 El Nino and the post-Pinatubo uptake.

Transcom 3 inversion intercomparison: Model mean results for the estimation of seasonal carbon sources and sinks

[1] The TransCom 3 experiment was begun to explore the estimation of carbon sources and sinks via the inversion of simulated tracer transport. We build upon previous TransCom work by presenting the seasonal inverse results which provide estimates of carbon flux for 11 land and 11 ocean regions using 12 atmospheric transport models. The monthly fluxes represent the mean seasonal cycle for the 1992 to 1996 time period. The spread among the model results is larger than the average of their estimated flux uncertainty in the northern extratropics and vice versa in the tropical regions. In the northern land regions, the model spread is largest during the growing season. Compared to a seasonally balanced biosphere prior flux generated by the CASA model, we find significant changes to the carbon exchange in the European region with greater growing season net uptake which persists into the fall months. Both Boreal North America and Boreal Asia show lessened net uptake at the onset of the growing season with Boreal Asia also exhibiting greater peak growing season net uptake. Temperate Asia shows a dramatic springward shift in the peak timing of growing season net uptake relative to the neutral CASA flux while Temperate North America exhibits a broad flattening of the seasonal cycle. In most of the ocean regions, the inverse fluxes exhibit much greater seasonality than that implied by the DpCO2 derived fluxes though this may be due, in part, to misallocation of adjacent land flux. In the Southern Ocean, the austral spring and fall exhibits much less carbon uptake than implied by DpCO2 derived fluxes. Sensitivity testing indicates that the inverse estimates are not overly influenced by the prior flux choices. Considerable agreement exists between the model mean, annual mean results of this study and that of the previously published TransCom annual mean inversion. The differences that do exist are in poorly constrained regions and tend to exhibit compensatory fluxes in order to match the global mass constraint. The differences between the estimated fluxes and the prior model over the northern land regions could be due to the prior model respiration response to temperature. Significant phase differences, such as that in the Temperate Asia region, may be due to the limited observations for that region. Finally, differences in the boreal land regions between the prior model and the estimated fluxes may be a reflection of the timing of spring thaw and an imbalance in respiration versus photosynthesis. INDEX TERMS: 0322 Atmospheric Composition and Structure: Constituent sources and sinks; 1615 Global Change: Biogeochemical processes (4805); 0315 Atmospheric Composition and Structure: Biosphere/atmosphere interactions; KEYWORDS: carbon transport, inversion

Interannual growth rate variations of atmospheric CO2 and its δ13C, H2, CH4, and CO between 1992 and 1999 linked to biomass burning

[1]xa0High-precision, multispecies measurements of flask air samples since 1992 from CSIROs global sampling network reveal strong correlation among interannual growth rate variations of CO2 and its δ13C, H2, CH4, and CO. We show that a major fraction of the variability is consistent with two emission pulses coinciding with large biomass burning events in 1994/1995 and 1997/1998 in tropical and boreal regions, and observations of unusually high levels of combustion products in the overlying troposphere at these times. Implied pulse strengths and multispecies emission ratios are not consistent with any other single process, but do not exclude possible contributions from covarying processes that are linked through climatic forcing. Comparison of CO2 with its δ13C indicates that most of the CO2 variation is from terrestrial exchange, but does not distinguish forcing by biomass burning from imbalance in photosynthesis/respiration of terrestrial ecosystems. Partitioning of terrestrial CO2 fluxes is constrained by H2, CH4, and CO, all of which are products of biomass burning but which have no direct link to net respiration of CO2. While CO is a strong indicator of biomass burning, its short lifetime prevents it from usefully constraining the magnitude of CO2 emissions. If the H2 and CH4 variations were dominated by biomass burning, they would imply associated carbon emissions in excess of mean annual levels of other years, of 0.6–3.5 and 0.8–3.7 Pg C for 1994/1995 and 1997/1998, respectively. The large range in emission estimates mainly reflects uncertainty in H2/CO2 and CH4/CO2 emission ratios of fires in these years.

OptIC project: An intercomparison of optimization techniques for parameter estimation in terrestrial biogeochemical models

We describe results of a project known as OptIC (Optimisation InterComparison) for comparison of parameter estimation methods in terrestrial biogeochemical models. A highly simplified test model was used to generate pseudo-data to which noise with different characteristics was added. Participants in the OptIC project were asked to estimate the model parameters used to generate this data, and to predict model variables into the future. Ten participants contributed results using one of the following methods: Levenberg-Marquardt, adjoint, Kalman filter, Markov chain Monte Carlo and genetic algorithm. Methods differed in how they locate the minimum (gradient-descent or global search), how observations are processed (all at once sequentially), or the number of iterations used, or assumptions about the statistics (some methods assume Gaussian probability density functions; others do not). We found the different methods equally successful at estimating the parameters in our application. The biggest variation in parameter estimates arose from the choice of cost function, not the choice of optimization method. Relatively poor results were obtained when the model-data mismatch in the cost function included weights that were instantaneously dependent on noisy observations. This was the case even when the magnitude of residuals varied with the magnitude of observations. Missing data caused estimates to be more scattered, and the uncertainty of predictions increased correspondingly. All methods gave biased results when the noise was temporally correlated or non-Gaussian, or when incorrect model forcing was used. Our results highlight the need for care in choosing the error model in any optimization.

Improving the responses of the Australian community land surface model (CABLE) to seasonal drought

[1]xa0Correct representations of root functioning, such as root water uptake and hydraulic redistribution, are critically important for modeling the responses of vegetation to droughts and seasonal changes in soil moisture content. However, these processes are poorly represented in global land surface models. In this study, we incorporated two root functions: a root water uptake function which assumes root water uptake efficiency varies with rooting depth, and a hydraulic redistribution function into a global land surface model, CABLE. The water uptake function developed by Lai and Katul (2000) was also compared with the default one (see Wang et al., 2010) that assumes that efficiency of water uptake per unit root length is constant. Using eddy flux measurements of CO2 and water vapor fluxes at three sites experiencing different patterns of seasonal changes in soil water content, we showed that the two root functions significantly improved the agreement between the simulated fluxes of net ecosystem exchange and latent heat flux and soil moisture dynamics with those observed during the dry season while having little impact on the model simulation during the wet seasons at all three sites. Sensitivity analysis showed that varying several model parameters influencing soil water dynamics in CABLE did not significantly affect the models performance. We conclude that these root functions represent a valuable improvement for land surface modeling and should be implemented into CABLE and other land surface models for studying carbon and water dynamics where rainfall varies seasonally or interannually.

Sensitivity of inverse estimation of annual mean CO2 sources and sinks to ocean-only sites versus all-sites observational networks

[1]xa0Inverse estimation of carbon dioxide (CO2) sources and sinks uses atmospheric CO2 observations, mostly made near the Earths surface. However, transport models used in such studies lack perfect representation of atmospheric dynamics and thus often fail to produce unbiased forward simulations. The error is generally larger for observations over the land than those over the remote/marine locations. The range of this error is estimated by using multiple transport models (16 are used here). We have estimated the remaining differences in CO2 fluxes due to the use of ocean-only versus all-sites (i.e., over ocean and land) observations of CO2 in a time-independent inverse modeling framework. The fluxes estimated using the ocean-only networks are more robust compared to those obtained using all-sites networks. This makes the global, hemispheric, and regional flux determination less dependent on the selection of transport model and observation network.

Measurements of biomass burning influences in the troposphere over southeast Australia during the SAFARI 2000 dry season campaign

[1]xa0Several studies have observed midtropospheric atmospheric composition anomalies and suggested a link to tropical biomass burning. Such anomalies complicate the use of trace gas profiles in remote regions to infer their surface sources/sinks based on the vertical gradients. The Southern African Regional Science Initiative (SAFARI 2000) campaign in Africa and coordinated downwind measurements in Australia provided an opportunity to confirm this link and elucidate the specific surface and atmospheric processes. Five aircraft missions were conducted by Commonwealth Scientific and Industrial Research Organisation (CSIRO) Atmospheric Research during the campaign. They were scheduled after African outflows of polluted air were observed in satellite images over the Indian Ocean flowing east toward Australia. Air samples collected from near the surface to 7 km were analyzed for a suite of trace gases (12CO2, CH4, CO, H2, N2O, and C2 and C3 hydrocarbons) and one isotopomer (13CO2) to provide vertical composition profiles. Ozone was monitored continuously during flight while a ground-based lidar was employed in the Melbourne region to detect aerosol layers. A preliminary statistical analysis on the Australian data confirms covarying midtroposphere enhancements in the biomass burning products. Making rudimentary corrections for photochemical evolution during transit, the trace gas enhancement ratios in affected air samples are comparable to emission ratios in fresh biomass burning plumes. The 13CO2/12CO2 ratios are also consistent with a source from terrestrial plants. Back-trajectory analysis for strongly enhanced samples suggests long-range transport from tropical regions in Africa or South America, the proof of which requires a follow-on analysis with a global chemistry transport model.

Using models to guide field experiments: a priori predictions for the CO2 response of a nutrient- and water-limited native Eucalypt woodland

The response of terrestrial ecosystems to rising atmospheric CO2 concentration (Ca ), particularly under nutrient-limited conditions, is a major uncertainty in Earth System models. The Eucalyptus Free-Air CO2 Enrichment (EucFACE) experiment, recently established in a nutrient- and water-limited woodland presents a unique opportunity to address this uncertainty, but can best do so if key model uncertainties have been identified in advance. We applied seven vegetation models, which have previously been comprehensively assessed against earlier forest FACE experiments, to simulate a priori possible outcomes from EucFACE. Our goals were to provide quantitative projections against which to evaluate data as they are collected, and to identify key measurements that should be made in the experiment to allow discrimination among alternative model assumptions in a postexperiment model intercomparison. Simulated responses of annual net primary productivity (NPP) to elevated Ca ranged from 0.5 to 25% across models. The simulated reduction of NPP during a low-rainfall year also varied widely, from 24 to 70%. Key processes where assumptions caused disagreement among models included nutrient limitations to growth; feedbacks to nutrient uptake; autotrophic respiration; and the impact of low soil moisture availability on plant processes. Knowledge of the causes of variation among models is now guiding data collection in the experiment, with the expectation that the experimental data can optimally inform future model improvements.

Quantification and attribution of errors in the simulated annual gross primary production and latent heat fluxes by two global land surface models

Differences in the predicted carbon and water fluxes by different global land models have been quite large and have not decreased over the last two decades. Quantification and attribution of the uncertainties of global land surface models are important for improving the performance of global land surface models, and are the foci of this study. Here we quantified the model errors by comparing the simulated monthly global gross primary productivity (GPP) and latent heat flux (LE) by two global land surface models with the model-data products of global GPP and LE from 1982 to 2005. By analyzing model parameter sensitivities within their ranges, we identified about 2–11 most sensitive model parameters that have strong influences on the simulated GPP or LE by two global land models, and found that the sensitivities of the same parameters are different among the plant functional types (PFT). Using parameter ensemble simulations, we found that 15%–60% of the model errors were reduced by tuning only a few (<4) most sensitive parameters for most PFTs, and that the reduction in model errors varied spatially within a PFT or among different PFTs. Our study shows that future model improvement should optimize key model parameters, particularly those parameters relating to leaf area index, maximum carboxylation rate, and stomatal conductance.

Linear and nonlinear effects of dominant drivers on the trends in global and regional land carbon uptake: 1959 to 2013

Changes in atmospheric CO2 levels, surface temperature, or precipitation have been identified to have significantly contributed to the estimated increase in the terrestrial carbon uptake rate over the last few decades; however, those analyses did not consider the interactions. Using the Australian community land surface model (Community Atmosphere Biosphere Land Exchange), we performed factorial experiments to quantify the importance of external drivers (climate drivers and atmospheric CO2) and their interactions on annual terrestrial carbon uptake (FL), excluding land use change and fires, from 1959 to 2013. Our model simulations show a trend of 0.025u2009±u20090.015u2009Pgu2009Cu2009yr−2 (or ~1.5%u2009yr−1) in global FL for 1959–2013, which is largely attributed to the positive influences of the increased atmospheric CO2 (0.050u2009±u20090.001u2009Pgu2009Cu2009yr−2) and negative influences of changes in climate (−0.026u2009±u20090.014u2009Pgu2009Cu2009yr−2). Globally, the contribution of the nonlinear effects of dominant drivers to the simulated trend in FL is small ( 35%), particularly in the boreal forests and semiarid regions. The interactions between temperature and CO2 or temperature and precipitation can dominate the simulated trend in parts of Europe, southeastern North America, southern China, and some semiarid regions. This modeling result suggests that the effects of nonlinear interactions of drivers on the trend of land carbon uptake should be considered in future studies.