Abstract. Inter-annual variations in the tropical land carbon (C) balance are a dominant component of the global atmospheric CO2 growth rate. Currently, the lack of quantitative knowledge on processes controlling net tropical ecosystem C balance on inter-annual timescales inhibits accurate understanding and projections of land–atmosphere C exchanges. In particular, uncertainty on the relative contribution of ecosystem C fluxes attributable to concurrent forcing anomalies (concurrent effects) and those attributable to the continuing influence of past phenomena (lagged effects) stifles efforts to explicitly understand the integrated sensitivity of a tropical ecosystem to climatic variability. Here we present a conceptual framework – applicable in principle to any land biosphere model – to explicitly quantify net biospheric exchange (NBE) as the sum of anomaly-induced concurrent changes and climatology-induced lagged changes to terrestrial ecosystem C states (NBE = NBECON+NBELAG). We apply this framework to an observation-constrained analysis of the 2001–2015 tropical C balance: we use a data–model integration approach (CARbon DAta-MOdel fraMework – CARDAMOM) to merge satellite-retrieved land-surface C observations (leaf area, biomass, solar-induced fluorescence), soil C inventory data and satellite-based atmospheric inversion estimates of CO2 and CO fluxes to produce a data-constrained analysis of the 2001–2015 tropical C cycle. We find that the inter-annual variability of both concurrent and lagged effects substantially contributes to the 2001–2015 NBE inter-annual variability throughout 2001–2015 across the tropics (NBECON IAV = 80 % of total NBE IAV, r =  0.76; NBELAG IAV = 64 % of NBE IAV, r = 0.61), and the prominence of NBELAG IAV persists across both wet and dry tropical ecosystems. The magnitude of lagged effect variations on NBE across the tropics is largely attributable to lagged effects on net primary productivity (NPP; NPPLAG IAV 113 % of NBELAG IAV, r = −0.93, p value < 0.05), which emerge due to the dependence of NPP on inter-annual variations in foliar C and plant-available H2O states. We conclude that concurrent and lagged effects need to be explicitly and jointly resolved to retrieve an accurate understanding of the processes regulating the present-day and future trajectory of the terrestrial land C sink.

Modeling the net carbon balance is challenging due to the knowledge gaps in the variability and processes controlling gross carbon fluxes. Terrestrial carbon cycle modeling is susceptible to several sources of bias, including meteorological uncertainty, model structural uncertainty, and model parametric uncertainty. To determine the impact of these uncertainties, we compare three model‐derived representations of the global terrestrial carbon balance across 1997–2009: (1) observation‐constrained model‐data fusion (CARBon Data Model FraMEwork, CARDAMOM), (2) the reanalysis‐driven Trends in Net Land‐Atmosphere Carbon Exchange (TRENDY) land biosphere model ensemble, and (3) the Coupled Model Intercomparison Project 5 (CMIP5) Earth System Model ensemble. We consider the spread in carbon cycle simulations attributable primarily to parametric uncertainty (CARDAMOM), structural uncertainty (TRENDY), and combined structural and simulated meteorological uncertainty (CMIP5). We find that the spread across the CARDAMOM ensemble long‐term mean—produced by parameter uncertainty—is larger than the spread of TRENDY and CMIP5 for net biosphere exchange (NBE) but similar for gross primary productivity (GPP). The carbon flux dynamics of CARDAMOM compares to models in TRENDY as well as models in TRENDY compare to each other in many regions for NBE seasonal (nine of 12), NBE interannual (11 of 12), and GPP seasonal variability (7 of 12), although not for GPP interannual variability (2 of 12). The simple model structure of CARDAMOM and systematic assimilation of observations is sufficient to produce carbon dynamics within the range of more complex models. These results are encouraging for the use of model‐data fusion products with empirically estimated uncertainty for global carbon cycle studies.

Abstract. Evapotranspiration (ET) from tropical forests serves as a critical moisture source for regional and global climate cycles. However, the magnitude, seasonality, and interannual variability of ET in the Congo Basin remain poorly constrained due to a scarcity of direct observations, despite the Congo being the second-largest river basin in the world and containing a vast region of tropical forest. In this study, we applied a water balance model to an array of remotely sensed and in situ datasets to produce monthly, basin-wide ET estimates spanning April 2002 to November 2016. Data sources include water storage changes estimated from the Gravity Recovery and Climate Experiment (GRACE) satellites, in situ measurements of river discharge, and precipitation from several remotely sensed and gauge-based sources. An optimal precipitation dataset was determined as a weighted average of interpolated data by Nicholson et al. (2018), Climate Hazards InfraRed Precipitation with Station data version 2 (CHIRPS2) , and the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks–Climate Data Record product (PERSIANN-CDR), with the relative weights based on the error magnitudes of each dataset as determined by triple collocation. The resulting water-balance-derived ET (ETwb) features a long-term average that is consistent with previous studies (117.2±3.5 cm yr−1) but displays greater seasonal and interannual variability than seven global ET products. The seasonal cycle of ETwb generally tracks that of precipitation over the basin, with the exception that ETwb is greater in March–April–May (MAM) than in the relatively wetter September–October–November (SON) periods. This pattern appears to be driven by seasonal variations in the diffuse photosynthetically active radiation (PAR) fraction, net radiation (Rn), and soil water availability. From 2002 to 2016, Rn, PAR, and vapor-pressure deficit (VPD) all increased significantly within the Congo Basin; however, no corresponding trend occurred in ETwb. We hypothesize that the stability of ETwb over the study period despite sunnier and less humid conditions may be due to increasing atmospheric CO2 concentrations that offset the impacts of rising VPD and irradiance on stomatal water use efficiency (WUE).

Abstract. The terrestrial carbon cycle plays a critical role in modulating the interactions of climate with the Earth system, but different models often make vastly different predictions of its behavior. Efforts to reduce model uncertainty have commonly focused on model structure, namely by introducing additional processes and increasing structural complexity. However, the extent to which increased structural complexity can directly improve predictive skill is unclear. While adding processes may improve realism, the resulting models are often encumbered by a greater number of poorly determined or over-generalized parameters. To guide efficient model development, here we map the theoretical relationship between model complexity and predictive skill. To do so, we developed 16 structurally distinct carbon cycle models spanning an axis of complexity and incorporated them into a model–data fusion system. We calibrated each model at six globally distributed eddy covariance sites with long observation time series and under 42 data scenarios that resulted in different degrees of parameter uncertainty. For each combination of site, data scenario, and model, we then predicted net ecosystem exchange (NEE) and leaf area index (LAI) for validation against independent local site data. Though the maximum model complexity we evaluated is lower than most traditional terrestrial biosphere models, the complexity range we explored provides universal insight into the inter-relationship between structural uncertainty, parametric uncertainty, and model forecast skill. Specifically, increased complexity only improves forecast skill if parameters are adequately informed (e.g., when NEE observations are used for calibration). Otherwise, increased complexity can degrade skill and an intermediate-complexity model is optimal. This finding remains consistent regardless of whether NEE or LAI is predicted. Our COMPLexity EXperiment (COMPLEX) highlights the importance of robust observation-based parameterization for land surface modeling and suggests that data characterizing net carbon fluxes will be key to improving decadal predictions of high-dimensional terrestrial biosphere models.

Abstract. The terrestrial carbon cycle plays a critical role in modulating the interactions of climate with the Earth system, but different models often make vastly different predictions of its behavior. Efforts to reduce model uncertainty have commonly focused on model structure, namely by introducing additional processes and increasing structural complexity. However, the extent to which increased structural complexity can directly improve predictive skill is unclear. While adding processes may improve realism, the resulting models are often encumbered by a greater number of poorly-determined or over-generalized parameters. To guide efficient model development, here we map the theoretical relationship between model complexity and predictive skill. To do so, we developed 16 structurally distinct carbon cycle models spanning an axis of complexity and incorporated them into a model–data fusion system. We calibrated each model at 6 globally-distributed eddy covariance sites with long observation time series and under 42 data scenarios that resulted in different degrees of parameter uncertainty. For each combination of site, data scenario, and model, we then predicted net ecosystem exchange (NEE) and leaf area index (LAI) for validation against independent local site data. Though the maximum model complexity we evaluated is lower than most traditional terrestrial biosphere models, the complexity range we explored provides universal insight into the inter-relationship between structural uncertainty, parametric uncertainty, and model forecast skill. Specifically, increased complexity only improves forecast skill if parameters are adequately informed (e.g., when NEE observations are used for calibration). Otherwise, increased complexity can degrade skill and an intermediate-complexity model is optimal. This finding remains consistent regardless of whether NEE or LAI is predicted. Our COMPLexity EXperiment (COMPLEX) highlights the importance of robust, observation-based parameterization for land surface modeling and suggests that data characterizing net carbon fluxes will be key to improving decadal predictions of high-dimensional terrestrial biosphere models.