Optimization of ecosystem model parameters through assimilating eddy covariance flux data with an ensemble Kalman filter

A novel hybrid modelling framework for GPP estimation: Integrating a multispectral surface reflectance based Vcmax25 simulator into the process-based model

Terrestrial gross primary productivity (GPP) is the key element in the carbon cycle process. Accurate GPP estimation hinges on the maximum carboxylation rate (Vcmax,025). The high uncertainty in deriving ecosystem-level Vcmax,025 has long hampered efforts toward the performance of the GPP model. Recently studies suggest the strong relationship between spectral reflectance and Vcmax,025. We proposed the multispectral surface reflectance-driven Vcmax,025 simulator using the fully connected deep neural network and built the hybrid modelling framework for GPP estimation by integrating the data-driven Vcmax,025 simulator in the process-based model. The performance of hybrid GPP model was evaluated at 95 flux sites. The result shows that the multispectral surface reflectance-driven Vcmax,025 simulator acquires the satisfactory estimation, with correlation coefficient (R), root mean square error (RMSE) and median absolute percentage error (MdAPE) ranging from 0.34 to 0.80, 14 to 43 μmol m-2 s-1 and 21 % to 66 % across different land cover types, respectively. The hybrid framework generates good GPP estimates with R, RMSE and MdAPE varying from 0.76 to 0.89, 1.79 to 6.16 μmol m-2 s-1 and 27 % to 90 %, respectively. Compared with EVI-driven method, the multispectral surface reflectance significantly improves the Vcmax,025 and GPP estimates, with MdAPE declining by 0.6 %-18 % and 1 % to 21 %, respectively. The Shapley value analysis reveals that red (620-670 nm), near-infrared (841-876 nm) and shortwave infrared (1628-1652 nm and 2105-2155 nm) are the key bands for Vcmax,025 estimation. This study highlights the potential of multispectral surface reflectance for quantifying ecosystem-level Vcmax,025. The new hybrid framework fully extracts the information of all available spectral bands using deep learning to reduce parameter uncertainty while maintains the description of photosynthetic process to ensure its physical reasonability. It can serve as a powerful tool for accurate global GPP estimation.

Spatiotemporal Evolution of the Carbon Fluxes from Bamboo Forests and their Response to Climate Change Based on a BEPS Model in China

Carbon flux is the main basis for judging the carbon source/sink of forest ecosystems. Bamboo forests have gained much attention because of their high carbon sequestration capacity. In this study, we used a boreal ecosystem productivity simulator (BEPS) model to simulate the gross primary productivity (GPP) and net primary productivity (NPP) of bamboo forests in China during 2001–2018, and then explored the spatiotemporal evolution of the carbon fluxes and their response to climatic factors. The results showed that: (1) The simulated and observed GPP values exhibited a good correlation with the determination coefficient (R2), root mean square error (RMSE), and absolute bias (aBIAS) of 0.58, 1.43 g C m−2 day−1, and 1.21 g C m−2 day−1, respectively. (2) During 2001–2018, GPP and NPP showed fluctuating increasing trends with growth rates of 5.20 g C m−2 yr−1 and 3.88 g C m−2 yr−1, respectively. The spatial distribution characteristics of GPP and NPP were stronger in the south and east than in the north and west. Additionally, the trend slope results showed that GPP and NPP mainly increased, and approximately 30% of the area showed a significant increasing trend. (3) Our study showed that more than half of the area exhibited the fact that the influence of the average annual precipitation had positive effects on GPP and NPP, while the average annual minimum and maximum temperatures had negative effects on GPP and NPP. On a monthly scale, our study also demonstrated that the influence of precipitation on GPP and NPP was higher than that of the influence of temperature on them.

Assimilating spatiotemporal MODIS LAI data with a particle filter algorithm for improving carbon cycle simulations for bamboo forest ecosystems

Bamboo forests are an important part of the forest ecosystem, which has strong carbon sequestration potential and plays an important role in the global carbon cycle. As a key parameter for simulating the carbon cycle using forest ecosystem models, the quality of leaf area index (LAI) data has a direct influence on the accuracy of modelling results. Here, we used the particle filter (PF) algorithm and PROSAIL model to assimilate MODIS LAI products, which were then used to drive a boreal ecosystem productivity simulator model to simulate the bamboo forest carbon cycle. The results showed that the relationship between the assimilated and observed LAI values was very significant, with an R² of 0.95 and an RMSE of 0.28, greatly improving the precision of MODIS LAI products. The R² values for the gross primary productivity (GPP), net ecosystem exchange (NEE), and total ecosystem respiration (TER) simulated by the assimilated LAI values and observed carbon fluxes were 0.65, 0.45 and 0.70, respectively, and the RMSE values were 1.10 g C m^-2 day^-1, 1.00 g C m^-2 day^-1 and 0.35 g C m^-2 day^-1, respectively. Compared with the results of the carbon cycle simulated by non-assimilated LAI, the R² values of the GPP, NEE and TER values that were simulated by assimilated LAI increased by 27.5%, 45.2% and 6.1%, and the RMSE values decreased by 29.9%, 23.7% and 22.2%, respectively. Therefore, coupling the PF and PROSAIL models can greatly improve the simulation precision for the large-scale bamboo forest carbon cycle. This study laid the foundation for simulating the carbon cycle over a large-scale bamboo forest based on low-resolution data in the future.

Diverse photosynthetic capacity of global ecosystems mapped by satellite chlorophyll fluorescence measurements

Photosynthetic capacity is often quantified by the Rubisco-limited photosynthetic capacity (i.e. maximum carboxylation rate, V_cmax). It is a key plant functional trait that is widely used in Earth System Models for simulation of the global carbon and water cycles. Measuring V_cmax is time-consuming and laborious; therefore, the spatiotemporal distribution of V_cmax is still poorly understood due to limited measurements of V_cmax. In this study, we used a data assimilation approach to map the spatial variation of V_cmax for global terrestrial ecosystems from a 11-year-long satellite-observed solar-induced chlorophyll fluorescence (SIF) record. In this SIF-derived V_cmax map, the mean V_cmax value for each plant function type (PFT) is found to be comparable to a widely used N-derived V_cmax dataset by Kattge et al. (2009). The gradient of V_cmax along PFTs is clearly revealed even without land cover information as an input. Large seasonal and spatial variations of V_cmax are found within each PFT, especially for diverse crop rotation systems. The distribution of major crop belts, characterized with high V_cmax values, is highlighted in this V_cmax map. Legume plants are characterized with high V_cmax values. This V_cmax map also clearly illustrates the emerging soybean revolution in South America where V_cmax is the highest among the world. The gradient of V_cmax in Amazon is found to follow the transition of soil types with different soil N and P contents. This study suggests that satellite-observed SIF is powerful in deriving the important plant functional trait, i.e. V_cmax, for global climate change studies.

Estimating carbon fixation of plant organs for afforestation monitoring using a process-based ecosystem model and ecophysiological parameter optimization

Afforestation projects for mitigating CO2 emissions require to monitor the carbon fixation and plant growth as key indicators. We proposed a monitoring method for predicting carbon fixation in afforestation projects, combining a process-based ecosystem model and field data and addressed the uncertainty of predicted carbon fixation and ecophysiological characteristics with plant growth. Carbon pools were simulated using the Biome-BGC model tuned by parameter optimization using measured carbon density of biomass pools on an 11-year-old Eucommia ulmoides plantation on Loess Plateau, China. The allocation parameters fine root carbon to leaf carbon (FRC:LC) and stem carbon to leaf carbon (SC:LC), along with specific leaf area (SLA) and maximum stomatal conductance (g smax) strongly affected aboveground woody (AC) and leaf carbon (LC) density in sensitivity analysis and were selected as adjusting parameters. We assessed the uncertainty of carbon fixation and plant growth predictions by modeling three growth phases with corresponding parameters: (i) before afforestation using default parameters, (ii) early monitoring using parameters optimized with data from years 1 to 5, and (iii) updated monitoring at year 11 using parameters optimized with 11-year data. The predicted carbon fixation and optimized parameters differed in the three phases. Overall, 30-year average carbon fixation rate in plantation (AC, LC, belowground woody parts and soil pools) was ranged 0.14-0.35 kg-C m-2 y-1 in simulations using parameters of phases (i)-(iii). Updating parameters by periodic field surveys reduced the uncertainty and revealed changes in ecophysiological characteristics with plant growth. This monitoring method should support management of afforestation projects by carbon fixation estimation adapting to observation gap, noncommon species and variable growing conditions such as climate change, land use change.

Influence of Snow on the Magnitude and Seasonal Variation of the Clumping Index Retrieved from MODIS BRDF Products

The foliage Clumping Index (CI) is an important vegetation structure parameter that allows for the accurate separation of sunlit and shaded leaves in a canopy. The CI and its seasonality are critical for global Leaf Area Index (LAI) estimating and ecological modelling. However, the cover of snow tends to reduce the reflectance anisotropy of the vegetation canopy and thus probably influences CI estimates. In this paper, we investigate the influence of snow on the magnitude and seasonal variation of the CI retrieved from Moderate-resolution Imaging Spectroradiometer (MODIS) Bidirectional Reflectance Distribution Function (BRDF) products based on field-measured CI and statistics from the global MODIS CI product. We find that the backup algorithm can effectively correct abnormally large CI values and obtain more reasonable CI retrievals than the main algorithm without any constraints in snow-covered areas. Validation indicates that the time-series CI product shows the potential in investigating the trajectories of the clumping effect in snow seasons. For evergreen forests, the clumping effect is relatively stable throughout the year; however, for deciduous vegetation types, CI values tend to display significant seasonal variations. This study suggests that the latest version of the global MODIS CI product, in which the backup algorithm is used to process the snow-covered pixels, has improved accuracy for CI retrievals in snow-covered areas and thus is probably more suitable as the input parameter for ecological and meteorological models.

Probing the limits of predictability: data assimilation of chaotic dynamics in complex food webs

The daunting complexity of ecosystems has led ecologists to use mathematical modelling to gain understanding of ecological relationships, processes and dynamics. In pursuit of mathematical tractability, these models use simplified descriptions of key patterns, processes and relationships observed in nature. In contrast, ecological data are often complex, scale-dependent, space-time correlated, and governed by nonlinear relations between organisms and their environment. This disparity in complexity between ecosystem models and data has created a large gap in ecology between model and data-driven approaches. Here, we explore data assimilation (DA) with the Ensemble Kalman filter to fuse a two-predator-two-prey model with abundance data from a 2600+ day experiment of a plankton community. We analyse how frequently we must assimilate measured abundances to predict accurately population dynamics, and benchmark our population model's forecast horizon against a simple null model. Results demonstrate that DA enhances the predictability and forecast horizon of complex community dynamics.

Confronting terrestrial biosphere models with forest inventory data

Efforts to test and improve terrestrial biosphere models (TBMs) using a variety of data sources have become increasingly common. Yet, geographically extensive forest inventories have been under-exploited in previous model-data fusion efforts. Inventory observations of forest growth, mortality, and biomass integrate processes across a range of timescales, including slow timescale processes such as species turnover, that are likely to have important effects on ecosystem responses to environmental variation. However, the large number (thousands) of inventory plots precludes detailed measurements at each location, so that uncertainty in climate, soil properties, and other environmental drivers may be large. Errors in driver variables, if ignored, introduce bias into model-data fusion. We estimated errors in climate and soil drivers at U.S. Forest Inventory and Analysis (FIA) plots, and we explored the effects of these errors on model-data fusion with the Geophysical Fluid Dynamics Laboratory LM3V dynamic global vegetation model. When driver errors were ignored or assumed small at FIA plots, responses of biomass production in LM3V to precipitation and soil available water capacity appeared steeper than the corresponding responses estimated from FIA data. These differences became nonsignificant if driver errors at FIA plots were assumed to be large. Ignoring driver errors when optimizing LM3V parameter values yielded estimates for fine-root allocation that were larger than biometric estimates, which is consistent with the expected direction of bias. To explore whether complications posed by driver errors could be circumvented by relying on intensive study sites where driver errors are small, we performed a power analysis. To accurately quantify the response of biomass production to spatial variation in mean annual precipitation within the eastern United States would require at least 40 intensive study sites, which is larger than the number of sites typically available for individual biomes in existing plot networks. Driver errors may be accommodated by several existing model-data fusion approaches, including hierarchical Bayesian methods and ensemble filtering methods; however, these methods are computationally expensive. We propose a new approach, in which the TBM functional response is fit directly to the driver-error-corrected functional response estimated from data, rather than to the raw observations.

Assimilation of multiple data sets with the ensemble Kalman filter to improve forecasts of forest carbon dynamics

The ensemble Kalman filter (EnKF) has been used in weather forecasting to assimilate observations into weather models. In this study, we examine how effectively forecasts of a forest carbon cycle can be improved by assimilating observations with the EnKF. We used the EnKF to assimilate into the terrestrial ecosystem (TECO) model eight data sets collected at the Duke Forest between 1996 and 2004 (foliage biomass, fine root biomass, woody biomass, litterfall, microbial biomass, forest floor carbon, soil carbon, and soil respiration). We then used the trained model to forecast changes in carbon pools from 2004 to 2012. Our daily analysis of parameters indicated that all the exit rates were well constrained by the EnKF, with the exception of the exit rates controlling the loss of metabolic litter and passive soil organic matter. The poor constraint of these two parameters resulted from the low sensitivity of TECO predictions to their values and the poor correlation between these parameters and the observed variables. Using the estimated parameters, the model predictions and observations were in agreement. Model forecasts indicate 15 380-15 660 g C/ m2 stored in Duke Forest by 2012 (a 27% increase since 2004). Parameter uncertainties decreased as data were sequentially assimilated into the model using the EnKF. Uncertainties in forecast carbon sinks increased over time for the long-term carbon pools (woody biomass, structure litter, slow and passive SOM) but remained constant over time for the short-term carbon pools (foliage, fine root, metabolic litter, and microbial carbon). Overall, EnKF can effectively assimilate multiple data sets into an ecosystem model to constrain parameters, forecast dynamics of state variables, and evaluate uncertainty.

Ecological forecasting and data assimilation in a data-rich era

Several forces are converging to transform ecological research and increase its emphasis on quantitative forecasting. These forces include (1) dramatically increased volumes of data from observational and experimental networks, (2) increases in computational power, (3) advances in ecological models and related statistical and optimization methodologies, and most importantly, (4) societal needs to develop better strategies for natural resource management in a world of ongoing global change. Traditionally, ecological forecasting has been based on process-oriented models, informed by data in largely ad hoc ways. Although most ecological models incorporate some representation of mechanistic processes, today's models are generally not adequate to quantify real-world dynamics and provide reliable forecasts with accompanying estimates of uncertainty. A key tool to improve ecological forecasting and estimates of uncertainty is data assimilation (DA), which uses data to inform initial conditions and model parameters, thereby constraining a model during simulation to yield results that approximate reality as closely as possible. This paper discusses the meaning and history of DA in ecological research and highlights its role in refining inference and generating forecasts. DA can advance ecological forecasting by (1) improving estimates of model parameters and state variables, (2) facilitating selection of alternative model structures, and (3) quantifying uncertainties arising from observations, models, and their interactions. However, DA may not improve forecasts when ecological processes are not well understood or never observed. Overall, we suggest that DA is a key technique for converting raw data into ecologically meaningful products, which is especially important in this era of dramatically increased availability of data from observational and experimental networks.