Seasonally varied controls of climate and phenophase on terrestrial carbon dynamics: modeling eco-climate system state using Dynamical Process Networks

Abstract

Context Prediction of climate impacts on terrestrial ecosystems is limited by the complexity of the couplings between biosphere and atmosphere-what we define here as eco-climate. Critical transitions in ecosystem function and structure must be conceptualized, modeled, and ultimately predicted. Eco-climate system macrostate is a pattern of physical couplings between subsystems; each macrostate must be modeled differently because different physical processes are important. Critical transitions are less likely where the elasticity of macrostate is weak or absent. This motivates a fundamentally new complex systems approach. Objective To model eco-climate macrostate, and its elasticity to seasonal climate forcing (air temperature and precipitation) and ecosystem biophysical state (phenophase). Methods This Dynamical Process Network approach uses information flow to model an ecoclimate system structure using timeseries observations from seven eddy-covariance tower sites in the United States. An aggregate power-law model estimates the elasticity of each location's macrostate to seasonal climate and phenophase. Results Macrostate varies by both season and ecosystem type. Evergreen forests are highly elastic to air temperature and are more likely than agricultural or deciduous systems to experience state changes as the climate warms. Precipitation and phenophase elasticity is stronger in some agricultural, grassland, and deciduous forest systems. Conclusions Different empirical model structures are needed based on season and location, to simulate Special issue: Macrosystems ecology: Novel methods and new understanding of multi-scale patterns and processes. have contributed equally to the work. Electronic supplementary materialThe online version of this article (ecosystem carbon dynamics and critical state transitions. Phenophase directly controls macrostate in some ecosystems. Flux data co-located with in situ ecological monitoring are essential for eco-climate model development and prediction using complex systems approaches.

Direct and Indirect Effects of Tides on Ecosystem‐Scale CO2 Exchange in a Brackish Tidal Marsh in Northern California

“…Since nonlinear estimators can be more powerful than linear estimators, information metrics that consider lags are well suited to analyze biophysical drivers of fluxes (Sturtevant et al, ). The robustness of information metrics to nonlinearity and asynchrony has already improved our understanding of the sensitivity and dynamics of biosphere‐atmosphere exchange to climate variability (Kumar & Ruddell, ; Ruddell et al, ), and the important processes and timescales influencing wetland methane exchange (Sturtevant et al, ).…”

Robust observations of land-to-atmosphere feedbacks using the information flows of FLUXNET

“…11 Information flow statistics are a robust and mature method for delineating PNs, and have been previously applied to the direct and explicit measurement of feedback between the land surface and atmosphere using flux tower observations. [35][36][37][38][39] PNs have been shown to accurately diagnose interactions between turbulent fluxes and the atmosphere in ecohydrological systems, [35][36][37]40 and have accurately described functional differences between starkly diverse land surface ecosystems at continental scales. 35,41,42 This paper's choice of Transfer Entropy to delineate PNs 43 is ideal to measure directional, scale-specific, and nonlinear couplings that characterize land-to-atmosphere feedbacks.…”

Identifying scale‐emergent, nonlinear, asynchronous processes of wetland methane exchange

“…Entropy metrics incorporating joint probabilities, such as mutual information and transfer entropy, quantify information overlap or flow between systems and indicate the dependency of one variable on another without assuming the analytical form or timing of the relationship. The robustness of information metrics to nonlinearity and asynchrony has already improved our understanding of the feedback network of processes influencing CO 2 and H 2 O exchange across ecosystems [ Ruddell and Kumar , , ] and the sensitivity and dynamics of biosphere‐atmosphere exchange to climate variability [ Kumar and Ruddell , ; Ruddell et al ., ]. Despite these benefits, combining wavelets and information theory to understand multiscale eco‐atmosphere interactions is rare, and thus far has only been applied to terrestrial upland sites [ Brunsell and Anderson , ; Brunsell and Wilson , ; Brunsell et al ., ].…”