Memory effects of climate and vegetation affecting net ecosystem CO2 fluxes in global forests

Forests play a crucial role in the global carbon (C) cycle by storing and sequestering a substantial amount of C in the terrestrial biosphere. Due to temporal dynamics in climate and vegetation activity, there are significant regional variations in carbon dioxide (CO2) fluxes between the biosphere and atmosphere in forests that are affecting the global C cycle. Current forest CO2 flux dynamics are controlled by instantaneous climate, soil, and vegetation conditions, which carry legacy effects from disturbances and extreme climate events. Our level of understanding from the legacies of these processes on net CO2 fluxes is still limited due to their complexities and their long-term effects. Here, we combined remote sensing, climate, and eddy-covariance flux data to study net ecosystem CO2 exchange (NEE) at 185 forest sites globally. Instead of commonly used non-dynamic statistical methods, we employed a type of recurrent neural network (RNN), called Long Short-Term Memory network (LSTM) that captures information from the vegetation and climate’s temporal dynamics. The resulting data-driven model integrates interannual and seasonal variations of climate and vegetation by using Landsat and climate data at each site. The presented LSTM algorithm was able to effectively describe the overall seasonal variability (Nash-Sutcliffe efficiency, NSE = 0.66) and across-site (NSE = 0.42) variations in NEE, while it had less success in predicting specific seasonal and interannual anomalies (NSE = 0.07). This analysis demonstrated that an LSTM approach with embedded climate and vegetation memory effects outperformed a non-dynamic statistical model (i.e. Random Forest) for estimating NEE. Additionally, it is shown that the vegetation mean seasonal cycle embeds most of the information content to realistically explain the spatial and seasonal variations in NEE. These findings show the relevance of capturing memory effects from both climate and vegetation in quantifying spatio-temporal variations in forest NEE.

[1]  M. Herold,et al.  Quantifying the effect of forest age in annual net forest carbon balance , 2018, Environmental Research Letters.

[2]  M. Aubinet,et al.  Inter-annual variability of Net Ecosystem Productivity for a temperate mixed forest: A predominance of carry-over effects? , 2018, Agricultural and Forest Meteorology.

[3]  Markus Reichstein,et al.  Upscaled diurnal cycles of land–atmosphere fluxes: a new global half-hourly data product , 2018, Earth System Science Data.

[4]  Martin Jung,et al.  Modelling Landsurface Time-Series with Recurrent Neural Nets , 2018, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium.

[5]  Atul K. Jain,et al.  Global Carbon Budget 2018 , 2014, Earth System Science Data.

[6]  Michael Dixon,et al.  Google Earth Engine: Planetary-scale geospatial analysis for everyone , 2017 .

[7]  Marco Körner,et al.  Temporal Vegetation Modelling Using Long Short-Term Memory Networks for Crop Identification from Medium-Resolution Multi-spectral Satellite Images , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[8]  M. Herold,et al.  An empirical spatiotemporal description of the global surface-atmosphere carbon fluxes: opportunities and data limitations , 2016 .

[9]  Markus Reichstein,et al.  Predicting carbon dioxide and energy fluxes across global FLUXNET sites with regression algorithms , 2016 .

[10]  C. Gough,et al.  Disturbance, complexity, and succession of net ecosystem production in North America's temperate deciduous forests , 2016 .

[11]  Susan G. Letcher,et al.  Carbon sequestration potential of second-growth forest regeneration in the Latin American tropics , 2016, Science Advances.

[12]  J. Canadell,et al.  Greening of the Earth and its drivers , 2016 .

[13]  Alfredo R. Huete,et al.  Leaf development and demography explain photosynthetic seasonality in Amazon evergreen forests , 2016, Science.

[14]  W. Shen,et al.  Precipitation legacy effects on dryland ecosystem carbon fluxes: direction, magnitude and biogeochemical carryovers , 2015 .

[15]  Xiaomin Sun,et al.  Lagged climatic effects on carbon fluxes over three grassland ecosystems in China , 2015 .

[16]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[17]  Markus Reichstein,et al.  Effects of climate extremes on the terrestrial carbon cycle: concepts, processes and potential future impacts , 2015, Global change biology.

[18]  Max Kuhn,et al.  caret: Classification and Regression Training , 2015 .

[19]  C. Woodcock,et al.  Improvement and expansion of the Fmask algorithm: cloud, cloud shadow, and snow detection for Landsats 4–7, 8, and Sentinel 2 images , 2015 .

[20]  E. Vivoni,et al.  Legacy effects in linked ecological-soil-geomorphic systems of drylands , 2015 .

[21]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[22]  Yoshua Bengio,et al.  Neural Machine Translation by Jointly Learning to Align and Translate , 2014, ICLR.

[23]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[24]  T. Spies,et al.  Disturbance legacies increase the resilience of forest ecosystem structure, composition, and functioning. , 2014, Ecological applications : a publication of the Ecological Society of America.

[25]  Quoc V. Le,et al.  Sequence to Sequence Learning with Neural Networks , 2014, NIPS.

[26]  Andrew R. Smith,et al.  Can current moisture responses predict soil CO2 efflux under altered precipitation regimes? A synthesis of manipulation experiments , 2014 .

[27]  Nina Buchmann,et al.  NEP of a Swiss subalpine forest is significantly driven not only by current but also by previous year's weather , 2014 .

[28]  P. Ambus,et al.  Corrigendum to 'Can current moisture responses predict soil CO2 efflux under altered precipitation regimes? A synthesis of manipulation experiments' published in Biogeosciences, 11, 2991-3013, 2014 , 2014 .

[29]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[30]  A. Desai.  Influence and predictive capacity of climate anomalies on daily to decadal extremes in canopy photosynthesis , 2014, Photosynthesis Research.

[31]  S. Seneviratne,et al.  Climate extremes and the carbon cycle , 2013, Nature.

[32]  R. DeFries,et al.  Current systematic carbon-cycle observations and the need for implementing a policy-relevant carbon observing system , 2013 .

[33]  B. Duval,et al.  Altered dynamics of forest recovery under a changing climate , 2013, Global change biology.

[34]  Feng Gao,et al.  Landsat Ecosystem Disturbance Adaptive Processing System LEDAPS algorithm description , 2013 .

[35]  R. Monson,et al.  Persistent reduced ecosystem respiration after insect disturbance in high elevation forests , 2013, Ecology letters.

[36]  M. Turner,et al.  Bark beetle effects on fuel profiles across a range of stand structures in Douglas-fir forests of Greater Yellowstone. , 2013, Ecological applications : a publication of the Ecological Society of America.

[37]  Tara N. Sainath,et al.  FUNDAMENTAL TECHNOLOGIES IN MODERN SPEECH RECOGNITION Digital Object Identifier 10.1109/MSP.2012.2205597 , 2012 .

[38]  S. Rambal,et al.  Gas exchange and leaf aging in an evergreen oak: causes and consequences for leaf carbon balance and canopy respiration. , 2012, Tree physiology.

[39]  Zhe Zhu,et al.  Object-based cloud and cloud shadow detection in Landsat imagery , 2012 .

[40]  Jeffrey G. Masek,et al.  Carbon consequences of forest disturbance and recovery across the conterminous United States , 2012 .

[41]  Christopher Potter,et al.  Simulating the impacts of disturbances on forest carbon cycling in North America: processes, data, models, and challenges , 2011 .

[42]  A. Arneth,et al.  Global patterns of land-atmosphere fluxes of carbon dioxide, latent heat, and sensible heat derived from eddy covariance, satellite, and meteorological observations , 2011 .

[43]  Manfred J. Lexer,et al.  Unraveling the drivers of intensifying forest disturbance regimes in Europe , 2011 .

[44]  R. B. Jackson,et al.  A Large and Persistent Carbon Sink in the World’s Forests , 2011, Science.

[45]  S. Seneviratne,et al.  Drought and ecosystem carbon cycling , 2011 .

[46]  J. Thepaut,et al.  The ERA‐Interim reanalysis: configuration and performance of the data assimilation system , 2011 .

[47]  Kenneth L. Clark,et al.  Ecosystem carbon dioxide fluxes after disturbance in forests of North America , 2010 .

[48]  Nuno Carvalhais,et al.  Identification of vegetation and soil carbon pools out of equilibrium in a process model via eddy covariance and biometric constraints , 2010 .

[49]  M. Turner.  Disturbance and landscape dynamics in a changing world. , 2010, Ecology.

[50]  Witold R. Rudnicki,et al.  Feature Selection with the Boruta Package , 2010 .

[51]  F. Woodward,et al.  Terrestrial Gross Carbon Dioxide Uptake: Global Distribution and Covariation with Climate , 2010, Science.

[52]  Daniel C. Donato,et al.  Forest Fire Impacts on Carbon Uptake, Storage, and Emission: The Role of Burn Severity in the Eastern Cascades, Oregon , 2009, Ecosystems.

[53]  Christopher I. Roos,et al.  Fire in the Earth System , 2009, Science.

[54]  D. Baldocchi.  ‘Breathing’ of the terrestrial biosphere: lessons learned from a global network of carbon dioxide flux measurement systems , 2008 .

[55]  D. Schimel.  Carbon cycle conundrums , 2007, Proceedings of the National Academy of Sciences.

[56]  C. Gough,et al.  The legacy of harvest and fire on ecosystem carbon storage in a north temperate forest , 2007 .

[57]  Linda S. Heath,et al.  Carbon sequestration in the U.S. forest sector from 1990 to 2010 , 2007 .

[58]  T. Vesala,et al.  Towards a standardized processing of Net Ecosystem Exchange measured with eddy covariance technique: algorithms and uncertainty estimation , 2006 .

[59]  A. Lucier,et al.  Forest carbon management in the United States: 1600-2100. , 2006, Journal of environmental quality.

[60]  T. Vesala,et al.  On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm , 2005 .

[61]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[62]  S. Gower.  Patterns and Mechanisms of the Forest Carbon Cycle1 , 2003 .

[63]  Daniel M. Kammen,et al.  Annual review of environment and resources , 2003 .

[64]  Peter E. Thornton,et al.  Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests , 2002 .

[65]  W. Keeton,et al.  Disturbances and structural development of natural forest ecosystems with silvicultural implications, using Douglas-fir forests as an example , 2002 .

[66]  W. Oechel,et al.  FLUXNET: A New Tool to Study the Temporal and Spatial Variability of Ecosystem-Scale Carbon Dioxide, Water Vapor, and Energy Flux Densities , 2001 .

[67]  P. Curtis,et al.  Effects of Forest Management on Soil C and N Storage: Meta Analysis , 2001 .

[68]  P. Reichert,et al.  A comparison of techniques for the estimation of model prediction uncertainty , 1999 .

[69]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[70]  Michael I. Jordan,et al.  Advances in Neural Information Processing Systems 30 , 1995 .

[71]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[72]  Geoffrey E. Hinton,et al.  Learning representations by back-propagation errors, nature , 1986 .

[73]  E. Odum.  The strategy of ecosystem development. , 1969, Science.