Biogeosciences www.biogeosciences.net 1726-4170 1726-4189 6 10 2009 10.5194/bg-6-2001-2009 http://www.biogeosciences.net/6/2001/2009/ http://www.biogeosciences.net/6/2001/2009/bg-6-2001-2009.html http://www.biogeosciences.net/6/2001/2009/bg-6-2001-2009.pdf 2001 2013 2009-10-06 Towards global empirical upscaling of FLUXNET eddy covariance observations: validation of a model tree ensemble approach using a biosphere model M. Jung mjung@bgc-jena.mpg.de M. Reichstein A. Bondeau Max Planck Institute for Biogeochemistry, Jena, Germany Potsdam Institute for Climate Impact Research (PIK), Potsdam, Germany Global, spatially and temporally explicit estimates of carbon and water fluxes derived from empirical up-scaling eddy covariance measurements would constitute a new and possibly powerful data stream to study the variability of the global terrestrial carbon and water cycle. This paper introduces and validates a machine learning approach dedicated to the upscaling of observations from the current global network of eddy covariance towers (FLUXNET). We present a new model TRee Induction ALgorithm (TRIAL) that performs hierarchical stratification of the data set into units where particular multiple regressions for a target variable hold. We propose an ensemble approach (Evolving tRees with RandOm gRowth, ERROR) where the base learning algorithm is perturbed in order to gain a diverse sequence of different model trees which evolves over time. <br><br> We evaluate the efficiency of the model tree ensemble (MTE) approach using an artificial data set derived from the Lund-Potsdam-Jena managed Land (LPJmL) biosphere model. We aim at reproducing global monthly gross primary production as simulated by LPJmL from 1998–2005 using only locations and months where high quality FLUXNET data exist for the training of the model trees. The model trees are trained with the LPJmL land cover and meteorological input data, climate data, and the fraction of absorbed photosynthetic active radiation simulated by LPJmL. Given that we know the "true result" in the form of global LPJmL simulations we can effectively study the performance of the MTE upscaling and associated problems of extrapolation capacity. <br><br> We show that MTE is able to explain 92% of the variability of the global LPJmL GPP simulations. The mean spatial pattern and the seasonal variability of GPP that constitute the largest sources of variance are very well reproduced (96% and 94% of variance explained respectively) while the monthly interannual anomalies which occupy much less variance are less well matched (41% of variance explained). We demonstrate the substantially improved accuracy of MTE over individual model trees in particular for the monthly anomalies and for situations of extrapolation. We estimate that roughly one fifth of the domain is subject to extrapolation while MTE is still able to reproduce 73% of the LPJmL GPP variability here. <br><br> This paper presents for the first time a benchmark for a global FLUXNET upscaling approach that will be employed in future studies. Although the real world FLUXNET upscaling is more complicated than for a noise free and reduced complexity biosphere model as presented here, our results show that an empirical upscaling from the current FLUXNET network with MTE is feasible and able to extract global patterns of carbon flux variability. Akaike, H.: A new look at the statistical model identification, IEEE Transactions on Automatic Control, 19(6), 716–723, 1974. Bates, J. M. and Granger, C. W. J.: The combination of forecasts, Operations Research Quarterly, 20, 451–468, 1969. Bondeau, A., Smith, P. C., Zaehle, S., et al.: Modelling the role of agriculture for the 20th century global terrestrial carbon balance, Glob.l Change Biol., 13(3), 679–706, 2007. Breiman, L.: Bagging predictors, Mach. Learn., 24(2), 123–140, 1996. Breiman, L.: Random forests, Mach. Learn., 45(1), 5–32, 2001. Breiman, L., Friedman, J., Olshen, R., and Stone J.: Classification and Regression Tree, Wadsworth and Brooks, 1984. Brovkin, V., Sitch, S., von Bloh, W., Claussen, M., Bauer, E., and Cramer, W.: Role of land cover changes for atmospheric CO2 increase and climate change during the last 150 years, Glob. Change Biol., 10(8), 1253–1266, 2004. Burnham, K. P. and Anderson, D. R.: Multimodel inference – understanding AIC and BIC in model selection, Sociological Methods and Research, 33(2), 261–304, 2004. Chandra, D. K., Ravi, V., and Bose I.: Failure prediction of dotcom companies using hybrid intelligent techniques, Expert Syst. Appl., 36, 4830–4837, 2009. Dietterich, T. G.: An experimental comparison of three methods for constructing ensembles of decision trees: Bagging, boosting, and randomization, Mach. Learn., 40(2), 139–157, 2000. Fader, M., Rost, S., and Müller, C.: Virtual water content of temperate cereals and maize: Present and potential future pattern, J. Hydrol., in review. Freund, Y. and Schapire,R. E.: Experiments with a new boosting algorithm, Proceedings of the 13th International Conference on Machine Learning, 148–156, 1996. Geurts, P., Ernst, D., and Wehenkel, L.: Extremely randomized trees, Mach. Learn., 63(1), 3–42, 2006. Hansen, L. and Salamon, P.: Neural network ensembles, IEEE Trans. Pattern Anal. Mach. Intell., 12, 993–1001, 1990. Haxeltine, A. and Prentice, I. C.: BIOME3: an equilibrium terrestrial biosphere model based on ecophysiological constraints, resource availability and competition among plant functional types, Global Biogeochem. Cy., 10, 693–710, 1996. Ho, T. K.: The random subspace method for constructing decision forests, Ieee Transactions on Pattern Analysis and Machine Intelligence, 20(8), 832–844, 1998. Jones, C., Collins, M., Cox, P., and Spall, S. A.: The Carbon Cycle Response to ENSO: A Coupled Climate-Carbon Cycle Model Study, J. Climate, 14, 4113–4129, 2001. Jung, M., Verstraete, M., Gobron, N., Reichstein, M., Papale, D., Bondeau, A., Robustelli, M., and Pinty, B.: Diagnostic assessment of European gross primary production, Glob. Change Biol., 14(10), 2349–2364, 2008. Jung, M., Vetter, M., Herold, M., et al.: Uncertainties of modeling gross primary productivity over Europe: A systematic study on the effects of using different drivers and terrestrial biosphere models, Global Biogeochem. Cy., 21, GB4021, doi:10.1029/2006GB002915, 2007. Karalic, A.: Employing linear regression in regression tree leaves, Proceedings of the 10th European Conference on Artificial Intelligence, 440–441, 1992. Knorr, W., Gobron, N., Scholze, M., Kaminski, T., Schnur, R., and Pinty, B.: Impact of terrestrial biosphere carbon exchanges on the anomalous CO₂ increase in 2002–2003, Geophys. Res. Lett., 34, L09703, doi:10.1029/2006GL029019, 2007. Kocev, D., Dzeroski, S., White, M. D., Newell, G., and Griffioen, P.: Using single- and multi-target regression trees and ensembles to model a compound index of vegetation condition, Ecol. Modell., 220, 1159–1168, 2009. Lasslop, G., Reichstein, M., Kattge, J., and Papale, D.: Influence of observation errors in eddy flux data on inverse model paramter estimation, Biogeosciences, 5, 1311–1324, 2008. Lasslop, G., Reichstein, M., Papale, D., Richardson, A. D., Arneth, A., Barr, A., Stoy, P., and Wohlfahrt, G.: Separation of net ecosystem exchange into assimilation and respiration using a light response curve approach: critical issues and global evaluation, Glob. Change Biol., doi:10.1111/j.1365-2486.2009.02041.x, in press, 2009. Liu, F. T., Ting, K. M., and Fan, W.: Maximizing Tree Diversity by Building Complete-Random Decision Trees, Advances in Knowledge Discovery and Data Mining, 9th Pacific-Asia Conference, PAKDD 2005, 2005. Liu, F. T., Ting, K. M., Yu, Y., and Zhou, Z. H.: Spectrum of variable-random trees, J. Artif. Intell. Res., 32, 355–384, 2008. Loh, W., Chen, C. W., and Zheng, W.: Extrapolation errors in linear model trees, ACM Trans. Knowl. Discov. Data, 1(2), 6, ISSN:1556-4681, 2007. Lucht, W., Prentice, I. C., Myneni, R. B., Sitch, S., Friedlingstein, P., Cramer, W., Bousquet, P., Buermann, W., and Smith, B.: Climatic control of the high-latitude vegetation greening trend and Pinatubo effect, Science, 296(5573), 1687–1689, 2002. Makridakis, S., Anderson, A., Carbone, R., Fildes, R., Hibdon, M., and Lewandowski, R.: The accuracy of extrapolation (time series) methods: Results of a forecasting competition, J. Forecast., 1, 111–153, 1982. Malerba, D., Esposito, F., Ceci, M., and Appice, A.: Top-Down Induction of Model Trees with Regression and Splitting Nodes, IEEE Trans. Pattern Anal. Mach. Intell., 26(5), 612–625, 2004. New, M., Lister,D. Hulme, M. and Makin I.: A high-resolution data set of surface climate over global land areas, Climate Res., 21, 1–25, 2002. Österle, H., Gerstengarbe, F.-W., and Werner, P. C.: Homogenisierung und Aktualisierung des Klimadatensdatzes der Climate Research Unit of East Anglia, Norwich, Terra Nostra, 6, 326–329, 2003. Papale, D. and Valentini, A.: A new assessment of European forests carbon exchanges by eddy fluxes and artificial neural network spatialization, Glob. Change Biol., 9(4), 525–535, 2003. Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W., Longdoz, B., Rambal, S., Valentini, R., Vesala, T., and Yakir, D.: Towards a standardized processing of Net Ecosystem Exchange measured with eddy covariance technique: algorithms and uncertainty estimation, Biogeosciences, 3, 571–583, 2006. Potts, D. and Sammut, C.: Incremental learning of linear model trees, Mach. Learn., 61(1–3), 5–48, 2005. Prentice, I. C., Cramer, W., Harrison, S. P., Leemans, R., Monserud, R. A., and Solomon, A. M.: A Global Biome Model Based on Plant Physiology and Dominance, Soil Properties and Climate, J. Biogeography, 19(2), 117–134, 1992. Qian, H., Joseph, R., and Zeng, N.: Response of the terrestrial carbon cycle to the El Nino-Southern Oscillation, Tellus B, 60(4), 537–550, 2008. Reichstein, M., Papale, D., Valentini, R., et al.: Determinants of terrestrial ecosystem carbon balance inferred from European eddy covariance flux sites, Geophys. Res. Lett., 34, L01402, doi:10.1029/2006GL027880, 2007. Reichstein, M., Falge, E., Baldocchi, D., et al.: On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm, Glob. Change Biol., 11(9), 1424–1439, 2005. Richardson, A. D., Hollinger, D. Y., Burba, G. G., et al.: A multi-site analysis of random error in tower-based measurements of carbon and energy fluxes, Agr. Forest. Meteorol., 136(1–2), 1–18, 2006. Schaphoff, S., Lucht, W., Gerten, D., Sitch, S., Cramer, W., and Prentice, I. C.: Terrestrial biosphere carbon storage under alternative climate projections, Climatic Change, 74(1–3), 97–122, 2006. Schwarz, G.: Estimating the dimension of a model, Ann. Stat., 6(2), 461–464, 1978. Sims, D. A., Rahman, A. F., Cordova, V. D., et al.: On the use of MODIS EVI to assess gross primary productivity of North American ecosystems, J. Geophys. Res.-Biogeosci., 111(G4), G04015, doi:10.1029/2006JG000162, 2006. Sitch, S., Brovkin, V., von Bloh, W., van Vuuren, D., Assessment, B., and Ganopolski, A.: Impacts of future land cover changes on atmospheric CO₂ and climate, Global Biogeochem. Cy., 19(2), GB2013, doi:10.1029/2004GB002311, 2005. Sitch, S., Smith, B., Prentice, I. C., et al.: Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model, Glob. Change Biol., 9(2), 161–185, 2003. Vens, C. and Blockeel, H.: A simple regression based heuristic for learning model trees, Intelligent Data Analysis, 10, 215–236, 2006. Vetter, M., Churkina, G., Jung, M., Reichstein, M., Zaehle, S., Bondeau, A., Chen, Y., Ciais, P., Feser, F., Freibauer, A., Geyer, R., Jones, C., Papale, D., Tenhunen, J., Tomelleri, E., Trusilova, K., Viovy, N., and Heimann, M.: Analyzing the causes and spatial pattern of the European 2003 carbon flux anomaly using seven models, Biogeosciences, 5, 561–583, 2008. Vogel, D. S., Asparouhov, O., and Scheffer, T.: Scalable look-ahead linear regression trees, in: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, edited by: ACM, San Jose, California, USA, 2007. Weber, U., Jung, M., Reichstein, M., Beer, C., Braakhekke, M. C., Lehsten, V., Ghent, D., Kaduk, J., Viovy, N., Ciais, P., Gobron, N., and Rödenbeck, C.: The interannual variability of Africa's ecosystem productivity: a multi-model analysis, Biogeosciences, 6, 285–295, 2009. Xiao, J. F., Zhuang, Q. L., Baldocchi, D. D., et al.: Estimation of net ecosystem carbon exchange for the conterminous United States by combining MODIS and AmeriFlux data, Agr. Forest Meteorol., 148(11), 1827–1847, 2008. Yang, L., Ichii ,K., White, M. A., Hashimoto, H., Michaelis, A., Votava, P., Zhu, A., Huete, A., Running, S., and Nemani, R.: Developing a continental-scale measure of gross primary production by combining MODIS and AmeriFlux data through Support Vector Machine Approach, Remote Sens. Environ., 110, 109–122, 2007.