Ecological Applications, 21(5), 2011, pp. 1429–1442 Ó 2011 by the Ecological Society of America Ecological forecasting and data assimilation in a data-rich era YIQI LUO, 1,7 KIONA OGLE, 2,3 COLIN TUCKER, 2 SHENFENG FEI, 1 CHAO GAO, 1 SHANNON LADEAU, 4 JAMES S. CLARK, 5 AND DAVID S. SCHIMEL 6 1 Department of Botany and Microbiology, University of Oklahoma, Norman, Oklahoma 73019 USA 2 Department of Botany, University of Wyoming, Laramie, Wyoming 82071 USA 3 Department of Statistics, University of Wyoming, Laramie, Wyoming 82071 USA 4 Cary Institute of Ecosystem Studies, Millbrook, New York 12545 USA 5 Department of Biology, Duke University, Durham, North Carolina 27708 USA 6 NEON, Boulder, Colorado 80301 USA Abstract. 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. Key words: data assimilation; data–model fusion; ecological forecasting; inverse analysis; optimization; predictions; prognosis; projections. THE NEED FOR ECOLOGICAL FORECASTING The capability to forecast the impacts of environ- mental change on our living environment and natural resources is critical to decision making in a world where the past is no longer a clear guide to the future (Clark et al. 2001). We are living in a period marked by rapid climate change (Solomon et al. 2007), profound alter- ation of biogeochemical cycles (Vitousek et al. 1997), unsustainable depletion of natural resources (Heinz Report 2008), proliferation of exotic species (D’Antonio and Vitousek 1992, Liao et al. 2008) and infectious disease (Smith et al. 2005), and deterioration of air and water quality (Gleick 2002, Akimoto 2003). Human populations are increasing at an alarming rate, and society is dependent on the extraction and utilization of natural resources to support regional and global economies. Predictable and increasing supplies of energy, food, fiber, freshwater, and clean air are necessary to maintain healthy human societies. To effectively mitigate and adapt to climate change, we need to develop robust methods to apply data and current knowledge to the problem of anticipating future states of ecosystems and then to assess resilience and, potentially, collapse of ecosystem services. Nascent ecological forecast models are in use in some areas. For example, ecosystem and biogeochemical cycling models have been incorporated into earth-system models to project terrestrial carbon sinks and sources and their feedback to climate change in the 21st century (Cox et al. 2000, Friedlingstein et al. 2006). Those model predictions have been incorporated into the assessment Manuscript received 15 July 2009; revised 19 July 2010; accepted 4 August 2010. Corresponding Editor: S. K. Collinge. For reprints of this Invited Feature, see footnote 1, p. 1427. 7 E-mail: yluo@ou.edu 1429 July 2011 DATA ASSIMILATION FOR ECOLOGICAL FORECASTING