Our generation of ecologists has been offered great opportunities to advance ecology with big data. Ecologists and scientists in our relevant disciplines have been making measurements at scales spanning 17 orders of magnitude, from molecules of CO2 (diameter = 3.3×10-10 m), genomics, biochemistry, cellular and organismal physiology, to the functioning of entire ecosystems, regions, and the Earth system (diameter = 1.3×107 m). For example, genomics research involving high throughput DNA sequencing has been used to generate molecular-scale data about cellular and organismal metabolism, while on the other extreme, dozens of satellites have been launched into space to observe metabolic processes at the planet scale. Many Earth surface observatory networks, such as NEON, have been generating massive amounts of ecological data, and thousands of manipulative experiments have been conducted in the laboratory and the field to study responses of biological and ecological processes to climate change. A deluge of ecological data has become available, with storage volumes already well beyond petabytes, with terabytes of data accumulating daily. Now we face the challenges on how we can use the big data to advance our process understanding and ecological forecasting.
Results/Conclusions
This poster will offer a few approaches and examples to illustrate uses of big data for ecological research. For example, big data analysis may involve interpretable models (e.g., decision trees), which are white-box machine learning algorithms, to obtain insight into processes to improve ecological forecasting. While simulation modeling has been used by ecologists for decades for prediction, but with limited data constraints, big data can be used to select alternative structures and train parameters of models via data assimilation. To fully capitalize on the information contained in big data, we may have to combine process-guided machine learning with data-driven modeling to understand ecological processes and forecast ecosystem responses to global change at regional and global scales