Zhaobin: Ecological research is entering the age of great science and large data

Speaking of the word "great science", the first is to think of the rocket, http://www.aliyun.com/zixun/aggregation/32260.html "> Astronomical telescope and Particle accelerator research, these studies need a lot of financial support." And the ecological research, which brings to mind the feet-like plant ecologists, the bird ecologists with the telescope, and the insect-scientists who hold the nets, and so on, few people associate their research with big science. In fact, conversely, this is particularly surprising! The areas studied by ecologists, such as the large majority land, are areas that are truly occupied by humans, but have not been categorized as great science. But, led by David Schimel, a large state of American ecologists is planning to correct that. They plan to shake the field of terrestrial ecology and bring it to the scale and scope of the great science that they are building, the National Ecological Observing Network (NEON, Nation Ecological Observatory receptacle), This is the topic that this article will discuss as a detailed case.

Big Data covers a growing range. The New York Times August 2, 2012, for example, explores the emerging cultural gene (barely meme), the big data. Usually, people look at big data from a technical enthusiast's point of view. The New York Times article describes large data as "the application of artificial intelligence tools, such as machine learning, the collection of huge data, which goes beyond the standard database." New data sources include traces of web browsing data, social network communications, sensor data, and monitoring data. ”

The

Large data is shaped by the natural world and affects how we understand the natural world and communicate with the natural world. Biology provides a huge amount of data at so many levels: almost limitless data from genetic research is flooding the byte of our largest digital warehouse, a massive collection of specimens from the Natural History Museum, photographic archives taken in daily life on a global scale, and a large number of "civic science" (Citizen Science) data in the database, such as national phenology receptacle, on this content, in the 2012 issue of the frontiers in Ecology and the environnement " Civic Science "There is a very important review article" from Caprio's Lilacs to the American national Phenology Receptacle "introduced the network; the collection of large-scale biogeochemical studies, such as the coastal" dead zone " is essentially a confluence of industrial nitrogen conversion, human agricultural practices, primary productivity and biological respiration. Interactive maps of these areas have been published by WRI.

Big data is affecting how we understand the world, because it is disintegrating what we have been told over the past 50 years to be the cornerstone of scientific understanding: to test falsified hypotheses with controlled experimentation in a "strong reasoning" framework. The idea that science must be falsified comes from Popper (Karl Popper), who was inspired by the John Pratt of John Platt, the high citation "strong inference" published in his 1964 (fervent Inference) Xiongwen advocated the use of standardized methods for biological research. Observation and Ecology's author, Rafe, said: "If you like it the first time you read his article, I do, but I urge you to read it in a critical manner." And said it was a bit like Ayn Rand (Ayn Rand) in high school, trying to combine her thoughts with reality as an adult. It is obvious that these philosophical ideas produce some customary rules: "Relevance does not imply causation", "mode cannot reveal mechanism", and scientific conclusion, if it does not overturn predetermined assumptions, serves only as "pre-trial investigation". Big Data makes these solid beliefs look a little odd. While these beliefs are sometimes still valuable, the reflexive nature of them is applied in the same way as scientists and non-scientists, which need to be reassessed. If we can't get 100% clear of some of the institutional causes (which no one can do), large data methods allow life scientists to discover more robust patterns in larger chaotic cycles, or they will allow us to be as close to the real picture as possible. At the same time, in the treatment of large data should be noted that large data will never completely replace the small data. The small data that takes time to observe nature, and the understanding of the scraps of small data, are the cornerstones of the whole of large data.

In the field of ecology, it is not a household name to mention neon, but it can be said to have added an unusually bright glow to this large data age. Initially, the project was only based on a search for financial support in Boulder, Colorado, and the project was exceptionally difficult. But after 10 years of discussion and planning, the National Science Foundation managed to persuade Congress to earmark $434 million trillion for construction, a medium space probe priced at about $80 million a year. In the 2011, the continental Scale eco-network program of the United States was finally launched, which will comprehensively monitor the environmental changes throughout the United States, fundamentally changing the traditional small-scale and regional research methods, and forming a land-field group cooperation. Researchers claim that the era of big data in the ecological field has come and that ecological observations will usher in a "torrent of data".

Eventually, the 60 regions covering the United States will be set up synchronously. Once this station network is completed in 2016, 15,000 sensors will be used to collect more than 500 types of data, including temperature, precipitation, air pressure, wind speed and wind direction, humidity, sunshine, air pollutant concentrations such as ozone, and the total amount of nutrients in the soil and streams, As well as areas of vegetation and microbial state. In each place these instruments will be installed in the same way, using the same measurements, adhering to the long-term data collection in a standardized manner, in the hope of achieving the need for statistical efficiency, and turning ecological research from some technical experimental observations into an industrial-scale enterprise. With such a foundation, it is possible to observe how ecosystems respond to climate change and land-use change and the invasion of new species. Ultimately, these observational data can be used to develop models for predicting the future of ecosystems, which can help policymakers assess the consequences of various programmes of action.

Fig. 1 American Neon Project divides the United States into 20 different ecosystem regions

Neon's researchers divided the United States into 20 different regions (Figure 1), each representing a specific ecosystem type. Each area is equipped with three sets of sensors. A set of fixed installations at the core sites for at least 30 years of continuous monitoring, the core sites of environmental conditions undisturbed and may be maintained. The other two sets are movable and moved to another place after 3-5 years of observation in one place, and these "floating" sites are used for comparisons within the region. Each of the sites, whether core or floating, has a sensor-filled observation tower that is 10 meters higher than the existing vegetation canopy. Within a dozens of-square-kilometer area around the tower, researchers have placed more sensors in soil and streams, measuring temperature, carbon dioxide and nutrient levels, and root growth rates and microbial activity. These sensors will record how different ecosystems use nutrients and water effectively, how vegetation responds to climate change, and how carbon dioxide moves between organisms and the atmosphere. This will help to seek an understanding of the carbon cycle and the consequences of climate change caused by greenhouse gases. To complement these ground measurements, the researchers will also conduct a yearly aerial survey at each of the core sites, observing the chemical characteristics of the leaves and the health of the forest canopy, and can be used for comparison with satellite observations. In addition, neon researchers can deploy a specially equipped aircraft equipped with a LIDAR (a radar optics), a spectrometer (the determination of chemical composition) and a high-resolution camera for assessing the effects of natural disasters such as floods, wildfires and pest outbreaks.

(Responsible editor: The good of the Legacy)