Environmental Engineering Reference
In-Depth Information
networks (e.g., NEON, GLEON, OOI; see earlier section on challenges associated with
global change) adds new opportunities for large-scale, cross-system analyses along with
new data management and processing challenges, and researchers will have to convert
the “deluge of data” (
Baraniuk 2011
) into scientific understanding.
Management and analysis of larger data sets requires more powerful computational
resources and creative analytical tools. Standardization of data formatting across analytical
platforms, development of sufficient descriptive documentation to accompany shared
databases, and policies for data access, security, and sharing are all important considera-
tions for projects involving large data streams (
Michener and Jones 2012
). Along with
these techniques, we must also supply appropriate ways to verify that final extracted
values can be trusted with or without raw information to examine. The developing field of
ecological informatics or ecoinformatics attempts to integrate environmental and informa-
tion sciences by providing tools to analyze and access the large amounts of collected data.
One example of ecoinformatics is the creation of “middleware” such as DataTurbine,
which can ingest data of all types, reorganize them, and produce transformed structures
suitable for archival storage in databases or file systems, and is compatible with statistical
software programs (
Benson et al. 2009
). The ability to analyze and store large data sets
using high-performance computing environments, such as cluster-based computing, cloud
computing, and heterogeneous computational environments, all offer potential solutions
to these challenges (
Schadt et al. 2010
).
A less obvious problem associated with the increased ease and decreased cost of collect-
ing data is the issue of data storage. The amount of data generated worldwide has been
increasing by 58% every year. For example, in 2010 the world generated 1250 billion
gigabytes of data, and in 2011, the amount of data collected was twice as large as the
world's storage capacity (
Baraniuk 2011
). The gap between production of data and
the ability to store it means that data will be unavailable for further analysis, lost,
overwritten, or deleted. Terabytes or even petabytes of data will require inventive ways
to be processed and stored, whether by new mathematical algorithms, compression techni-
ques, or other strategies.
Finally, data analysis and modeling approaches that effectively use these newly available
data streams are also being developed (
Schmidt and Lipson 2009
). For example, data assim-
ilation methods that generate weather forecasts can be used to create tools for ecological
forecasting (
Luo et al. 2011
). Although dealing with the increasing amount of information
and data is challenging, it will provide unprecedented opportunities to improve our under-
standing of ecosystems, and aid in developing novel scientific questions.
THE CHANGING CULTURE OF SCIENCE
The way that we do science affects the kind of science that we do. Although perhaps a
less obvious driver of scientific change than technological or conceptual advances, changes
in scientific culture will affect ecosystem science in coming decades. Here are a few
examples.
The increasing scale, complexity, and urgency of many problems in ecosystem science
is pushing ecologists to work in larger, more intellectually diverse teams (
Greene 2007
).
