HR: 08:30h
AN: IN21D-03 INVITED [Abstracts]
TI: Geoscience Informatics and the National Science Foundation
AU: * Leinen, M
EM: mleinen@nsf.gov
AF: National Science Foundation, 4201 Wilson Blvd., Arlington, VA 22230
United States
AB:
The central importance of data to the geosciences, the diversity of data types, the volume of data, and the importance of
integrating data collected by many individual researchers to study Earth and space systems, have often put geoscientists and
space scientists at the forefront of developing ways to handle data.
Recently, however, changes in the role of data in our science are changing how we deal with data:
(a) Data volume is increasing - modern observing tools are providing larger
volumes of data, more people are collecting data;
(b) Scientists are tackling increasingly complex problems that often require different types from different sources;
(c) There is growth in observatory science that is generating real-time
data streams. These pose a number of challenges. Part of the scientific
utility of observatories comes from being able to detect and respond to events
in real-time. In many cases quality control must be done automatically.
(d) Environmental data from unattended observing systems is likely to become used the way weather data is most often used by
researchers - not directly but indirectly as a data product obtained from assimilation of raw data into a numerical model of
the system.
These changes have and will continue to require changes in how data are managed. For example, data discovery is an
increasing challenge. Researchers must pay more attention to metadata - ensuring that it is created as data are collected,
processed and archived, ensuring that it is standardized, ensuring that metadata records are machine readable and permit
machine-reasoning about the relations between data.
NSF is responding to these challenges in several ways. First, virtually all large projects must have sophisticated data
management components. Second, we are funding many geoinformatics projects that propose to develop new tools for data
management. Third, NSF computer sciences funds research efforts in data management - several in partnership with
geoscientists. Finally, NSF is in the process of evaluating our policies for funding data management in order to be
responsive to the challenges of ensuring access to data and to state of the art tools for using data.
DE: 0525 Data management
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005