sky
observatory
Astronomy
Pipelines
tape
Archive

local
Archives

 

scope, with notable demand for high performance and data quality. Scientific data ranges from medical and biological to community science and from large-scale institutional to local laboratories. Here, we focus on the big amounts of data collected or produced by instruments archived in databases and managed by DBMSs. The database community has expertise that can be applied to solve the problems in existing scientific databases.

 

observation and simulation Scientific data originates through observation and/or simulation. 16 Observational data is collected through detectors; input is digitized, and output is raw observational data. Simulation data is produced through simulators that take as input the values of simulation parameters. Both types of data are

virtual
observatories
Web

 

often necessary for scientific research on the same topic; for instance, observational data is compared with simulation data produced under the same experimental setup. Consider three examples, one each for observational, simulation, and combined:

Observational scientific data. The
SDSS located at Ohio State Univer-
sity and Johns Hopkins University, is
a long-running astronomy project.
Since 2000, it has generated a detailed
3D map of about 25% of the sky (as
seen from Earth) containing millions
of galaxies and quasars. One reason
for its success is its use of the SQL
Server DBMS. The SDSS uses the tele-
scope at Apache Point Observatory,
NM, to scan the sky at regular intervals
to collect raw data. Online processing
is done on the data to detect the stars
and galaxies in the region. This online
processing also helps detect and fine-
tune the telescope’s alignment. The
image data is then recorded onto tape
for archival purposes.

figure 2. Workflow of earthquake-simulation data.

Mesh
generation
Mesh

u2(t)

simulation u1(t)

 

Analysis
visualization
Ground model
time-varying output