8.3 Technical Advances and Problem Evolution
ties, networking, and data collection somehow
solve all the "hard" problems in ecological data
management. In fact, the technological advances do
solve some problems, but they also change the nature of others and create new ones as well. Only by
looking at the issues that persist through radical
technological change can we hope to identify the
really fundamental issues.
By reclassifying project management concerns,
we can focus on what might be called today's more
technology driven approach to ecosystem study. In
this new classification, we find some sets of issues
that have grown in importance to demand their own
specialized subclass and some that persist, but in
greatly reduced importance. There are only a few
key issues that persist in importance relatively unchanged by technological advances.
8.3.1 Data Assembly
Today we exist in an environment in which a huge
volume of digital data is available and in which the
critical decision is that of which data sets to assemble together, rather than how to overcome the
inherent difficulties in manual data collection. New
data collection is still important, but data collection
problems are often dwarfed by our opportunity (and
need) to choose between several, dozens, or even
hundreds of existing data sets that could augment
or complement a particular study. For example,
compare the researcher who yearns for a single reliable data set of key weather-related values with
one who has the option of selecting one of literally
hundreds of such data sets being collected daily
across dozens of regions. The practical problem of
finding and reading the data shifts to a more
methodological problem of why to select a particular data set. In addition, the availability of data
sets with consistent measurements across a large
space can cause significant shifts in project goals.
Whereas researchers were once inherently constrained to relatively small study areas, they now
can consider larger areas or multiscale studies that
address ecological issues at scales ranging from
small plots of land to watersheds, regions, and
larger.
8.3.2 Processing and Storage
Constraints on processing and storage resources
have not disappeared, but they have been relaxed
by orders of magnitude so that familiar practical
limits have essentially disappeared. Whereas the
scarce computing resources of yesterday tended to
be rigidly controlled and rationed out to users by
111
central support organizations, nowadays virtually
all project staff, from the most senior researchers
and analysts to the most junior clerks and dataentry technicians, control their own powerful desktop computer. Once it took several days to months
for staff members to collect data sets as inputs, digitize them appropriately, schedule processing time,
create processing software, and determine the execution parameters that would be used to produce a
single derived data set. Today a single staff member can rapidly derive dozens of such data sets on
a desktop computer without ever leaving his or her
desk. He or she can use network facilities to collect the inputs, standard software to do the processing, and local on-line disk storage to hold the
results. The availability of local on-line storage is
particularly significant; it has exploded to the point
where it is now feasible to place tens of gigabytes
of on-line storage on individual desktop computers. Removable media systems are still in use, but
the reliable, standard, and inexpensive systems of
today bear little similarity to those of the past. In
particular, read-write media such as Zip and Jaz
disks or read-only media such as CD-ROMs make
it easy to replicate and distribute data sets. In fact,
the devices themselves can be easily moved and
distributed; many are designed specifically to be
carried around in a backpack. Thus, although computing hardware, media, and software are still not
free, the problem has shifted from finding a single
reliable option to selecting between numerous affordable alternatives.
8.3.3 Networking and Dissemination
With advances in networking and on-line data storage, as exemplified by the capabilities of the Wodd
Wide Web, the distinctions among data sets prepared specifically for on-line processing, off-line
storage, or off-line dissemination have rapidly
crumbled. As information resources in even modestly sized labs grow, the parent organizations must
recognize that the work that supports data dissemination to external users, typically viewed as extra,
is in fact the core of what the organization must do
to use its own data effectively. Vast stores of data
can be kept on line, and modern networks permit
essentially instantaneous delivery of data in standard formats to either internal or external users.
Thus the problems of external media and data formats are virtually eliminated. What remains as the
primary constraint for potential users, both internal
and external, is determining the content and context for the available data sets and assuring their
appropriateness for project goals.
ties, networking, and data collection somehow
solve all the "hard" problems in ecological data
management. In fact, the technological advances do
solve some problems, but they also change the nature of others and create new ones as well. Only by
looking at the issues that persist through radical
technological change can we hope to identify the
really fundamental issues.
By reclassifying project management concerns,
we can focus on what might be called today's more
technology driven approach to ecosystem study. In
this new classification, we find some sets of issues
that have grown in importance to demand their own
specialized subclass and some that persist, but in
greatly reduced importance. There are only a few
key issues that persist in importance relatively unchanged by technological advances.
8.3.1 Data Assembly
Today we exist in an environment in which a huge
volume of digital data is available and in which the
critical decision is that of which data sets to assemble together, rather than how to overcome the
inherent difficulties in manual data collection. New
data collection is still important, but data collection
problems are often dwarfed by our opportunity (and
need) to choose between several, dozens, or even
hundreds of existing data sets that could augment
or complement a particular study. For example,
compare the researcher who yearns for a single reliable data set of key weather-related values with
one who has the option of selecting one of literally
hundreds of such data sets being collected daily
across dozens of regions. The practical problem of
finding and reading the data shifts to a more
methodological problem of why to select a particular data set. In addition, the availability of data
sets with consistent measurements across a large
space can cause significant shifts in project goals.
Whereas researchers were once inherently constrained to relatively small study areas, they now
can consider larger areas or multiscale studies that
address ecological issues at scales ranging from
small plots of land to watersheds, regions, and
larger.
8.3.2 Processing and Storage
Constraints on processing and storage resources
have not disappeared, but they have been relaxed
by orders of magnitude so that familiar practical
limits have essentially disappeared. Whereas the
scarce computing resources of yesterday tended to
be rigidly controlled and rationed out to users by
111
central support organizations, nowadays virtually
all project staff, from the most senior researchers
and analysts to the most junior clerks and dataentry technicians, control their own powerful desktop computer. Once it took several days to months
for staff members to collect data sets as inputs, digitize them appropriately, schedule processing time,
create processing software, and determine the execution parameters that would be used to produce a
single derived data set. Today a single staff member can rapidly derive dozens of such data sets on
a desktop computer without ever leaving his or her
desk. He or she can use network facilities to collect the inputs, standard software to do the processing, and local on-line disk storage to hold the
results. The availability of local on-line storage is
particularly significant; it has exploded to the point
where it is now feasible to place tens of gigabytes
of on-line storage on individual desktop computers. Removable media systems are still in use, but
the reliable, standard, and inexpensive systems of
today bear little similarity to those of the past. In
particular, read-write media such as Zip and Jaz
disks or read-only media such as CD-ROMs make
it easy to replicate and distribute data sets. In fact,
the devices themselves can be easily moved and
distributed; many are designed specifically to be
carried around in a backpack. Thus, although computing hardware, media, and software are still not
free, the problem has shifted from finding a single
reliable option to selecting between numerous affordable alternatives.
8.3.3 Networking and Dissemination
With advances in networking and on-line data storage, as exemplified by the capabilities of the Wodd
Wide Web, the distinctions among data sets prepared specifically for on-line processing, off-line
storage, or off-line dissemination have rapidly
crumbled. As information resources in even modestly sized labs grow, the parent organizations must
recognize that the work that supports data dissemination to external users, typically viewed as extra,
is in fact the core of what the organization must do
to use its own data effectively. Vast stores of data
can be kept on line, and modern networks permit
essentially instantaneous delivery of data in standard formats to either internal or external users.
Thus the problems of external media and data formats are virtually eliminated. What remains as the
primary constraint for potential users, both internal
and external, is determining the content and context for the available data sets and assuring their
appropriateness for project goals.