of data. It would be useful to offer guidance related
Solutions
Communication is key when contracting out GIS
to the dos and don'ts of data maintenance scheduling.
services. There needs to be a closer relationship
This would allow for the flexibility necessary to sup-
between the contracting office (CO) and the contract-
port all USACE projects and mission areas.
ing officer's technical representative (COTR). The
The survey found the following problems in the
COTR's responsibility is to monitor procurements and
area of geospatial data maintenance.
deliverables at the technical level to ensure that the
USACE is getting what it needs and expects from the
Problem: Budgeting for data maintenance
contractor. For every geospatial dataset created, the
Description
deliverable must include a corresponding geospatial
Data maintenance is often ignored in project bud-
metadata file that conforms to standards set by the
gets. USACE-owned data is usually kept for histori-
Federal Geographic Data Committee.
cal purposes, but data owned by other state or federal
agencies may be thrown away if disk space becomes
Geospatial data and systems maintenance
a constraint. When an outside source makes a data
Data maintenance refers to the updating of a data
request from a District or Division, funds are not avail-
collection within a time frame that is meaningful
able to make updates, so the data is sent out `as is.'
for the data type or for the project purpose. For exam-
ple, data maintenance on river bottom cross-sections
Solutions
taken for navigation purposes requires much more
Data maintenance must be put into project budgets
frequent updates than a dataset of county-wide terrain
for mission-required data. Overall, most respondents
elevation for a mapping project. Data maintenance
said that data maintenance is needed, but funding is
also requires backup of datasets and maintenance of
not available so maintenance is done on an ad hoc
the hardware and software necessary to create and
basis. This is based on a corporate culture that is ori-
read data files.
ented to the short-term and doesn't look at data as
USACE generally does not provide uniform guid-
a long-term investment. Labor for updates must be
ance related to maintenance of geospatial data. This
charged to projects and must be included in the proj-
flexibility in requirements for updating is good con-
ect budget.
sidering the differences in the projects in each District
or Division. A greater awareness of systems mainten-
Problem: Data backups
ance is sorely needed, however. This is discussed in
Description
the problem section below.
A wide variety of routines exist for geospatial
Data layers that are actively maintained vary by
data backups. Some locations do daily backups for
District. In some Districts, landcover may be updated
all changes that occur and full backups once a week.
every decade. In Districts with strong environmental
Other locations do backups when they have a chance.
The situation often varied within a District between
an annual basis. Data collection update cycles can
CADD and GIS data backups and who was responsi-
range from 15-minute cycles for automated systems
ble for the data. In some instances, the backup sched-
to 10-year cycles for less important or less dynamic
ule is increased when weather conditions indicate the
data layers.
potential for a natural disaster.
USACE engineer manuals sometimes recommend
that mission-specific data be updated at regular inter-
Solutions
vals. Based on these recommendations, some mis-
The best backup and storage schema are at loca-
sion-specific data is actively maintained by Divisions,
tions where the relationship between the geospatial
including
data users and information management people is
strong. Backup and storage recommendations can be
Channel improvement master plan data
found in the technical section of this report. There is
absolutely no valid reason or excuse for not having
water, rainfall, and salinity
a consistent, logical backup and storage plan. Each
Geodetic control data
location paid for the original collection of data, and
Navigation-related data.
this investment must be protected.
Because maintenance schedules for geospatial data
Problem: Software and hardware updates
are so diverse, it would not be useful to create broad-
Description
based guidance on how often to update various themes
As was seen previously in the data management
4