Data sets are packages of information that you can use before, during or after a disaster. It is important in planning to determine who has what data, if you can get access to it, and if it is compatible with your systems.
The following are common data sets that provide baseline data, or information that is available prior to any events occurring. These are useful for planning purposes and exercises.
- Baseline data: Topography; Political boundaries; Demography; Land ownership / use, Critical facilities
- Scientific Data: Hydrography / hydrology; Soils; Geology; Seismology
- Engineering and Environmental Data: Control structures: locks, dams, levees; Building inventories/codes; Transportation, bridges, tunnels; Utility infrastructure, pipelines, power lines; Water quality; Hazardous sites; Critical facilities
- Economic Data
- Census Data
Other data sets are only available during a response. Some data sets are specific to the incident. These are usually dynamic and depend heavily on solid damage assessment and data exchange agreements.
- Critical infrastructure status: Road and bridge closures; Airport status; Utility status – water / electricity / gas / telephone
- Secondary hazards: Condition of dams and levees; Fires and toxic release potential
- Resource Information: Personnel deployment; Equipment deployment
- Medical system condition: Hospital status; Nursing home status
- Casualty information: Injuries and deaths; Medical evacuations; Location of trapped persons; Evacuation routes; Shelters
Keep in mind that not everyone uses data for the same purpose. Many do a damage assessment and this data is all called “damage assessment”, but it not useful to each other due to different needs and standards. Some examples include:
- FEMA looks a major infrastructure and systems, and overall impact.
- SBA looks at impact to businesses.
- USACE looks at impact to locks, dams, and their projects.
- Red Cross looks at impact to individuals and families.
- Insurance companies look at impact to policy holders.
Most of the data sets include a geographic location, such as an address, road, or other positioning information. This helps transform the data set from just existing in a database to being analyzed through a GIS tool. More on that in the GIS section.
A database is a location that stores data. There can be simple databases, and very complex databases. Remember that we’re looking at this from the perspective of an emergency manager. You should be familiar with the terms and some other basic information, but leave the complex database creation to the SMEs.
Here are a few terms that are used in database discussions:
- Database – an organized collection of data
- Table – data organized in rows and columns
- Attribute or field – a variable or item, think of a cell in a spreadsheet
- Record – a collection of attributes
- Domain – the range of values an attribute may have
- Key – unique data used to identify records
- Index – organize and order records
- Data dictionary & schema – documentation of connections between all the parts
Quick Tip: never buy or accept a database from someone without a properly documented data dictionary and schema. Having that will save you hassle in the future when you need support or to change it.
A database can either store all the data in a single table, or spread the data across multiple tables. Remember that large amounts of data are best handled in a multi-table database, but this also creates the problems of trying to share data as all the data must be linked back. It shouldn’t be a problem in a properly designed and documented database.
A single table database is like keeping data in Microsoft Excel. It is simple to create all the rows and columns on one sheet. Lots of data points and duplicate data points may be better organized in a multi-table database. Key fields are used to link records across many tables. Tables can be a one to one (1:1) relationship. For example, if one table contained your personal information and another table contained your transcript that would be a one to one relationship. Tables can be a one to many (1:many) relationship. A table of course descriptions may link by class name to everyone’s individual grades. Once course taken by many people. The course description could be updated once without needing to touch every record of all the people that took it. If you want to dive deeper, start at http://en.wikipedia.org/wiki/Database_model.
Meta data is a way of providing information about data, or anything else really. Metadata makes data retrieval and understand much easier. It also makes data gathering more complicated and difficult since there is more work required on the data gathering side. In my experience, everyone agrees that good metadata is good.
Like everything else, garbage in = garbage out.
http://www.clientdatastandard.org/dcds/schema/1.0 is an example of data and metadata in capturing. Each field is described as to the use and what it contains.
Another way to look at meta data. The nutrition facts is meta data for the banana. The banana is meta data for the information listed in the nutrition facts.
Next: Data standards