375
placed on products for supply chain monitoring, remote sensors on satellites or
UAVs, streamflow and water quality gauges, smart-agriculture measurements of
soil moisture, and so on. Surveys are census methods administered by researchers
for collecting usage, production, or transportation data. Census methods are generally employed at the establishment scale (space) and annual to decadal scale (time)
using an affordable statistical sampling technique, with data released in aggregated
form to preserve privacy. Inventory data is collected by businesses for their own
internal purposes. Inventories account for how much of a product is available at an
establishment, on order, or en route. Inventory data is a central component of private
sector supply chain management. Inventory data tends to be private and may be
considered trade secrets. These three bottom-up data types are critical for understanding the “last mile” of the FEW supply chain and the flows of FEW commodities in general, along with emergency management.
FEW systems data has several attributes, including but not limited to structure,
quality, spatial resolution, temporal resolution, process resolution, and scale. Data
management best practices are essential both for users and originators of data at
every stage of the data life cycle. Important privacy and ethics principles compromise the accessibility of bottom-up data and compromise the utility of top-down
data resources. Major strengths and weaknesses in data availability and accessibility
are reviewed for application domains of the FEW system.
This chapter does not attempt to identify all current FEW system data sources
(an impossible task!) but does employ numerous examples. Most of these examples
are drawn from the USA. Each chapter of this book surveys exemplary datasets that
are germane to its specific concepts and applications, so the reader should look to
the chapters for guidance on where to find especially useful data for each topic. The
chapter’s end matter surveys some particularly useful datasets for the FEW system,
and the reader should inspect these sources for further education.
14.2 Data Attributes: Structure, Quality, Scale
and Resolution
14.2.1 Data Structure (and Type)
FEW systems are coupled natural-human systems that may often be types of socioecological systems. The native conceptual model for FEW systems is usually a multitype (or multiplex, Baggio et al. 2016) network with many qualitatively different
types of agents, behaviors, processes, boundaries, stocks, and flows. The special
type of mathematics that characterizes this system is a graph or network—specifically, the process network (Ruddell and Kumar 2009). This type of graph features
nodes that represent many types of natural or human processes or agents in the
system; these processes and agents act on the FEW system network by transforming
inputs into outputs, producing and consuming goods and services, and causing
14 Data
placed on products for supply chain monitoring, remote sensors on satellites or
UAVs, streamflow and water quality gauges, smart-agriculture measurements of
soil moisture, and so on. Surveys are census methods administered by researchers
for collecting usage, production, or transportation data. Census methods are generally employed at the establishment scale (space) and annual to decadal scale (time)
using an affordable statistical sampling technique, with data released in aggregated
form to preserve privacy. Inventory data is collected by businesses for their own
internal purposes. Inventories account for how much of a product is available at an
establishment, on order, or en route. Inventory data is a central component of private
sector supply chain management. Inventory data tends to be private and may be
considered trade secrets. These three bottom-up data types are critical for understanding the “last mile” of the FEW supply chain and the flows of FEW commodities in general, along with emergency management.
FEW systems data has several attributes, including but not limited to structure,
quality, spatial resolution, temporal resolution, process resolution, and scale. Data
management best practices are essential both for users and originators of data at
every stage of the data life cycle. Important privacy and ethics principles compromise the accessibility of bottom-up data and compromise the utility of top-down
data resources. Major strengths and weaknesses in data availability and accessibility
are reviewed for application domains of the FEW system.
This chapter does not attempt to identify all current FEW system data sources
(an impossible task!) but does employ numerous examples. Most of these examples
are drawn from the USA. Each chapter of this book surveys exemplary datasets that
are germane to its specific concepts and applications, so the reader should look to
the chapters for guidance on where to find especially useful data for each topic. The
chapter’s end matter surveys some particularly useful datasets for the FEW system,
and the reader should inspect these sources for further education.
14.2 Data Attributes: Structure, Quality, Scale
and Resolution
14.2.1 Data Structure (and Type)
FEW systems are coupled natural-human systems that may often be types of socioecological systems. The native conceptual model for FEW systems is usually a multitype (or multiplex, Baggio et al. 2016) network with many qualitatively different
types of agents, behaviors, processes, boundaries, stocks, and flows. The special
type of mathematics that characterizes this system is a graph or network—specifically, the process network (Ruddell and Kumar 2009). This type of graph features
nodes that represent many types of natural or human processes or agents in the
system; these processes and agents act on the FEW system network by transforming
inputs into outputs, producing and consuming goods and services, and causing
14 Data