385
long to use for operational decision-making. The coarsest possible FEW systems
data is Paleo data, which describes ancient patterns in FEW systems before modern
data became available. Long-term climate models are similarly coarse, with resolution on an (arguably) decades-to-centuries timescale.
Fine time resolution data can sometimes be used as “early warning” of problems
and for operational decision-making. Fine dynamics also allow scientists to establish cause and effect in FEW systems robustly and to build accurate models (as
opposed to simply observing long-term trends). However, fine temporal resolution
data below the Annual scale is relatively rare and unavailable in FEW systems, even
for establishments and processes that are identified at a fine spatial resolution. The
higher-frequency and near-real-time sensor stream data needed for operations pose
serious software engineering challenges.
The decadal timescale is most common for census data publication. Decadal
timescale FEW system process dynamics include for instance changes in technology, economic and population growth, demographics, changing culture and consumption patterns, collapse and conflict, natural capital accounts and natural
resource sustainability, infrastructure development, changing trade patterns, changing legal environments, climate, strategic reserves and storage systems, elections,
and policy. Annual timescale dynamics are often aggregated up to decadal scales for
publication and reporting.
The annual timescale is the most common establishment scale survey data reporting resolution. However, these annual data tend to represent processes and dynamics that are either slower or faster. Annual data tend to be coarse aggregations of
hourly and seasonal processes and patterns, summarizing total outcomes but losing
most of the process dynamics arising from finer scales. Annual data also tend to
reflect decadal trends but contain more noise than the decadal data with respect to
those trends. Downscaling annual data reports to the more appropriate seasonal
scale, or obtaining seasonal data, is a fundamental problem for FEW systems
research and operations.
Seasonal timescale data is rarely collected empirically but is a common resolution for model output. This seasonal timescale corresponds to everything from days
to months but is most commonly 1 month. Seasonal processes in FEW systems
include Weather norms and averages impacting agricultural production, transportation, and energy demand, drought, high or low water levels, the water year with low
storage after summer and high storage in spring, water demand for agriculture during warm seasons, annual harvests of grain and crops, urban water demand patterns
(peaking in summer for irrigation), urban energy demand patterns (peaking in winter for heating and in summer for cooling), liquid fuel demand cycles (higher in
summer due to more travel), storage system filling and discharge, streamflow,
renewable energy production, water stress, water quality (especially temperature),
and utility pricing.
Hourly timescale data refers to scales from seconds to days, but nominally 1 h—
because increments of 1 h resolve the key diurnal patterns of the human life cycle
and business activities impacting the FEW system but also because weather and
ecological functions change dramatically between day and night. Urban electrical
14 Data
long to use for operational decision-making. The coarsest possible FEW systems
data is Paleo data, which describes ancient patterns in FEW systems before modern
data became available. Long-term climate models are similarly coarse, with resolution on an (arguably) decades-to-centuries timescale.
Fine time resolution data can sometimes be used as “early warning” of problems
and for operational decision-making. Fine dynamics also allow scientists to establish cause and effect in FEW systems robustly and to build accurate models (as
opposed to simply observing long-term trends). However, fine temporal resolution
data below the Annual scale is relatively rare and unavailable in FEW systems, even
for establishments and processes that are identified at a fine spatial resolution. The
higher-frequency and near-real-time sensor stream data needed for operations pose
serious software engineering challenges.
The decadal timescale is most common for census data publication. Decadal
timescale FEW system process dynamics include for instance changes in technology, economic and population growth, demographics, changing culture and consumption patterns, collapse and conflict, natural capital accounts and natural
resource sustainability, infrastructure development, changing trade patterns, changing legal environments, climate, strategic reserves and storage systems, elections,
and policy. Annual timescale dynamics are often aggregated up to decadal scales for
publication and reporting.
The annual timescale is the most common establishment scale survey data reporting resolution. However, these annual data tend to represent processes and dynamics that are either slower or faster. Annual data tend to be coarse aggregations of
hourly and seasonal processes and patterns, summarizing total outcomes but losing
most of the process dynamics arising from finer scales. Annual data also tend to
reflect decadal trends but contain more noise than the decadal data with respect to
those trends. Downscaling annual data reports to the more appropriate seasonal
scale, or obtaining seasonal data, is a fundamental problem for FEW systems
research and operations.
Seasonal timescale data is rarely collected empirically but is a common resolution for model output. This seasonal timescale corresponds to everything from days
to months but is most commonly 1 month. Seasonal processes in FEW systems
include Weather norms and averages impacting agricultural production, transportation, and energy demand, drought, high or low water levels, the water year with low
storage after summer and high storage in spring, water demand for agriculture during warm seasons, annual harvests of grain and crops, urban water demand patterns
(peaking in summer for irrigation), urban energy demand patterns (peaking in winter for heating and in summer for cooling), liquid fuel demand cycles (higher in
summer due to more travel), storage system filling and discharge, streamflow,
renewable energy production, water stress, water quality (especially temperature),
and utility pricing.
Hourly timescale data refers to scales from seconds to days, but nominally 1 h—
because increments of 1 h resolve the key diurnal patterns of the human life cycle
and business activities impacting the FEW system but also because weather and
ecological functions change dramatically between day and night. Urban electrical
14 Data