Environmental Engineering Reference
In-Depth Information
are stored in a database, while two-dimensional quantifications are saved in one
(or several) image file(s). In both cases, dynamics (the changing character of the
objects) are not originally included in the files. The description of the movement of
a water body is as well the result of joining single measurements similar as the
changes of a habitat or air circulation. Calculations of changes of quantities over
time result in speed, rates and directions of changes.
If and how a model is dynamic (just using a pre-calculated file describing the
dynamics of an entity, or continuously executing calculations directly with
the model) depends on the architecture of the GIS model framework of the IEM:
the tighter the coupling between model, GIS and data is (see Sect.
22.2
) the easier
GIS and model(s) can exchange data. A loose coupling via data files might be
limited to a stepwise calculation of a dynamic process, especially considering the
lack of (standardized) appropriate file structure. A tight coupling might evoke the
tight embedding of the model within a GIS or vice versa.
Actually, GIS need some technical and conceptual extensions to deal with
dynamics (Derby et al. 2005). Again, GIS uses methods of integration to access
continuous
data. The location of the underlying data is the key to (spatially) join
(see Sect.
22.3
) time series as well as to merge periodic data sets. By spatially
matching multiple layers or periodic data GIS are able to calculate, to analyse, and
to model spatial relationships over time, identifying trends, patterns and changes
within a defined location. The GIS matches
layers
of data from the same area by
their geo-position; one data set laying on top of the other. The difference between
the superimposed data layers yields the
delta
needed to calculate parameters of
dynamics. Besides the simple calculation of the difference between the quality
of two (or more) layers, GIS offers the calculations of distances, length and areas of
the objects that changed.
Temporal/Spatial Aggregation
In addition to one-dimensional point data files frommeasurements and two-dimensional
data from mappings and remote sensing stored in raster or vector files, model data most
commonly are stored in n-dimensional data arrays. Platform-independent data inter-
faces like netCDF (
Network Common Data Format
) offer a sophisticated method to
store data as well as metadata. Time series in multi-dimensional files accessed via
interfaces, enable the GIS to
cut
single temporal layers from the file. Processing
complex dynamic data requires the aggregation of information from multipart files
(Shao et al. 2000). Integration necessitates methods to quantify and qualify the changes
and to adjust the parameters of models capturing field data.
Data analysis generalizes changes within the arrays, analysing speed, location
and extension of the changes. Beyond the single change from one time-step to the
next, general trends and directions are relevant to understand the system observed.
Boundary layers, diffusion of pollutants and movement of organisms are resulting
topics.
