|
Monitoring data analysis: real-time capture and processing
 Environmental information systems have to manage time series of monitoring data efficiently.
All ESS model and information systems share a set of generic functions with a common data structure to
 |
integrate on-line monitoring data including real-time links |
|
display and analyse time series of monitoring data |
|
link the monitoring data to other system components such as the GIS
(spatial interpolation functions) and the simulation models
(initial and boundary conditions, real-time data assimilation, comparison and calibration) |
| A web-based ToolKit for monitoring data: capture, processing, model integration, statistical analysis, display, storage |
Web-based, real-time system such as AirWare and WaterWare include a a set of functions for the capture, display, and analysis of time series of observation and monitoring data, model validation (direct comparison of model results and monitoring data) as well as the animation of dynamic model results.
Examples of this data management tools and their integration with simulation models are WaterWare and AirWare. The on-line manuals describe the features of a number of new time series display and analysis functions.
|
The basic functionality includes the display of time series data, scrolling, zooming, aggregation, unit conversions, and reading back numerical values for individual dates and time stamps from the graphical display.
Data are shown for individual observation stations (or station objects), that can be selected either from a list of stations, or from the map.
|
|
|
The analysis for a station's data is performed in the spatial context defined for the station, for example, a measurement network or a watershed.
While the data of an individual station are displayed in the time series graph, a list of neighbouring stations can be used to switch on or off other stations for the analysis.
|
|
|
A date selector allows to define sub-periods of the time series, or find the maximum common temporal coverage for a set of stations.
Test include a test for spatial homogeneity, where the data from one station are compared against the average of its neighbouring stations, and deviations from a user-defined threshold around that mean are indicated.
|
|
| In a related test, the exceedance of individual values (raw data or aggregates for a specific period like a day, week, or month) are shown against an upper and lower limit drawn around the arithmetic mean of the variable analysed. A similar test shows the data against an upper threshold, for example and air quality standard, and counts the number of instances when the standard is violated. These frequencies are then compared against the substance specific rules for compliance with the respective environmental legislation. |
|
|
Another test for the comparison of stations, in particular for cumulative variables like precipitation or runoff, is the comparison of mass curves.
Here the cumulative distributions are compared for two stations or groups of stations to highlight both absolute and relative differences between the data sets.
|
|
| To analyse the seasonality of temporal variability of time series, a simple display of temporal auto-correlation values with increasing time shifts can be used. The resulting pattern, in this example monthly shifts of precipitation data, show a clear annual pattern. Depending on the temporal resolution or aggregation of the data, daily (hourly), weekly (daily), monthly (daily), and inter-annual (monthly) patterns can be analysed and displayed. |
|
|
Spatial interpolation of oberservation data uses a distance-weighted interpolation function; depending on the type of variable, alternative methods can be used.
Interactive control over the display style, isoline display, and animation features are additional features of this module.
|
|
|
