Name of Program
MATLAB Kategorie:MATLAB stand-alone application
Short Description of Functionality
The application ncprofil can be used to display data on profiles and time. Data can also be based on profiles in terms of adjascent control volumes.
ncprofil is the predecessor of PLOTPROFILZEIT, which is still available but will not be further developed. The difference is a re-engineering and adaptation to the new standards for Matlab-Proghome Tools. In a next step, the netcdf reading routine will be updated. In the re-engineering, the old input format XTRDATA was not included.
Possible ways to display data:
- Type = Hovmoeller: Physical quantities, which are based on a profile and are dependant on time, can be displayed with a hovmoeller diagram. The spatial dimension is displayed along the x-axis and time is displayed along the y-axis. The physical value at each point is color-coded.
- Type = Percentiles: The temporal variation of the values can also be displayed in this way. The spatial dimension is again displayed along the x-axis, but the physical value itself is displayed along the y-axis. The temporal average at each point is displayed as a solid line over the x-dimension. The temporal variation is displayed as a grey area in the background, which is the area between to chosen percentile values, default is 5 % and 95 %.
- Type = Prognose: From the percentiles, predictive values can be calculated, according to the procedure applied in the Weser Estuary. For this, each percentile is rounded up by a rounding parameter (i. e. 1 cm for water level). Then, along river stretches of i. e. 5 km, the maximum rounded value is found and this is interpreted as the predicitve value for this area. Beside the display of the percentiles (see above), this predictive value is shown as a red line. The x-axis is the spatial dimension, while the y-axis shows the physical quantity.
- Type = Allevents: The temporal evolution is displayed by showing all available synoptic values as single lines along the profile. A color spectrum is used to show the temporal evolution, as long as not too many events are overlapping.
- Type = Multidata: This allows no display of the temporal variation. Instead, values from different runs can be compared. Single values from different data sets are displayed in one diagram. The variables must not contain a time dimension, thus can either be averages of a tidal characteristic analysis or results of a long-term analysis.
- Input_Format = CVMesh2 for Netcdf analysis results fuer Netcdf as well as aggregated but not analysed simulation results
- Input_Format = Mesh1 for Netcdf simulation output on profiles (DATACONVERT result of BDF profile file)
- Input_Format = Mesh0 for Netcdf results on profiles (Profile output of UnTRIM2009)
- Steering file (example in $PROGHOME/examples/ncprofil, type ncprofil_steer.dat)
- UNTRIM results in above mentioned formats.
- Figures (*.pdf, *.png, *.eps or *.fig)
- Excel sheet (optional)
Visualization with MATLAB.
Further, some statistic values are calculated by the program in case of Type = Percentiles. This used the following Matlab functions:
- percentiles: prctile
- mean: nanmean (nan meaning that NaNs are not taken into account)
- median: nanmedian
- maximum: nanmax
- minimum: nanmin
Program(s) to run before this Program
NCAGGREGATE, NCANALYSE, UNTRIM
Program(s) to run after this Program
MATLAB (erstellt mit Version R2016b)
Needs installation of MATLAB Compiler Runtime (under linux: /usr/local/MATLAB/MATLAB_Runtime/v85)
Working group POS with A. Zorndt and U. Schiller
Examples can be found unter $PROGHOME/examples/plotProfilZeit/.
Descriptions of the method can be found here: http://de.mathworks.com/help/stats/descriptive-statistics.html .
Thanks to Ulrike Schiller for your help in updating the program.
back to Program Descriptions