Ice Buoy Times Series Plots
FILL IN HERE! DATA PRODUCT NOT YET READY!
Revision History
- 20121003: Initial Release
- 20130725: changes to date include changing the colour map, making the colourbar show discrete colours, accepts new data product options, including averaging.
Parameters
Note:
Formats
This data is available as a PNG image or PDF file.
Description
Instruments deployed on profiling platforms, such as the Vertical Profiler System (VPS), produce time series scalar data that may be plotted against time and water depth in a contour plot. This facilitates visualization of water property changes over time, such as salinity or temperature. Below is an example plot from our testing environment with a limited time range.
The plots at slip into temperature, heating cycle 1 and heating cycle 2 data. The sensor elevation is set with zero being ice level at time of deployment of the sensor. The temperature data has a sampling rate of 6 hours and each heating cycle has a sample rate of 24 hours. no resampling and averaging available.
Here is an example pdf: CambridgeBay_SafePassageBuoy_IceBuoy_20160227T000000Z_20160430T235959Z-Heat1-736413.6288.pdf. Note that the naming convention goes as '-(data type)-(Time that the first set of data)' in the MODE field, see dp home.
To archive the best results with this data product, select 4 to 10 days of time when not resampling. The data gridding routine and parameters are setup to work best with the amount of data provided by a 7 day search, when the data has a sample of period 1 second. To extend the plot to longer time ranges, select resampling. A resample period of one minute will provide ample data from 10 days up to 6 months; for searches greater than 6 months, choose 15 minute resampling. If there is insufficient data, the plot will have a lot of extrapolation which can be misleading. Extrapolation occurs on the plot where the gridding routine only has a few points to draw from or where the data points are in a single direction, such as when there are no nearby data points and where the data doesn't surround the contours. The data points are marked on the plot with black dot markers or a thick black line (thicker than the thin black contour lines. Some extrapolation is necessary to fill out the contours, but too much leads to data being 'invented' where none actually exists. The extrapolation is limited to 2 hours and 25 decibars, anything less causes gaps when plotting nominal VPS data. The example plot shown above, and the PDF example, both exhibit extrapolation, as the time range is too short to contain enough VPS profiles through the water column to fill out the plot. Ideally, profile plots should contain at least 12 VPS profiles or casts through the water column spaced regularly over the time range plotted. When the VPS is operating normally, it runs on a schedule with at least 3 profiles/casts per day. The more profiles in the plot, the smoother and more accurate it will be.
Discussion
To comment on this product, click Add Comment below.