"&yunits=0"in your command line.
&yunits=0(for a single station query)
wget -np -q -O output.txt -i address.txt, where address.txt contains something similar to the example 1 above but replacing formchk=1 with wget=1.
0 12 * * * wget -np -q -O output.txt -i address.txt, it will retrieve the selected data every day at 12h00 and save it in output.txt
-npoption means 'no parents'. It ensures that wget will not go up in the directory tree.
-qoption means 'quiet'. Error messages and other informations will not be displayed.
-ioption must precede the address filename and
-Othe output filename.
wget -A.png -r -l1 -np -nd -q -i address.txt
-roption means 'recursive'. This means the targets of the links (here images) in the html page will be retrieved
-A.pngoption means 'only retrieve png files'. This means you won't get html files or whatever
-ndoption means 'no directory'. This avoid building a local copy of the distant directory hierarchy
-l1option means 'one level'. Specify recursion maximum depth level: only one needed here
wget --spider -l1 -np -i address.txt
wget -O myimage.png http://[nest address]/data/upload/chart.png
--spidermeans you just connect to the address, but retrieve nothing (i.e. we build the image this way).
-pis not the same as
|stations||APTY, AATB, ATHN, ARNM etc.|
|allstations||1 (selects all the stations)|
|tabchoice||revori, ori, 1h|
|dtype||corr_for_efficiency, corr_for_pressure, uncorrected, pressure_mbar|
|tresolution||0, 2, 5, 10, 30, 60, 120, 360, 720, 1440, 39276, 525969 (0 correspond to "best")|
|date_choice||last, bydate (, bygle, byforbush)|
|last_days||1 to 9999|
|last_label||days_label, hours_label or mins_label|
|force||1 or 0|
|yunits||1 or 0 (% or counts/mbar for a single station query)|
|yscale||1 or 0|
|yscalemin||-9999 to 9999 or min|
|yscalemax||-9999 to 9999 or max|
|anomalous||1 or 0|
|shift||0 to 999|
|start_day||1 to 31|
|end_day||1 to 31|
|start_month||1 to 12|
|end_month||1 to 12|
|start_year||1956 to curent year+1|
|end_year||1956 to current year +1|
|start_hour||0 to 23|
|end_hour||0 to 23|
|start_min||0 to 59|
|end_min||0 to 59|
|output||plot, ascii, both|
|asciifrac||1 or 0|
|odtype||corr_for_efficiency, corr_for_pressure, uncorrected, pressure_mbar|
|envdata||measured_temperature_inside, measured_temperature_outside, measured_wind_speed_m_s, measured_relative_humidity|
With the quick plots, you don't have to go through the whole form to get a plot of last data or last GLE.
Parameters are chosen to match to the most common requests:
Credits: Magnetic and geomagnetic poles coordinates are retrieved from Kyoto World Data Center for Geomagnetism
Vertical cutoff iso-rigidities (2020) are retrieved from Gerontidou et al. (https://doi.org/10.1016/j.asr.2021.01.011)
Stations colors: Black circles on the map above correspond to historical stations which are now closed.
Some stations may have the same grey color: these are new stations (in NMDB) which are under test.
Using the form, just click on the color box next to the station name to change this station color*
You can also apply the same color to all stations (one color button) or reset all colors (reset colors).
All/online stations : Check all stations with the "all stations" checkbox, or select only the stations which are online (i.e. with a lag lower than 15 minutes)
Select this option if you want to retrieve the most recent data of NMDB
Select this option if you want to retrieve data for a specific time period: give the start date and end date
Select this option if you want to retrieve the data corresponding to a specific Ground Level Event.
The start-time of the plot is given by the pre-increase period of the GLE. The duration of the event can be modified in the options.
Forbush decrease date:
Select this option if you want to retrieve the data corresponding to a specific Forbush decrease.
The start-time of the plot is given by the date and time of the onset minus 1 hour.
Note that the current list is restricted to large Forbush decreases (min 8%).
|Corrected for efficiency|
|Corrected for pressure|
|Smoothed sunspot number|
|Monthly sunspot number|
|GOES proton plot (max 15 days)|
|Kp 3-hourly index (max 1 year)|
|Collapse time labels|
|Wind speed (m/s)|
NMDB database holds 3 tables for each station:
Modify the time resolution: "best" will give you the best time resolution available
in the specified table, usually 1 minute for recent data. This is the default.
In the case you decrease time resolution, data will be average from the original resolution.
Important Choosing '1h' in the time resolution field is not the same as choosing the 1hour validated table (see also Note 1 & 2)
In order to reduce computation time and to display the results in a fast manner, values may be automatically averaged to a lower time resolution and/or could be retrieved from the 1hour table, depending on the query size. Check the plot title or the log file.
Example query 1: 1min time resolution for 1year and 1 single station => 1min values averaged over 30mins.
Example query 2: 1min time resolution for 6months and 8 stations => 1hour values averaged over 2hours.
When the user asks for a time resolution lower or equal to 2 hours, data is automatically averaged from the 1 hour values.
You can use the "force" parameter to force a better time resolution than what is automatically determined. But:
- This will increase the computation time.
- There is still a limitation. This shall give you a better time resolution but may not give you the best one.
1 Carr. rot. = 1 Carrington rotation = 27.2753 days
Averaging by this amount should correspond to Carrington's
rotations but MySQL results could be weird: use with caution.
This is the width of the smoothing window. The smoothing is a classical
unweighted sliding-average smooth, i.e. each value V of the time serie
is replaced by the average of 2n+1 (= width) values: n values before and n values after V.
The width should be positive and lower than 100. If a width value is even, then
width+1 will be used instead.
The width of the window is a number of points, not a duration.
i.e. if you get 10 min averaged data and Smooth width is set,
the 10 min averaged data points will be smoothed (not the 1min data)
|Select this option to include the most extreme values for the scaling of the plot.
Otherwise the scale is computed using the MAD (Median Absolute Deviation) which discards these outliers.
|This option will either: |
where MAD is the Median Absolute Deviation (i.e. we compute the mean after excluding possible outliers).
In the second case the same mean value "M" is computed for each time series "T".
What you get on the plot is:
|n * offset + level * ( T / M ),|
|where "offset" represent the distance between parallel plots (set to 4 by default, but user-adjustable),
"n"=0,1,2,3... and "level" is set to 20 by default. Note that these parameters have no physical meaning.
They are adjusted to represent multiple times series in a convenient way.
Example for 3 stations:
Tkiel plot = 0 + 20*Tkiel/Mkiel
Toulu plot = 4 + 20*Toulu/Moulu
Tkerg plot = 8 + 20*Tkerg/Mkerg
|Interval between parallel plots:|
|The offset defined above.
|Specify your own Y axis scale:|
|Define your own limits for the Y axis, the min and the max. You can define
either min or max or both. The option "ignore anomalous values" will not be taken into account in this case.
|Define your own baseline interval
If you want to compare 2 or more stations, it might be useful to set up
a common baseline interval, the same way baseline intervals are defined for GLEs.
Time series will be scaled using the mean value within this interval (you might also set the offset define above to 0).
The chosen interval must be within the time range defined in "date selection".
Note that this will only be possible in relative scale and that it will disable some navigation options.
|Adjust pre-increase level:|
|In the case you want to plot a specific GLE or FD, this option will adjust time series from all stations to a
common level (pre-increase level, see 'relative scale' in 'scaling options' for pre-increase level calculation).
The values are then given as the increase in % above this level. This option
is not compatible with "ignore anomalous values for scaling".
By default the GLE duration is set to 10 hours. By "duration" we mean time after the end of the baseline interval,
a GLE usually starts within the next hour.
The baseline for FD is given by the hour before the event. The default duration is arbitrarily computed as
2.5*tmin+5 (tmin=time from the onset to FD minimum).
|Normalize flux to 1|
|Every time serie is scaled so that the maximum equal 1.
|Specify the event duration:|
|You can set the duration parameter to any integer between 1 hour and 99 hours.|
|If ascii is chosen for a GLE/FD together with relative scale, the relative scale is displayed in
the ascii file (only if a single station is selected)
|the date timestamp in the ascii file will be eg 1996.4285 instead of 1996-06-05 20:00:00
The leap years are taken into account in the calculation.
|Display null values|
|If there is no data (neither corrected, uncorrected or pressure) for a given timestamp and a given station, |
there is no entry in NMDB database for this timestamp. When retrieving ascii data from NMDB, the default
is to display only what is in the DB. If you want to see data gaps filled with null values, check this option.
Note that the gaps may already be filled with null values if more than one station is retrieved
- also sets 'show gaps' of 'style options'
- If a given station has no data at all for the date range requested (in multi station requests), it will still appear in the table (default is to discard this column)