Weather Chart FAQs

How do I make a chart?
What is the data source?
How do I get the freshest data?
Why does a two-day chart has fewer points than a one-day chart.
What data is shown in the table?
How should I interpret the time axis?
Why are there two wind options, Wind (0-36) and Wind (0-40)?
Why do some charts grow taller?
Why do some sites show all zeros for barometric pressure?
What is the delta-P charting option (Columbia Basin sites)?
How is the direction data aggregated, especially when the wind is out of the north?
What kind of software is used to make this all work?
What are the little numbers, in the upper left hand corner, to the right of the home link?

How do I make a chart?  top

Use the four drop-down lists to control what data is presented; then click the "Press to Chart" button.  The drop-down lists set the following parameters (from left to right):

1. Location:  select from any of the listed weather monitoring sites. Note, you can change between between a list of Minnesota sites and Washington/Oregon sites by clicking on the words "Minnesota Weather" (or "Columbia Basin Weather") in the upper left corner. Use the same link to return. The defaults are Roosevelt for the Columbia Basin sites and Mankato for the Minnesota sites.
2. Sensor:  select wind (default), temperature, pressure, or delta-pressure (available on for the Columbia Basin sites)
3. End date:  select the last day of data to be shown in the chart (default is today)
4. "Days":  the total number of days to be presented in the chart (default is 1)

The "<" and ">" buttons (to the right of the "Days" control) step the chart backward ("<") or forward (">") in time.  The size of the step is determined by the value of the "Days" control. The ">" button (to the right of the stations select list) acts to step through the stations list and update the chart all in one click.

Some recent additions:

The URL (web page address) is now parameterized. This allows you to e-mail a link for a particular day and location, or to save a shortcut that will give you the latest data at your favorite site.

To create a URL that gives the latest data, just trim the URL back to the sensor parameter. For example, here is a link to the latest wind data for the Faribault, MN airport:

http://whitebear.gac.edu/waconia/Weather/Rosy_combo.ASP?region=MN&jS=T&hT=T&pT=F&Location=KFBL&Sensor=Wind40

And check out the link in the words "Columbia Basin Weather" (or "Minnesota Weather") that takes you to the Minnesota (or Washington) sites. Use the same link to return.

The "sites" link on the Minnesota side shows an image of the MN airports that I collect data from. I've added a few new MN airports to my gleaning process (Faribault, New Ulm, Paynesville, and Albert Lea). Faribault should be handy for those folks that sail at Cannon Lake.

http://whitebear.gac.edu/waconia/Weather/mn_sites.jpg

And finally there is a new secondary-y axis, for direction, when charting wind. The details are in the FAQ below on the two wind options.

What is the data source?  top

Data is collected from three sources:

Waconia's first data source was the HMS web page (http://hms.pnl.gov/stainfo.htm ).  A Perl script runs every 5 minutes and collects all the text from that page and then parses out wind, temperature, and pressure data.  This data is written to a MS Access database on the Gustavus Physics department's public server. Here is a map of the 30 HMS sites:
http://hms.pnl.gov/stamap.htm



Later, NOAA data was added from http://iwin.nws.noaa.gov/iwin/wa/hourly.html and http://iwin.nws.noaa.gov/iwin/or/hourly.html.  These NOAA sites provide data for Portland, The Dalles, and other cities not covered by HMS.  This additional gleaning is used to make the delta-P charting option available.



And now in 2006, data collection for Minnesota has begun. The source site is here:
http://weather.aero/tutorials/dataserver/metars/MetarsXMLOutput.php
Here is a Google Earth image: MN sites:
Here is a Google Earth kmz file (open with Google Earth): MN sites (interactive):

How do I get the freshest data?  top

You can set Waconia to auto-refresh every 5 minutes by clicking the "start timer" link. This feature automatically syncs your browser to refresh 10 seconds after the 5-minute updates on the Whitebear server and insures that you have the freshest data available. Another useful indicator is the difference (in minutes) between the current time and the time stamp on the latest data record. This is shown in the last set of parentheses in the string of numbers to the right of the home link.

A two-day chart has fewer points than a one-day chart.  Why?  top

Charts of two or more days of data will show hourly aggregates of the raw 15-minute data.  For example,  if the HMS web site presents data at 14:00, 14:15, 14:30, and 14:45, the chart will represent those four points with a single aggregate point at 2 p.m. PST (14/24 = 0.583 day fraction).

What data is shown in the table? top

The table at the top shows today's latest reading for the selected site.  This is a raw fifteen minute reading and is equivalent to the current data shown at the source site.  It is also equivalent to the last point on a one-day chart.

How should I interpret the time axis?  top

Here are a few things to remember when interpreting the time axis on the chart:

• A data point at 2:15 p.m. is an aggregate of high-frequency sampling during the 15-minutes prior to the 2:15 timestamp of the report.  In other words, a point at 2:15 reflects conditions between 2:00 p.m. and 2:15 p.m.
• When charting more than one day, the axis converts to a day format and the charted points are hourly aggregates.  A point at the half-day mark corresponds with an aggregate of the 12th hour (12n = 12 noon) reports (that is, conditions between noon and 1 p.m.).  A point at the day mark corresponds to the 0th hour (12 midnight) reports (that is, conditions between midnight and 1 a.m.).
• In the summer, day light savings time is automatically taken into account.
  • Timestamps in the table, both the server time and the "latest reading" time, indicate day light savings time with a "PDT" (Pacific Daylight Time).  A "PST" indicates Pacific Standard Time.
  • Whether the chart itself uses PST or PDT is determined by the end time of the chart selected.  Multi-day charts that end with a date within PDT will use (for the whole chart) PDT time values.
  • For the curious out there, the database is all recorded in PST.  Code in the ASP page handles the conversion to PDT when needed for clarity to the user.

Why are there two wind options, Wind (0-36) and Wind (0-40)? top

Both options provide wind direction and wind speed information. The direction is defined by the secondary y-axis on the right side of the chart.

The 0-40 option is the most used wind direction charting option. In this case the direction data is scaled so that the compass points correspond with the 5 mph grid lines on the chart.

The 0-36 option is a good choice for people that are comfortable thinking of direction in terms of degrees (0-360). 180 degrees on the right axis corresponds with 18 mph on the left axis; a scaling factor of 1/10. One frequent user, a sailplane pilot, prefers this wind charting option.

In either case, the secondary axis makes it straightforward to interpret the direction data.

For the curious, some history and development details: The chart-image module used in Waconia (ASPChart) unfortunately does not have an option for a secondary y-axis. The secondary axis that you currently see is actually a hand-built static image (a transparent gif), overlaid on the chart. Through the magic of HTML, CSS, and JavaScript, the two images are revealed simultaneously to the give the appearance of a single image. In earlier releases of Waconia, a simple compromise (using no secondary axis and no overlay image) was implemented, through which the wind-direction data was scaled by the ratio 40/360 (for the 0-40 option) or 1/10 (for the 0-36 option). Without a secondary axis, the user was left to interpret the direction points using only the speed axis and the clues in the image legend. Thanks to the image overlay in the current release, the legend has now been simplified; it's way better.

Why do some charts grow taller ?  top

There is an image scaling feature for the wind, temperature and delta-p charts.  

• If the wind exceeds 40 mph, the number of pixels in the chart scales (grows) with the max speed in the chart; this give an accurate visual impression of those big wind days.  This feature is especially useful when you look at the wind up on Rattlesnake mountain.
• For the temperature chart, the auto scaling is based on the difference between the minimum and maximum points.  If this differential exceeds 40 degrees, the number of pixels (along the temperature axis) scales with the differential.  If the differential is less than 40 degrees, the data is centered within the 40 degrees of range of the temperature axis.
• For the delta-p chart, if the min to max range exceeds 0.26 (the default range on the y axis), the charts image size scales.

Why do some (Columbia Basin) sites show all zeros for barometric pressure ?  top

Sites with no pressure sensor show zeros.  All NOAA sites have pressure sensors. The following HMS sites have a pressure sensor:

1. FFTF
2. Roosevelt
3. HMS
4. PFP
5. YAKB
6. 100K
7. 100N
8. 200E
9. 300A
10. Richland

What is the delta-P charting option (Columbia Basin sites)?  top

This option plots differential pressures between four sites: Portland, TheDalles, Roosevelt, and Richland.  The differentials are defined in the chart legend.  For example, the trace named "Portland - The Dalles" is calculated as the pressure difference between Portland and TheDalles (using sea-level corrected pressures).

Pressure differentials can be useful in predicting wind at Roosevelt WA.   Pressure gradients often move from west to east so a good differential between Portland and The Dalles can be a good indicator of wind to come at Rosy.  Differentials over about 0.05 inches Hg generally indicate sailable wind.

How is the direction data aggregated, especially when the wind is out of the north?  top

For charts of two or more days, I don't have a great way to aggregate near-North direction data (that is, directions that are bouncing back and forth between slightly greater than 0 and slightly less than 360).  So I simply use the average of the 4 direction readings.  This method can produce an erroneous result of approximately 180 (or 20 using Waconia direction notation) if the direction is near North (example:  (3 + 359 + 2 + 358)/ 4 = 180.5).

In the one-day charts this is not an issue because there is no aggregation of the HMS 15-minute data.

What kind of software is used to make this all work?  top

Perl, Python, ASP and VBScript (Active Server Pages) are the main tools.  For a little more detail, take a look at the "about the technology behind the viewer" page.

What are the little numbers in the upper left hand corner, to the right of the home link?  top

There are three categories of information in this string of numbers:

The first set indicates the number of image requests in the last 24 hours (1d = one day) and the last 72 hours (3d = 3 days).

And just to the right of these numbers is the time on the server when the page was generated.

And farthest to the right, in parentheses, is the difference in minutes between the server time and the time stamp of the latest data record.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.