Actual surface data is obtained from the Southern Region Headquarters of the National Weather Service.
http://www.srh.noaa.gov/product_sites.php?pil=CF6&max=61
A handful of stations do not provide complete data from the website above and I am forced to find it manually via other sources. These statons are listed below with their data source.
Oshkosh, WI; WI-WN-4 CoCoRaHS
Ann Arbor, MI; MI-WS-11 CoCoRaHS
Houghton, MI; MI-HG-1 CoCoRaHS
Escanaba, MI; NWS Station GLADSTONE NO 2
Traverse City, MI; NWS station TRAVERSE CITY MUNSON
Champaign, IL; IL-CP-47 CoCoRaHS
January Forecast "Snowfall" results
9 out of the 40 cities finished within +/- 30% of forecast total.
January Forecast "Days With Measurable Snowfall" results
28 out of the 40 cities finished within +/- 2 days of forecast total.
For more information visit the OSNW3|WxClimate Maps webpage.
If there are any questions or thoughts on my research and analysis of the theory or how I presented the material just let me know in the comments section of the blog. Thanks for reading!
Showing posts with label Google Visualization. Show all posts
Showing posts with label Google Visualization. Show all posts
Jan 23, 2013
Too Early To Make Predictions For Spring Storm Outbreaks?
The other day on the AccuWx Forums LRC Thread, Apo Agathos, questioned if it was too early to make predictions for spring storm outbreaks? It has become painfully obvious that many weather enthusiasts are once again very close to giving up on winter and setting their sites towards chasing thunderstorms.
Recently I completed the work to get the daily maximum temperatures stuffed into Google Motion Charts. This output allowed me to analyze daily Midwest temperature changes as an entity. With the 'spring storm' question on my mind I was on the lookout for a drastic change from warm to cool which can potentially heighten the possibility.
The first notable frontal passage is visible in early March.
The second comes during the same pattern the next cycle in late April.
Click on the images to navigate to the Motion Chart. Choose the correct month from the menu on the right. Click the play button and watch the pretty colors of the temperature flux in the Midwest. The chart is fully controllable allowing the user to change the chart type, select locations, scroll through time, etc. The forecast numbers are updated daily and are grouped by month.
When using the theory to project weather conditions months in advance it is necessary to give or take a day for the given time frame. Currently the forecast data is derived using a 52 day cycle length. The 500mb plots for this pattern in previous cycles are below. A link to the surface plot can be obtained by clicking on the 500mb plot image.
Nov 18, 2012
Jan 10, 2013
If there are any questions or thoughts on my research and analysis of the theory or how I presented the material just let me know in the comments section of the blog. Thanks for reading!
Recently I completed the work to get the daily maximum temperatures stuffed into Google Motion Charts. This output allowed me to analyze daily Midwest temperature changes as an entity. With the 'spring storm' question on my mind I was on the lookout for a drastic change from warm to cool which can potentially heighten the possibility.
The first notable frontal passage is visible in early March.
The second comes during the same pattern the next cycle in late April.
Click on the images to navigate to the Motion Chart. Choose the correct month from the menu on the right. Click the play button and watch the pretty colors of the temperature flux in the Midwest. The chart is fully controllable allowing the user to change the chart type, select locations, scroll through time, etc. The forecast numbers are updated daily and are grouped by month.
When using the theory to project weather conditions months in advance it is necessary to give or take a day for the given time frame. Currently the forecast data is derived using a 52 day cycle length. The 500mb plots for this pattern in previous cycles are below. A link to the surface plot can be obtained by clicking on the 500mb plot image.
Nov 18, 2012
Jan 10, 2013
If there are any questions or thoughts on my research and analysis of the theory or how I presented the material just let me know in the comments section of the blog. Thanks for reading!
Jan 2, 2013
December Snow Forecast Results & New Locations Added to the Trends
Recently I added three new locations to the Trend output. (thanks for the inquiries gentlemen, you know who you are)
Dayton, OH
Davenport, IA
Seattle, WA
These locations have now been added to the snow forecast outputs as well. Also, with the close of December I've put together the results of the December snow forecast. They can be accessed by clicking on the "Results" link next to the "December" map link on the bottom right side of the maps webpage linked below.
The result columns can be sorted by clicking on the table headers. The Error slider will list the stations based on the criteria selected.
For anyone who checks the results out, I would be very interested in your thoughts in regards to accuracy, presentation, etc. For example, is six inches too much of a difference between forecast and actual to give the forecast a passing grade? How many Days With Snow from actual would be considered a fail? Is there another category of analysis I should be presenting? This use of the theory is a bunch of crock and should be halted at once.
All feedback/ideas are welcome.
Something to note with precipitation forecasts for Jan and Feb. Something I call the 'cycle ratio' of the algorithm updates Jan-11. Consider it a new initialization of the modeled precip for future cycles. What this actually means is that the numbers that are forecast now will likely change on Jan-11. The next 'cycle ratio' update occurs Mar-6.
Dayton, OH
Davenport, IA
Seattle, WA
These locations have now been added to the snow forecast outputs as well. Also, with the close of December I've put together the results of the December snow forecast. They can be accessed by clicking on the "Results" link next to the "December" map link on the bottom right side of the maps webpage linked below.
The result columns can be sorted by clicking on the table headers. The Error slider will list the stations based on the criteria selected.
For anyone who checks the results out, I would be very interested in your thoughts in regards to accuracy, presentation, etc. For example, is six inches too much of a difference between forecast and actual to give the forecast a passing grade? How many Days With Snow from actual would be considered a fail? Is there another category of analysis I should be presenting? This use of the theory is a bunch of crock and should be halted at once.
All feedback/ideas are welcome.
Something to note with precipitation forecasts for Jan and Feb. Something I call the 'cycle ratio' of the algorithm updates Jan-11. Consider it a new initialization of the modeled precip for future cycles. What this actually means is that the numbers that are forecast now will likely change on Jan-11. The next 'cycle ratio' update occurs Mar-6.
Dec 1, 2012
But What Is It Good For?
"I confess that in 1901, I said to my brother Orville that man would not fly for fifty years... Ever since, I have distrusted myself and avoided all predictions." -- Wilbur Wright, 1908
After a few muses from @kcwxguy I found myself knee deep fiddling with Google Visualization. It soon became a fresh way to visualize my forecast model data based on LRC theory. The winter of 2012-13 begins my third year following the cycling weather patterns of the northern hemisphere. An eagerness to learn has grown into a passion. From a once inconceivable backyard snowfall forecast in 2010, to a winter outlook for Oshkosh in 2011, to seasonal forecast numbers for the entire Midwest in 2012-13. The idea has certainly grown on me.
As we get "pattern change" shoved down our throats the next two weeks I want to reiterate the fact that this "pattern change" is not something new to the eyeballs that have been following the cycling patterns of 2012-13. The particular pattern has been projected for months. It is evident in all of the trends for the cities that will be affected. I will focus on temperatures. Below is a screenshot animation from the OSNW3|WxClimate site of a Google Geo Chart depicting the maximum temperature deviation for the Midwest cities selected for trends this season. There are three images in the animation. An image for Dec 1-14, Dec 15-28 and Dec 29-31. Clicking on the image will take you to the maps website for a larger view.
What isn't on the media radar yet, well at least I haven't seen it, is the extreme warm-up that could take place in the Midwest the last few days of 2012. Below is another screen capture of the OSNW3|WxClimate site. This Google Motion Chart is showing all of the cities max temperature departures from average for the month of December. By following the thick blue line it is evident something warm will be recurring near the end of the year. The same idea is shown above in the third frame of the animation.
(From the Event Mapping page click on "Motion Chart" to use the chart yourself)
We all know a weather prediction made one month out is seldom followed through to fruition. Most of us will not trust a computer generated forecast more than 3 days out. A major asset of the theory, in my opinion, is it's ability to predict significant "pattern change". Give or take a day, I am confident this will all take place, even before the populous had hints from long range computer models.
If there are any questions or thoughts on my research and analysis of the LRC or how I presented the material just let me know in the comments section of the blog. Thanks for reading!
After a few muses from @kcwxguy I found myself knee deep fiddling with Google Visualization. It soon became a fresh way to visualize my forecast model data based on LRC theory. The winter of 2012-13 begins my third year following the cycling weather patterns of the northern hemisphere. An eagerness to learn has grown into a passion. From a once inconceivable backyard snowfall forecast in 2010, to a winter outlook for Oshkosh in 2011, to seasonal forecast numbers for the entire Midwest in 2012-13. The idea has certainly grown on me.
As we get "pattern change" shoved down our throats the next two weeks I want to reiterate the fact that this "pattern change" is not something new to the eyeballs that have been following the cycling patterns of 2012-13. The particular pattern has been projected for months. It is evident in all of the trends for the cities that will be affected. I will focus on temperatures. Below is a screenshot animation from the OSNW3|WxClimate site of a Google Geo Chart depicting the maximum temperature deviation for the Midwest cities selected for trends this season. There are three images in the animation. An image for Dec 1-14, Dec 15-28 and Dec 29-31. Clicking on the image will take you to the maps website for a larger view.
What isn't on the media radar yet, well at least I haven't seen it, is the extreme warm-up that could take place in the Midwest the last few days of 2012. Below is another screen capture of the OSNW3|WxClimate site. This Google Motion Chart is showing all of the cities max temperature departures from average for the month of December. By following the thick blue line it is evident something warm will be recurring near the end of the year. The same idea is shown above in the third frame of the animation.
(From the Event Mapping page click on "Motion Chart" to use the chart yourself)
We all know a weather prediction made one month out is seldom followed through to fruition. Most of us will not trust a computer generated forecast more than 3 days out. A major asset of the theory, in my opinion, is it's ability to predict significant "pattern change". Give or take a day, I am confident this will all take place, even before the populous had hints from long range computer models.
If there are any questions or thoughts on my research and analysis of the LRC or how I presented the material just let me know in the comments section of the blog. Thanks for reading!
Subscribe to:
Posts (Atom)