Aerology

Roger has been nice enough to provide Aerology a page on Tallbloke’s Talkshop to keep a recent status on the Aerology forecasting method accuracy evaluation and inclusion to ForecastWatch. Roger and I have been discussing our collaborative research for the past two years and I’m excited to share the status of the forecast’s accuracy evaluation as things develop.

The forecasting method behind Aerology is based upon the interlocking cyclic pattern in the inner planet returns, and lunar orbital dynamics cycles, that is also synchronized to the magnetic rotation of the sun. I have been generating forecast maps by this methodology since 1990, but the current evolution of the website has been posted online since December 27, 2008.

Eric Floehr recently contacted me for inclusion into ForecastWatch – a service dedicated to helping weather forecasters evaluate their forecasting methods in relation to other forecast services. I have long wanted to get an independent evaluation of how the analog forecast I generate from past cycles of raw data from 19, 37, and 55 years ago compiled together in regard to actual repeatability to show how the cyclic pattern does compared to the numerical models that always fall apart in 7 to 10 days simply because they do not consider any of the lunar tidal forces as important.

It has been my experience from looking at these patterns since 1990 that they should be at least equal or better than the 5 day out forecast the models generate. IF they can beat that level of accuracy – then it would to me prove that the repeating patterns are the XX% of the drivers behind the weather.

With the basic idea of mine that these forces and their patterns should be incorporated into the numerical models, to take advantage of the increase in length of time they will then be accurate so as to give farmers and construction workers, as well as severe weather warning forecasters more time to change plans or issue watches, warnings, and evacuations, to save many lives, as well as much property damage, and crop loss.

The forecast data he’ll be evaluating was forecast (using the Aerology analog forecasting method) 3 years ago. To have daily forecasts with any kind of accuracy out past two weeks will be a first for the world as far as main stream meteorologists are concerned. From further evaluating the results he gets I will be guided in my upgrade of the forecast method and further generation of longer series of daily maps.

Currently, we’re waiting to hear back on the incorporation of the data into ForecastWatch and the analysis of forecast’s accuracy evaluation – these processes should be completed within a month or so. I’ll continue to post updates to this page as the analysis develops, the evaluation is released and as we explore the possibility of incorporation of these patterns into numerical models.

Thanks again to Roger for giving me the opportunity to share this exciting development on Tallbloke’s Talkshop. Looking forward to posting news as it becomes available.

For more information on Aerology and the analog forecasting method behind it, please visit aerology.com.

New Map format 1 degree contour line 32F is white contour line, the composite first then the four past cycles to compare how repetitive the cycles might be.

new map foremat tmin composite

http://tallbloke.files.wordpress.com/2011/01/20120131_tmin-first2-1-1993.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmin-second12-26-1974.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmin-third1-11-1957.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmin-fourth1-28-1939.png

Maximum temp maps composite first then the cyclic components start below.

http://tallbloke.files.wordpress.com/2011/01/20120131_tmax.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmax-first2-1-1993.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmax-second12-26-1974.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmax-third1-11-1957.png

http://tallbloke.files.wordpress.com/2011/01/20120131_tmax-fourth1-28-1939.png

Posted these to show how much reparation there is between cycles of daily weather patterns 6558 days apart. Now able to post color scale for temperatures at this time but will update asap.

These maps added January 6th 2012.  by Richard Holle

 

Comments
  1. Heard from Eric today the problems he is having with the data analysis comes from the use of all valid data from all stations on line in the past cycles, due to the moving and great dieing of the thermometers only 869 stations are on his “current list” and re-identification of all past stations relative to the current accepted location closest to the originals it rather time consuming.

    “It also means that the same forecast point isn’t always used, but that likely doesn’t matter too much with your forecasts, so long as they are close. But it would have been easier to process had their been consistent lat/longs across files.”

    “The second, and more difficult lat/long problem, is that in the same file there are often the exact same lat/long but with different forecast values. Take “39.033334,-76.933334″, just north of Washington, DC, for example. It often appears many times in a single file, often with different values. Just as one example, for March 2, 2011, the forecast high could either by 71 or 72, and the forecast low can be either 34, 37, 39, or 41. How should I deal with cases like this?”

    It seems that in the original data sort and consolidation to the tables much of this type of problem was averaged out, but some of the District of Columbia area got missed, along with the whole state of Pennsylvania.

    “The last thing, for now, is that did you know you don’t have any points, in any of the files, that are in Pennsylvania?”

    That’s what QA is supposed to be about, finding errors and correcting them, I have made an appointment to talk to the original programmer, about the inclusion of more standard methods of data processing, and a more complete list of suggestions from Eric, for consideration in the rewriting of the software to remove errors and bugs we have found. I originally wanted to use the densest amount of data possible, by using all of the stations available from the entire data set.

    “Because of the state of the files, it’s going to take a lot of additional processing to generate forecasts for the verification points I have observation data for. Any thoughts you have on making that easier certainly would make the accuracy calculations go easier.”
    “Eric”
    I want to include Canada and Alaska stations in the 2nd generation set of maps, I am much encouraged by the depth of QA help I am going to get out of this evaluation process. Ryan the programmer and I will try to sort out the repairs for the data errors found already and come up with a better process that better represents “Standard best practices” now that we know what they should be from our first attempt four years ago, with out good QA feedback.

  2. Ryan is loading the Canadian data into the tables this weekend and the 0″ precipitation problems is an easy fix. To change “less than or equal to zero = null” just remove the “than or = to” part of the CSV generation code line. I need to find a complete data set for as many stations as possible for the UK to do that forecast and Australia as well.

    The reference frame and required mask to the outline of the additional maps is simple enough, once the raw is tabled in the same format just an edit to the reference frame of the map set in question is all that we need to change. Have a max of 2TB hard drive space as of Friday on the server available for my use.

    Will be changing the contour line temperature resolution to 1 degree F instead of 10 degrees F, Just need to make the new color graduation file to show the new scale, and change the map draw resolution from 10 to 01.

  3. Canadian data is now all extracted and tabled, we will be adding Alaska and Canada as one separate map with its own pages, to augment the USA maps, Australian Data has been found and will be purchased from the BOM as soon as it is updated to include the first quarter of 2011. [first or second week of June it will ship].

    working on adding an additional 6558 day long cycle starting in 1938 as it might better reflect the slower solar cycle we are in. In the process of tweaking out the data sorting/compiling process of making the CSV files for the maps to accommodate the changes needed due to the additional learning I have been doing in the past four years.

    Still working on the number of nearest neighbors used in the griding process to best show the detail in precipitation patterns, as well as the UHI around cities and cooler areas around large lakes and reservoirs in the more detailed temperature maps.

    Things are still slow going because of the massive data base, total of all reporting stations with valid data for each day for all four cycles will be used to generate each daily map, if the next day two stations have no records the maps will reflect a slight loss of resolution in the areas of missing data.

    My opinion is that most of the earths surface is not a prime siting area for a weather station, so reality is that if the readings are with in the +/- 3 degrees of the actual temperatures forecast, then that seems to me to prove that the cycle repeats as well as the data records can provide, and this will be as good as it can get.

    I will be happy if I can average +/- 5 degrees of the actuals for the forecast days, because that is as good as the NWS gets out to 3 days into the future. Waiting for the update from the programmer on next steps to work on….

  4. [...] HomeAbout meAerologyBlog rulesHall of fameHall of ShamePredictions!SuggestionsTagline data [...]

  5. Had a few problems for Canada, with the conversion of metric to SAE upon proofing the data I found we had forgotten to change the decimal placement to get hundredths of inches instead of whole inches into the existing format so could see almost no Precip was still reduced by 1/100 of the real value, and 1/10 of the real snow values.

    Recalculated the amounts post conversion and we are now finishing up the retabling of the proofed data in the right format. Hoping to get CSV files grids, and Map samples at the new higher resolution to post on here ASAP. (By the weekend)

    Got the Australian data set ordered and shipped yesterday A$776.50 or US$ 831.00 for all three sets Temperatures, Precipitations, and an additional set of dew points, wind data, and surface pressures.

    I think the dew point plotting will carry extra info on the specific heat content of the air masses, I expect to better follow current than just the air temps do. Might take a couple of months to get the Australian data made into maps, but do plan on starting showing of the forecast maps from about June 2010, with the same color scale format actuals from June 2010, until the end of the data set April 2011.

    I want to display the two sets, the composite forecast and the actuals for the same dates, side by side, then switch to just the forecast composites past that date. I will keep expanding the outlook into the future for up to 10 years out, for all three the USA, Canada, and Australia, as I am able to process and post maps into the future.

  6. Sample screen shots of the new map resolution, (for 2-14-2011 Low temps) showing the increased amount of detail available in the data base that was not originally obvious, down to the surface land use and topographical contouring that creates micro climates that mankind has taken advantage of to urbanize and farm.

    I think this will greatly enhance the usability of the forecasts for farmers, ranchers, construction workers moving dirt, forestry services, and mining operations.

    http://research.aerology.com/project-progress/map-detail/

  7. tallbloke says:

    Very nice maps!

  8. Got an update from Eric today, we have both been busy with our own projects and he has supplied me with a quote for the forecast setup services and a years subscription to his service, described;

    Where we stand on the progress;
    “As for delivery, I could do an analysis of the forecast format, etc.
    like I did last time to make sure it’s suitable for import by the end
    of September. Then once any corrections are made, do the full
    setup/import code. If that went well (there were few corrections
    needed, etc.) I could import September 2011 before October 15, which
    would allow you to then see the September aggregations and data in
    mid-October.”

    working with huge data sets makes for slow progress, when I have limited funds available, but slow progress is better than no progress. The programer and I are still working out the details of the masking of the maps, and the rebuilding of the site layout to allow all of the additional maps in the higher resolution format. I am hoping to bring on line all of the additional features and the ongoing validation of the forecast accuracy by the first of the year 2012.

  9. tallbloke says:

    Hi Richard, I watched the demo. It looks like a very powerful tool for checking your hypotheses. Solid progress, well done!

  10. I think the evaluation process will point out the short term shifts in the solar/lunar patterns with the synod conjunctions with the outer planets. I expect the resultant departures to follow the rise in temps to conjunction then fall for about 5 days following. As well as the locations of the shift in precipitation patterns and amounts. It is my hope that the evaluation process will help me build a workable algorithm to incorporate the outer planet patterns of the disruption in the inner planet lunar declinational tidal patterns that will define the other 20% of the temporal disruptions, but more extreme excursions of the natural variability patterns.

  11. Got a look at the output from Eric’s evaluation of the forecast abilities of the repeating patterns of the inner planets, starting yesterday for the month of October. It is clear [from even the state wide scale] of the shift in phase and spatial/temporal relationships that the signal from the interference of the outer planets shows up in the digital data as well as the GOES satellite photos. Most of November 2011 until mid April will be outer planet interference free from conjunctions, which should give me a nice calm period to see how well the cyclic pattern works as a fore cast.

    More details when I get permission to quote the results.

    Hope to have new detailed maps going online for the USA by middle of the next month.

  12. tallbloke says:

    Exciting news indeed Richard! I think you should start making a few specific predictions here and appraising the results after the fact. Success there will definitely start arousing more interest. Correlation-prediction-validation. This process can put your hypothesis on a strong footing without the need to fully explicate mechanism at this stage.

  13. Here is a sample of the data, code still being tweaked into final form.

    Uploaded the csv files for the new maps October 2011, through end of January 2012, for North America to a dropbox file available to all on the internet. Sent these to Eric for processing for accuracy evaluation, will update again when I hear from him.

    http://dl.dropbox.com/u/37184697/CSVFiles.zip

    38 MB zip file in total for download expands to 173MB as 3,049 csv files of the map composite for North America.

    These result are from the days back from today search results I am now using. the numbers are different for the precips than the temperatures, due to the seasonal shift in temperatures is much more sensitive to 27 day shifts in calendar dates from different past patterns than the location and by the amount of precip changes by smaller amounts is due to dew point shifts, even though the air mass movements globally are still much the same flow patterns.

    The look back dates used for Precip are 6,992, 13,550, 20,108, and 26,666 before the present cycle date a forecast is sought for.

    The look back days used for Temperatures are 6,938, 13,496, 20,108, and 26,666 before present date of forecast inquiry.

    These are for the new map format, with the color scale for the new 1 degree increment temperature contours. These new csv files are the latest data I am using to generate the new maps, and have been working out the bugs in the automated process for the new site revision, the process has been ongoing now for 8+ months and we are finally to the point of doing the site layout graphic rework and text up dates for the about page, and new features for the home page and additional Canadian maps pages.

    I made a compressed file for the Jpg maps themselves but the size is 2.65 Gig and too big to use dropbox (max 300Meg files), Rog if you have a ftp site where I could upload the maps (2.7 gig of space needed) I could do that instead.

    Did manage to upload a set of Tmax and Tmin maps for 2012-01-31 forecast with the respective four past cycles of data for visual comparison of the repeatability between cycles. Feel free to clean up any mess I have left on your site Roger and thanks for the page space.

  14. Tornado prediction for spring with mention of heavy snow and ice storms as well, posted on WUWT.

    In response to http://www.nasa.gov/topics/earth/features/se-us-severe.html
    http://wattsupwiththat.com/2012/03/01/tornado-outbreak-tracking/

    Richard Holle says:
    March 1, 2012 at 6:47 pm

    Once again the Lunar declinational tidal effect is responsible but goes unmentioned, The moon was maximum North declination on the 1st of March, the solar declination seasonal tide is incoming from the South adding to the effect and making the resultant tropical air mass surge two days sooner than the usual, peak production on the day of Maximum North lunar declination and three days after.

    I have had daily forecast maps for this expected precipitation posted for 51 months now;
    http://www.aerology.com/national.aspx

    Details on how it works are posted on the site, in the blog/research section.

    You can watch the incoming lunar tidal bulge sweep in from the Southwest in their short movie, and the back side more polar air mass brings in the negative static/ionic charges that gave added power to the temperature front to drive the condensation high enough to generate the tornadoes.

    Over the next three days as the fetch of moisture slides East across Texas into the Gulf states, and the moon starts to head South again, Just as we are having a heliocentric conjunction with Mars on the 3rd, this is very likely to bring on another round of tornadoes.(they got that part right anyway) see my maps for these days as well.

    Last year the big outbreak was enhanced by the heliocentric conjunction with Saturn, on the 3rd of April, which will be occurring on the 15th of April this year, so you can expect more outbreaks to occur from the 4-10-2012 Maximum South lunar declination and four day after window, another much larger 2 or 3 day burst as the moon crosses the equator on 4-17-2012. Then the heavy action through the end of the month of April, ending in last hurrah of big snows into the first week of May. Appalachian Ice storm seems to be on the 3rd through 5th of May, buy your replacement power poles early.

  15. Higher definition global circulation video showing the lunar declinational tides in the atmosphere, three cycles from 10 degrees North of the equator to max North,then back through the cycles to the same point again. Christmas of 2009 through March 8th 2010.

  16. Hans says:

    Richard,

    Your work is very impressive. Just to do cyclic patterns recognition on images as you are doing will lead forward a lot even if you don´t understanding a bit of what is driving the atmospheric motion involved.
    Your emphases on the lunar sidereal period is spot on and a key for investigating tidal energy transfer in the atmosphere (wind and precipitation generation)
    Your 6558 cycle is new to me. What is the reason you don´t use the SAROS cycle?

    Holle cycle
    222 Tmsyn = 6555.79 days
    240 Tmsid = 6557.20
    238 Tmanom = 6557.98
    241 Tmdrac = 6558.14

    SAROS cycle
    223 Tmsyn = 6585.32
    241 Tmsid = 6584.52
    239 Tmanom = 6585.54
    242 Tmdrac = 6585.36

    A question: According to what I have learnt the tide at the South Pole (Weddel Sea I guess) is only once a day. Do you know if it is caused by the sidereal lunar tide or the synodic lunar tide? I would be grateful for your answer. I would also be very interested in what period you consider be the QBO correct average QBO period and how it relates to the celestial bodies you consider vital for its creation?

  17. tchannon says:

    Rhetorical wondering: how does this relate to what Piers Corbyn thinks goes on.

  18. Hans;Your 6558 cycle is new to me. What is the reason you don´t use the SAROS cycle?

    Holle cycle
    222 Tmsyn = 6555.79 days
    240 Tmsid = 6557.20
    238 Tmanom = 6557.98
    241 Tmdrac = 6558.14

    SAROS cycle
    223 Tmsyn = 6585.32
    241 Tmsid = 6584.52
    239 Tmanom = 6585.54
    242 Tmdrac = 6585.36

    There is a four fold symmetry to the Rossby wave and jet stream driven by the “240 Tmsid = 6557.20″ declinational tidal movement, if I use the 241 cycles in the Saros cycle they are out of phase and the interference kills the synchronicity of the weather data, by using the 240 cycle period the signal to noise ratio goes up 14 times, compared to the Saros cycle. This effect is quite visible if you look at the different periods I sorted tornado production numbers by;
    the Metonic, Saros, and my 6558 day cycle

    http://research.aerology.com/lunar-declinational-affects-on-tornado-production/

    The differences in the location in the incoming MPH per each of the four repeating cycles shifts around and the average of the four together is almost useless in forecasting repeating global weather patterns. The movie segment above shows three of the four patterns with every other being more closely the same, (should be a better match between the end ones) than to the successive ones, much like meanders on a stream.

    The shapes and repeatability of the waveforms varies from region to region due to the standing wave influences of the North/South mountain ranges, Andes, Urals, Rockies, and the large block of the Himalayas cause disruptions of the laminar zonal flow down stream East, but the four folded pattern shift is also seen in all zones, most pronounced just East of the blocking mountains of interest.

    I have not heard before of the single daily tide at the poles, I can offer no answer to this question.

    The new article on the Mexico earth quake today the 20th is more in line with planetary studies I was working on 20 years ago before I found a large data base for USA COOP observations and I got totally distracted into looking at the weather effects, as I thought a good long range weather forecast had much more promise for helping all of humanity feed and shelter themselves.

    But what I was finding then is the perigee/apogee cycles and the declinational tides were key to pinpointing latitude of expected quake production, with surges in larger quake production with in 30 degrees of a heliocentric conjunction with one or more of the outer planets. All of that data and results, are locked up in an old Mac, (backed up on a shoe box full of 3.5 inch floppies, some HD, most not) as I haven’t turned it on for 15+ years now.

    QBO period is IMO due to the shift in the interference between the daily solar tides, lunar phase, and declinational tidal patterns acting on the body of the Pacific ocean surface in the lee of the Himalayas, which allows the tidal effects to build cumulative and destructive interactive trade wind and zonal wind flows, resulting in the ENSO effects and the Madden Julian oscillation being just named subsections of the total lunar tidal effects, modulated by the other planets gravitational and electromagnetic effects, that act to hide the clear pattern that would be seen with out their confounding influences.

    [Sorry for the slow response, checking new Austrailan maps I just got the data tabled and CSV files pulled for the four cycles and the composites and am proofing the data base and testing patterns of weather interactions I have not seen before, getting ready for the addition of more forecast areas to test my hypothesis on, not been checking up on other threads much the past week.] Richard

  19. Richard Holle says:
    April 13, 2012 at 1:06 pm

    http://wattsupwiththat.com/2012/03/01/tornado-outbreak-tracking/#comment-909920

    Seems to still be as valid now as then…
    Richard Holle says:
    March 1, 2012 at 6:47 pm

    Once again the Lunar declinational tidal effect is responsible but goes unmentioned, The moon was maximum North declination on the 1st of March, the solar declination seasonal tide is incoming from the South adding to the effect and making the resultant tropical air mass surge two days sooner than the usual, peak production on the day of Maximum North lunar declination and three days after.

    I have had daily forecast maps for this expected precipitation posted for 51 months now;
    http://www.aerology.com/national.aspx

    Details on how it works are posted on the site, in the blog/research section.

    You can watch the incoming lunar tidal bulge sweep in from the Southwest in their short movie, and the back side more polar air mass brings in the negative static/ionic charges that gave added power to the temperature front to drive the condensation high enough to generate the tornadoes.

    Over the next three days as the fetch of moisture slides East across Texas into the Gulf states, and the moon starts to head South again, Just as we are having a heliocentric conjunction with Mars on the 3rd, this is very likely to bring on another round of tornadoes.(they got that part right anyway) see my maps for these days as well.

    Last year the big outbreak was enhanced by the heliocentric conjunction with Saturn, on the 3rd of April, which will be occurring on the 15th of April this year, so you can expect more outbreaks to occur from the 4-10-2012 Maximum South lunar declination and four day after window, another much larger 2 or 3 day burst as the moon crosses the equator on 4-17-2012. Then the heavy action through the end of the month of April, ending in last hurrah of big snows into the first week of May. Appalachian Ice storm seems to be on the 3rd through 5th of May, buy your replacement power poles early.

  20. CSV files for the composite map data from February through July 2012;
    http://www.aerology.com/Feb2012-Jul31.rar

  21. Over in suggestions;
    Ninderthana says:
    May 12, 2012 at 3:45 pm (Edit)

    “”I think that you might be missing a very important paper here:

    http://www.agu.org/pubs/crossref/2011/2011JD016598.shtml

    Monthly lunar declination extremes’ influence on tropospheric circulation patterns

    JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, D23121, 5 PP., 2011

    Short-term tidal variations occurring every 27.3 days from southern (negative) to northern (positive) maximum lunar declinations (MLDs), and back to southern declination of the moon have been overlooked in weather studies. These short-term MLD variations’ significance is that when lunar declination is greatest, tidal forces operating on the high latitudes of both hemispheres are maximized. We find that such tidal forces deform the high latitude Rossby longwaves. Using the NCEP/NCAR reanalysis data set, we identify that the 27.3 day MLD cycle’s influence on circulation is greatest in the upper troposphere of both hemispheres’ high latitudes. The effect is distinctly regional with high impact over central North America and the British Isles. Through this lunar variation, midlatitude weather forecasting for two-week forecast periods may be significantly improved.””

    I have been in contact with the author of this paper;
    I would be happy to discuss and possible collaborate on lunar declination related climate stimuli.

    If you haven’t been able to obtain the article in its entirety, I have attached a pdf.

    I am currently working on my Ph.D at ASU working with Dr. Cerveny and occasionally with Dr. Balling. My dissertation’s topic is concerned with solar variation and its different avenues of interaction with the climate system.

    The article was a re-imagining of my Master’s research examination, which looked at Lunar declination vs. Phoenix’s climate. Its findings convinced us to look at a larger data-set, the NCEP Reanalysis.

    I am currently teaching a few Labs this semester at ASU. They conclude at the end of the month. I should be around for most of the summer for a meeting.
    Otherwise, please email me with ideas, questions, interesting articles or anything else for that matter. Let me know how I can be of help.

    Dan Krahenbuhl
    School of Geographic Sciences and Urban Planning
    Arizona State University

  22. I have a staging site up for the reviewing of problems before going live with the new site revisions, latest note from developer “BTW, you are generated through 9/1 now, so it is looking like almost 2 months per day in processing. So around 9-10 days to finish out the 2 years of maps. I will spin up a spare laptop tonight to help with some of the later maps that aren’t already queued up on the desktop.”

    http://staging.aerology.com/

    Take a look and any comments on needed improvements will be considered, but is fairly much cast in stone on the basics, detailed local forecasts will be added later when software can be located for postal codes for Canadian and Australian locations. Dual sae and metric scales will be added to the precips when ready.

  23. Web site just went live on the updated maps and expanded areas, there are no local area forecasts yet they will be added as soon as the site gets moved to a better hosting server, that can handle twice the map load than the current server can. In about two weeks or less, the staging site is off line now.

  24. As the discovery process and correction of undetected errors is progressing, there will be improvements to what you see today… The auto griding software chose 20 degree search radii for the nearest neighbor selection process instead of the 5 degree I asked for, as a result the definition is smoothed about three times more than I wanted, losing much of the temperature UHI detail I wanted to show and washing out the precipitation details, spreading lower amounts way out from the area rain or snow fell (especially in Australia) and lost the higher amounts completely, so it looks like they never get any significant amount of rain any where.

    This will all be rectified when I am moved onto a larger server with 16Tbt HD and 96gig of ram, we will rerun the generation of the maps from the same csv data files, and overwrite these maps with the more detailed ones with higher contrast features, in about a week or two.

    At that time we will be returning the local map features currently withheld due to lack of server space.

  25. Forecastwatch service has basically shown that the accuracy is ~50% lower than the 1 to 3 day forecasts by the NWS but closer to their 3 to 5 day out forecasts. For the past 7 months the Aerology.com forecast has been with in 3 degrees ~23% of the time both highs and lows for the days. Compared to the average of other forecasters (1 to 3 day forecast hit rate with in 3 degrees) of ~48% to 58% with them changing leaders almost every month.

    Eric compares ~875 GHCN (city/airport) stations with my data for the same stations to get his results, so is testing 1 out of 19 sites I use (total ~17,080). Most of the sites I use that he does not are rural, ag stations, or coop elevators, farm implement dealers, etc. so the results that my forecasts run on average ~-3.25 degrees F bias off from the current cycle seems to be UHI effects. With the increase in definition (when the corrected maps get posted) you can see areas of heat islands that may be due to natural topography, which the original inhabitants took advantage of, and the incoming settlers found to their liking as well.

    I think it is highly likely that areas that were naturally sheltered had their own HI effects and people settled these areas first, as they expanded into the modern cities (with preference for traffic access, and steady water supply), you will continue to see UHI effects in these same areas. I think there was a predisposition for Heat Island effects in the areas where they are seen today, so man has only enhanced it, not made it from scratch. There are lots of urban areas seen in these maps with lower temps in parts of the outer areas around Midwestern cities, like St. Louis, due to wide scale landscaping of the out lying neighborhoods, compared to current or former farm ground, but still have UHI effects in the downtown areas.

  26. Moved onto the commercially hosted server yesterday DNS transfer to new location still fu*ked up or lagged out at this time only the research.aerology.com blog section is active at this time (did not have it hosted on the old or on the new server).

    When I get access to the new server we will be running tests with 6 degree, 7 degree, and 8 degree nearest neighbor search radii for best fit showing most area coverage with highest resolution. it is a trade off between smearing data clear across 20 degrees or 1200 miles (like GISS does it in the Arctic but infilling all areas), or showing higher resolution clearly demarcated local details of greater density, and loosing (the grid sample areas each 3 miles square) areas not with in X ,6,7,or 8, degrees (which ever I decide to go with that is most useful) of three reporting stations, to get a triangulation plotting of contour that is valid.

    This is almost the same problem the ACORN data set in Australia ends up with homogenizing the nearest 40 stations to get “corrected adjusted data”, which ends up raising the average of the total adjusted stations because all of the cold areas from the Alps in the South East get washed out on the station homo level. Will show the changes in the data base as the process of homogenization of the real data started as soon as I can get the server running on track and start pulling out maps for separate dates over long periods to make into comparison movies.

  27. At this time the new server is on line and for a brief period of time [maybe a couple of days] (until we figure out the optimal search radius and reload the maps with the improved series.) Some sample maps with a 5 degree search radius for North America and a 2 degree search radius for Australia are posted for the dates starting on http://www.aerology.com/Home/Index?location=Australia&mapType=Tmin&date=3/9/2014, they are at 20 degree search radius prior to this date.

    On the 10 and 11th both are shown at 6 degree search radius http://www.aerology.com/Home/Index?location=Australia&mapType=Tmin&date=3/11/2014 . Then they continue on at 2 degrees for Aus and 5 for N A. In case you are interested in how it affects the data smoothing, smearing of the temps, precips etc.

  28. The trials for the search radius for use were completed with the dates for 6 degrees March 10-15 2014,
    7 degrees March 16-20 2014, 8 degrees March 21-25 2014, and 9 degrees March 26-30 2014.

    I will be going with a search radius of 8 degrees, so we will be reloading all the maps on the site overwriting those on line today with the new ones, as they are generated from the same csv files, same data sets, with the new gsb file overlay.

  29. adolfogiurfa says:

    What is it happening in front of NY & Canada?:
    http://weather.unisys.com/surface/sst_anom.gif

  30. There is a calm spot there that is allowing the sea surface to warm by virtue of not being cooled by winds, it has been sluggish in moving as the Jupiter enhancement of the lunar atmospheric tidal bulge has been building up to the 3rd today but will be flushed out of the area as the effect sweeps through this week, can be seen here at this time;
    http://www.atmos.washington.edu/~ovens/loops/wxloop.cgi?wv_east_enhanced+50+-update+3600

    You can still see the cold streak from Sandy along the eastern sea board maintained by the Eastern Canadian cold air flow out of the Hudson Bay area passing through.

  31. adolfogiurfa says:

    @Richard: It is good to learn something new. Thanks!

  32. RichardLH says:

    @Richard.

    Do you have a reference for the gravitation field modulation and distribution here on Earth from the Sun and Moon (over time)?

    As a vertical and horizontal change plot. i.e. Tidal forces (in G) both lateral and vertical on a Mollweide projection or a true globe.

    This is, I believe, an important inter body force for which I am unable to find good references.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s