Turbulence Ahead

IN 2004 BLIP users will find many changes from 2003's BLIPMAP/BLIPSPOT forecasts.  Overall the changes are necessary for future betterment, though several individual consequences will not be for the better.  These changes may be discomfiting to some users (particularly those with weak stomachs ).  In several respects BLIPMAP convenience has diminished in favor of providing additional information.  While I have provided information on these changes as they occurred in many "news" items on the forecast index and viewer pages, for user convenience I wanted to gather together the main changes in a singe webpage.  I have attempted to provide a summary description and links to further information when available.  I plan on updating this page on occasion and will use the "news" items to let user know about such changes. 

The website domain name and the locations of many forecasts have changed

      So a new cookie and changes in browser bookmarks will be required.  A forum posting provides new website details and forecast locations

Two model forecasts are now available: RAP and NAM

      So choices and evaluations must be made.  There will be differences, sometimes very significant, between the two forecasts due to the inherent uncertainties of atmospheric prediction [DrJack sez: Meteorology is not engineering].  Those who like certainty will be nonplussed - though I don't know of many glider pilots who have found certainty while flying in the atmosphere!  Each model has strengths and weaknesses - neither is always "best".  Having two different forecasts will allow a savvy pilot to obtain a better forecast than can be provided from a single forecast.  At present, however, how those model strengths/weaknesses translate into actual forecasts is not known due to a lack of prior experience - and in any case is likely to depend upon the forecast location and forecast parameter.  No one will begin 2004 as a fully "forecast savvy pilot" - while last year's experience will be useful, the NAM model is entirely new and this year's RAP will not perform exactly as last year's RAP.  The necessary experience will take time and effort to develop, but one has to start somewhere.  For 2004 those with an interest in the atmosphere will need to see how the RAP and NAM forecasts each validate against actual flying experience.  Those who have trouble coping with RAP-NAM differences can simply look at forecasts from a single model and be happy (sometimes ignorance can be blis)s. 
      The most important practical differences between the RAP and NAM models are that the NAM forecasts out to longer times while the RAP is available at more times and is updated more frequently.
      More info about reasons for NAM BLIPMAP development and NAM vs RAP model comparisons

Forecasts from "new" RAP model will differ from the "old" RAP model

      During its development BLIPMAPs had been privileged to use data from the "research" GSD laboratory, but BLIPMAPs have now evolved into an operational status and will instead be using forecasts from the operational RAP model.  While the models themselves are identical, they do have different initializations/assimilations and in the atmosphere small changes can on occasion produce relatively large changes.  But for that matter all RAP models have changed from last year due to model "improvements" so this year's RAP would not in any case produce results equivalent to last year's results under identical conditions - such changes are considered to be a feature.  One consequence is that last year's experience with RAP BLIPMAPs cannot be used unequivocally when evaluating this year's BLIPMAP forecasts. 
      In addition, on a practical level (1) the operational RAP forecasts only go out to 12 hours, so forecasts will not available the evening before, and (2) at present operational 6 hour forecasts are not being produced from the "non-3-hour" runs, so only 12Z, 15Z, 18Z, 21Z, 0Z, and 3Z forecasts are available. 

Forecast availability problems will initially be more problematical

     A fundamental issue is that BLIPMAPs now are in competition with all other public Weather Service users for access to the public operational servers, which are heavily overloaded.  Forecast availability will be impacted, particularly for the NAM model since it's files are 14 times larger than the RAP files and require considerable download time. At present this impact is still being evaluated.  In addition, I have had to greatly alter my program to find and download both the RAP and NAM model files and any alteration can introduce an unanticipated bug.  Further, the sorts of occasional problems that occur on the operational server are still uncertain and so not yet treated by the program - that will require additional later alterations.  The total effect will be to make the times of forecast availability more uncertain, particularly for the NAM model.  However, these problems should diminish with time as problems become better defined and are resolved.  [FYI, it is interesting to note that on the GSD server the time at which a given forecast would appear varied, but once it was available the time to download it was a constant; on the other hand, for the NCEP server the availability time of files is fairly constant, but the time required to download it is highly variable!]

Subscription will be required to obtain most forecasts

      Those who are more interested in their pocketbook than in soaring forecasts will be disappointed.  More info about this season's subscription plan.

PS:  If you think things are turbulent on the user side of BLIPMAPs, you should see things from this side!