WINDIP FORECASTS
WINDIP = Wind Information Prediction
Created by Dr. John W. (Jack) Glendening  

WIND Information Predictions summarize upper-air wind forecasts and give a simple mt. wave prediction.


DrJack sez:
At present WINDIP forecasts are not being emailed, so the forecasts are available only here on the web. 

Users should check each file's header date for currency.
RENO          Latest         Yesterday
Provides NWS model predicted winds at 700 and 300 mb at 12 hr intervals, also indicating when wave is expected based upon Doug Armstrong's criteria. 
BISHOP          Latest         Yesterday
The Armstrong wave criteria is utilized with 700 and 300 mb wind predictions for the Owens Valley.
MOHAVE          Latest         Yesterday
The Armstrong wave criteria is utilized with 700 and 300 mb wind predictions for the Mohave desert. 

UKIAH          Latest         Yesterday
As an experimental forecast, a modified wave criteria is utilzed with 850 and 300 mb wind predictions for Ukiah and a ridgeline which produces wave at Williams.  The 850 mb criteria is uncertain and feedback of the program's successes or failures, so that criterion can be adjusted, should be sent to

SALINAS          Latest         Yesterday
As an experimental forecast, a modified wave criteria is utilzed with 850 and 300 mb wind predictions for Salinas and a ridgeline expected to produce wave at Fremont Peak.  The 850 mb criteria is uncertain so that criterion may later be adjusted based on user feedback of the program's successes or failures.  The present expectation is that "*" markers will indicate wave conditions at Fremont Peak while "+" markers will indicate downwind wave in the mts. northeast of Hollister but this may be optimistic, i.e. likely to predict wave when none exists. 

Waynesboro VA          Latest         Yesterday
An experimental forecast akin to the SALINAS forecast above, utilzing 850 and 300 mb wind predictions for Staunton VA and ridgelines perpendicular to a 320 degree wind.

First Available:  The WINDTIP program is initiated around 7:30 am Pacific Standard Time (1530Z), when all model predictions based upon that day's 1200Z sounding observations become available, and results should be posted within 30 minutes of that time.

DESCRIPTION
    WINDIPs provide upper-air winds at "ridgetop" (either 700 or 850 mb) and 300mb heights - as forecast by the NWS RUC (Rapid Update Cycle), ETA, NGM (Nested Grid Model), AVN (Aviation), and MRF (Medium-Range Forecast) models - and a rudimentary prediction of mountain wave occurrence based upon empirical criteria.  Forecasts for the ETA, NGM, and AVN models are given at 12 hr intervals out to each model's forecast limit (48 hrs for the ETA and NGM models and 72 hrs for the AVN model), allowing results from different models to be compared and so providing an indication of prediction reliability/uncertainty.  The RUC model provides forecasts at 3 hr intervals, allowing better timing of the jet maximum.  The MRF model provides 4 to 10 day forecasts, allowing longer-term planning.  Each model's first forecast line gives the forecast period (in hrs) and date/time, with the corresponding 300mb and "ridgetop" wind predictions on the following two lines.  Each wind forecast consist of three predictions: a wind speed (in kt), a wind direction, and a windspeed component perpendicular to the assumed ridge axis (in kt).  An initial asterisk (*) or plus (+) is also present if the predicted windspeeds meet the empirical wave criteria. An asterisk (*) indicates that the winds are roughly perpendicular to main ridgeline, whereas a plus (+) is used when the perpendicular criteria are not met and the wave is thus more likely to be spotty and occur in unusual locations, e.g. downwind of individual peaks or minor ridgelines having an orientation differing from the main ridgeline.  An initial alert line summarizes the times and models for which wave is predicted within the next 72 hrs
    Please note that the primary purpose of this program is to provide advance warning of wave events rather than to predict wave on the actual day that it occurs (it was originally developed to help me decide when it was worth making a day-long drive to Minden to fly wave).  And it provides no information on the lowest or highest altitudes at which wave will occur. 

Wind Prediction Format
Line1:  WINDIP header line giving station identifier and date
Line2:  Alert summary line listing models forecasting wave for each forecast period
Line3:  Header line giving ridgeline axis direction and script start-end date:time
Line4:  RUC model forecast times ala: ff=mm/dd:hhZ
            where ff=forecast_period(hr)
                        mm=month dd=day hh=hourUCT  [ff=00 for initial analysis]
Line5:  RUC 300mb forecast winds ala: ww dd uu
            where ww=wind_speed(kts) dd=wind_direction(true)
                        uu=ridge_wind_component(kts,perpendicular_to_ridgeline_axis)
Line6:  RUC "ridgetop" forecast winds ala: ww dd uu
Line7-9:  ETA model forecasts
Line10-12:  NGM model forecasts
Line13-15:  AVN model forecasts
Line16-18:  MRF model forecasts
Blank windspeed indicates model forecast found but wind data missing
"---" windspeed indicates model forecast not found

Wave Predictions based upon criterion developed by Doug Armstrong of Reno NWS
Asterisk (*) or Plus (+) precedes windspeed when two mt. wave forecast indicators are met
      (1) windspeed >= 25 kts at ridgetop level
      (2) windspeed >= 65 kts at 300mb
      where both criteria must be met simultaneously for a predicted mt. wave
      (300mb height is roughly 31,000ft)
Asterisks (*) indicate ridge-prependicular components greater than 18 & 46 kt at rigetop & 300mb
      (i.e. when criteria windspeeds are within 45° of the ridge perpendicular)
Pluses (+) are used when the perpendicular criteria are not met.

Notes:
    Predictions for the same forecast period are provided from several models so that forecast variability at a given time will be apparent (a educational aspect of the WINDIP is to demonstrate the variability inherent in meteorological forecasting to those who have been used to relying on a single model!).  The forecast uncertainly is greatest when only a single model predicts wave and least when all models predict wave to occur on a certain day/hour.  Model variability results from differences in the models' resolution, physical equations, and initial conditions.  The fact that there can on occasion be relatively large differences between different model predictions should be noted by those who are used to seeing, and accepting, a single weather prediction.
    The ETA model is considered to be the most accurate predictor at a given forecast time, since it has higher resolution than the other models, but it is known to underpredict windspeeds near 300mb and thus will tend to underpredict mt. wave formation.
    Models which forecast further ahead in time do so by having coarser resolution and thus lesser accuracy. I personally feel that forecasts beyond 5 days (120 hrs) are of little value for wave forecasting, as "teaser" waves often appear and disappear.   Trends in windspeed are expected to be more accurate than are absolute magnitudes of the predicted speeds.
    The empirical criteria used to predict mt. wave occurrence are necessarily simplistic and omit several factors which affect wave formation, notably atmospheric stability effects which can produce trapped waves.  (It should also be noted that the empirical mt. wave criteria were developed for the Reno area and their use in locations with significantly different ridge heights would be questionable.)  The actual behaviour of mt. waves is more complex than can be predicted on a simple "yes-no" basis from winds at two heights, so the intent of the WINDIP is (1) to provide a quantiative overview allowing quick comparison between wind forecasts at different times or for different models and (2) to provide an "early warning" alerting pilots to the likelihood of wave occurrence.
    Additional analysis would be required to obtain more details, such jet stream analysis maps which indicate the strength and horizontal location of the observed/forecast wind jet maximum and thus provide a spatial perspective of the "big picture" at 300 mb; jet stream analyses are available for the AVN model at http://virga.sfsu.edu/crws/jetstream.html and for the ETA model at http://weather.unisys.com/eta/4panel/eta_300_4panel.html
    Also useful is real time data, such as the 1 km resolution satellite photos at http://www.wrh.noaa.gov/satellite/1km/vis1.html, which can confirm the existence of wave prior to a flight.


Link to the latest LWIP mt. wave forecasts
Link to the latest BLIPMAP thermal forecasts
Link to DrJack's home page


Google Search drjack.info