Discrepancies

[Željko Vranić]

Hi!

Although your product iz beyond great I'm having problems with understanding why "Top of usable lift" forecast is different across all three (four) views that you provide. 

For example: Forecast for this Friday 22.5.2009 (solar noon time) for town Koprivnica in Croatia (Balkan, Europe):

- xcskies map with light blue color indicates 1219m as top

- 3day forecast for that point indicates 1100m (red value)

- forecast for route that goes through that point gives around 1600m

- using java applet for SKEW-T - top depends on initial temperature difference of thermal on ground (which value you use: 1.5 or 2?) and also depends on temp-diff between thermal and profile temp that one marks as at least 1.2m/s lift (how to do that anyway?). 

Why this data is so different? 

Ok, xcskies map and 3day are close but route forecasts are way off and I tend to mostly believe in them. Route forecasts seem to be usually much closer to observed conditions (looking at flights that day)

It seems to me that route forecast is very close to heights that you get if one uses SKEW-T with 1.5 degees C starting thermal diff, finding intersection and subtracting around 100m (depending on SKEW-T profile shape, but when this shape is very straight right until final inversion this is good enough method. Although, for my own calculations I use 2degrees diff.

Anyway, maximum recorded heights are most of the time higher than any of your forecast so I'm wondering If you are giving some average value? I wouldn't want an average value because in reality there are many types of thermals and in flight I will most of the time choose the best ones that take me to maximum heights. I don't want this maximum height to be averaged with weaker thermals. I understand that "top of usable lift" should indicate real top value achieved with best thermals.

It would help me to understand how you calculate values for all that three sources of yours (not exact algorithm with carefully tuned parameters which is probably your trade secret - but an indication - what goes into each calculation.)

This discrepancies in results force me to always try to calculate top lift from skew-t data myself (and for that I don't even need your site because applet is originally hosted on http://rucsoundings.noaa.gov/)

I still use xcskies map to compare conditions over different areas, look at cloud cover and wind, but for top of usable lift color I know I always have to add 300-500m.

Another thing that bothers me is that in your forecasts lift always slowly decreases with height (being maximum at several hundred meters above ground) no matter how straight and adiabatic temperature profile on skewt looks that day. This couldn't be right. First, this low lift is very hard to center and not usable much until proper thermal forms from several smaller ones. Secondly, at least in my experience, lift is most of the time (from height where it is fully formed) pretty constant until it starts to approach the inversion. And when top of usable lift is above predicted cloud base - lift seems to be even stronger as one approaches the cloud (cloud suck phenomenon due to increased lift generated by releasing latent heat - it is always present in some amount, not just when cloud is really big).

And last thing - maybe you should clearly indicate somewhere that GFS data that SKEWT applet uses is generated for 0.5degrees (40*50km network) discrete points so choosing a point between this points won't give you interpolated results but result from nearest point. And this nearest point isn't obvious if you don't enable lat/lon grid. Applet itself indicates distance between choosed point and forecast point but distance is in "nm/degrees" (I still don't know how to convert it to km's)

Sorry for so many remarks. I just wan't this great product to be even better than it already is.

Thanks

Regards

Zeljko