XCSkies and Skysight

[[email protected]]

Up to this point, I've used XCSkies, but I keep hearing more and more about Skysight. I'm trying to determine if I should switch over for the 2020 soaring season. (For U.S. thermal soaring, no ridge/wave.) I've done a trial, but it was only seven days and I didn't really get enough experience with it to determine if I should make the switch.

They both have the HRRR model, but the only other option on Skysight is it's own model which I know nothing about. XCSkies have several models to chose from, of varying vertical/horizontal resolutions.

It also seems to me that Skysight updates the model less frequently the day prior and day of soaring compared to the RAP/HRRR models on XCSkies. That would negate a lot of the benefit of the HRRR model, wouldn't it?

If I'm wrong about any of these assumptions, please educate me. I'm still a neophyte when it comes to using weather products for soaring.

Can someone with knowledge on the models and these websites' methods for employing them describe which is more powerful in predicting actual soaring conditions?

What makes Skysight worth the cost, being twice as expensive as XCSkies?

[[email protected]]

You may be interested in the SkySight webinar we did last year as part of the SSA webinar series. www.ssa/webinars.com

[[email protected]]

My apologies. The 1st address I wrote was incorrect. Please use the following:

https://www.ssa.org/Webinar/TOP%203%20SECRETS%20OF%20SKYSIGHT%20AND%20HOW%20TO %20USE%20IT%20LIKE%20A%20PRO.mp4

[John Foster]

On Saturday, June 6, 2020 at 9:17:57 PM UTC-6, wrote:
> Up to this point, I've used XCSkies, but I keep hearing more and more about Skysight. I'm trying to determine if I should switch over for the 2020 soaring season. (For U.S. thermal soaring, no ridge/wave.) I've done a trial, but it was only seven days and I didn't really get enough experience with it to determine if I should make the switch.
>
> They both have the HRRR model, but the only other option on Skysight is it's own model which I know nothing about. XCSkies have several models to chose from, of varying vertical/horizontal resolutions.
>
> It also seems to me that Skysight updates the model less frequently the day prior and day of soaring compared to the RAP/HRRR models on XCSkies. That would negate a lot of the benefit of the HRRR model, wouldn't it?
>
> If I'm wrong about any of these assumptions, please educate me. I'm still a neophyte when it comes to using weather products for soaring.
>
> Can someone with knowledge on the models and these websites' methods for employing them describe which is more powerful in predicting actual soaring conditions?
>
> What makes Skysight worth the cost, being twice as expensive as XCSkies?

I'm still a neophyte with SkySight, but even so, I prefer it to XC Skies. My experience so far (limited) is that SkySight is more accurate. The colors have an adjustable translucency where the base topography can show through better, and SkySight also has the ability to use Google satellite imagery for the base map, which is quite helpful. You also get a "point forecast" with SkySight for any point on the map, as well as Skew-T charts for any point on the map, if you like to use them. There is a route-planning function with SkySight that you don't get with XCSkies, as well as a route suggestion function with SkySight. All in all, there are many more useful features to SkySight than XC Skies--well worth the money, in my limited opinion.

[Jonathan Foster]

On Saturday, June 6, 2020 at 11:11:17 PM UTC-6, John Foster wrote:
> On Saturday, June 6, 2020 at 9:17:57 PM UTC-6, wrote:
> > Up to this point, I've used XCSkies, but I keep hearing more and more about Skysight. I'm trying to determine if I should switch over for the 2020 soaring season. (For U.S. thermal soaring, no ridge/wave.) I've done a trial, but it was only seven days and I didn't really get enough experience with it to determine if I should make the switch.
> >
> > They both have the HRRR model, but the only other option on Skysight is it's own model which I know nothing about. XCSkies have several models to chose from, of varying vertical/horizontal resolutions.
> >
> > It also seems to me that Skysight updates the model less frequently the day prior and day of soaring compared to the RAP/HRRR models on XCSkies. That would negate a lot of the benefit of the HRRR model, wouldn't it?
> >
> > If I'm wrong about any of these assumptions, please educate me. I'm still a neophyte when it comes to using weather products for soaring.
> >
> > Can someone with knowledge on the models and these websites' methods for employing them describe which is more powerful in predicting actual soaring conditions?
> >
> > What makes Skysight worth the cost, being twice as expensive as XCSkies?
>
> I'm still a neophyte with SkySight, but even so, I prefer it to XC Skies. My experience so far (limited) is that SkySight is more accurate. The colors have an adjustable translucency where the base topography can show through better, and SkySight also has the ability to use Google satellite imagery for the base map, which is quite helpful. You also get a "point forecast" with SkySight for any point on the map, as well as Skew-T charts for any point on the map, if you like to use them. There is a route-planning function with SkySight that you don't get with XCSkies, as well as a route suggestion function with SkySight. All in all, there are many more useful features to SkySight than XC Skies--well worth the money, in my limited opinion.

[MNLou]

I believe you are asking a false question "either / or". Why not both?

I use 2 models and really like it. When they converge, I'm happy. When they diverge, I try to understand why.

Lou

[Ron Gleason]

On Saturday, 6 June 2020 21:17:57 UTC-6, wrote:
> Up to this point, I've used XCSkies, but I keep hearing more and more about Skysight. I'm trying to determine if I should switch over for the 2020 soaring season. (For U.S. thermal soaring, no ridge/wave.) I've done a trial, but it was only seven days and I didn't really get enough experience with it to determine if I should make the switch.
>
> They both have the HRRR model, but the only other option on Skysight is it's own model which I know nothing about. XCSkies have several models to chose from, of varying vertical/horizontal resolutions.
>
> It also seems to me that Skysight updates the model less frequently the day prior and day of soaring compared to the RAP/HRRR models on XCSkies. That would negate a lot of the benefit of the HRRR model, wouldn't it?
>
> If I'm wrong about any of these assumptions, please educate me. I'm still a neophyte when it comes to using weather products for soaring.
>
> Can someone with knowledge on the models and these websites' methods for employing them describe which is more powerful in predicting actual soaring conditions?
>
> What makes Skysight worth the cost, being twice as expensive as XCSkies?

Here is a great discussion and overall approach to weather darta gathering and forecasting

https://www.cloudbasemayhem.com/episode-120-lisa-verzella-and-understanding-xcskies-and-weather-forecasting/

Lisa Verzella is a long time champion and record holder HG and PG pilot. SHe now works for the NWS in SLC.

If you do not listen to the podcast Cloudbase Mayhem you should subscribe, great discussions, approaches to flying safely discussions, weather forecasting around the world, etc. DO not get turned off by the fact that it is PG focused, it covers many aspects of of soaring and hits topics any field of aviation needs to cover and understand.

Enjoy

Ron GLeason

[[email protected]]

On Sunday, 7 June 2020 05:17:57 UTC+2, wrote:
> They both have the HRRR model, but the only other option on Skysight is it's own model which I know nothing about. XCSkies have several models to chose from, of varying vertical/horizontal resolutions.
>
> It also seems to me that Skysight updates the model less frequently the day prior and day of soaring compared to the RAP/HRRR models on XCSkies. That would negate a lot of the benefit of the HRRR model, wouldn't it?
>
> What makes Skysight worth the cost, being twice as expensive as XCSkies?

You might enjoy my lecture from a few weeks ago, https://www.youtube.com/watch?v=Ez9Xjo4llT8

To directly answer some of your questions though:
1) If you are only going to use government weather models (GFS/RAP/HRRR etc), HRRR is the one you want - it's head and shoulders ahead of all the others for soaring purposes. It has a very high level of detail, and very precise physics, and updates hourly. The only problem is the forecast range, just ~36 hours. The other models are a significant regression for soaring purposes compared to HRRR (or SkySight).

HRRR validates very well against conventional validation parameters (better than SkySight mostly due to radar assimilation and cycling), but they do little to no validation of boundary layer 'soaring' parameters.
Because we run our own model, we can and do validate then adjust the model accordingly to get the best results for soaring pilots. Running weather models comparable to national weather services is extraordinarily expensive, we have ~1000 cores forecasting 24x7, and this factors into our pricing as does the ongoing development we do and maintaining the integrations with LX9000's, Oudies etc.

2) We have updates scheduled for the SkySight model at ~2am/8am/2pm for today and ~5pm for tomorrow, and beyond that update every forecast 1-2x a day. Running any model yourself is a balance of forecast duration, cost and detail, and by scheduling for these times we think we can offer maximum benefit at useful times for minimal cost. HRRR updates on SkySight hourly for today, and every 6 hours for tomorrow. I don't think hourly updates for longer than 24 hours ahead add meaningful information.

You can contact me directly via the website if you have any further questions.

[[email protected]]

Any opinions on Top Meteo's forecast models?

https://europe.topmeteo.eu/en-gb/

[Bret Hess]

It seems that NAM forecasts (12 km and 3 km) on XCSkies are very helpful for soaring in Utah. I haven't seen a benefit of HRRR over NAM. I'd be interested in hearing from other pilots who fly in the mountains comparing NAM with HRRR or Skysight for thermal predictions.

I'd also like to hear a discussion of why a 3 km resolution model is a great improvement for soaring over a 12 km model, for example. Higher resolution models generate more detailed forecasts, and maybe they show better what kind of variation you might see, but the variations at this scale don't seem pinned to the earth at an accuracy of 3 km.

[[email protected]]

On Saturday, 13 June 2020 20:54:15 UTC+2, Bret Hess wrote:
> It seems that NAM forecasts (12 km and 3 km) on XCSkies are very helpful for soaring in Utah. I haven't seen a benefit of HRRR over NAM. I'd be interested in hearing from other pilots who fly in the mountains comparing NAM with HRRR or Skysight for thermal predictions.
>
> I'd also like to hear a discussion of why a 3 km resolution model is a great improvement for soaring over a 12 km model, for example. Higher resolution models generate more detailed forecasts, and maybe they show better what kind of variation you might see, but the variations at this scale don't seem pinned to the earth at an accuracy of 3 km.

You'll see detailed explanations in my talk above about resolution (https://www.youtube.com/watch?v=Ez9Xjo4llT8) including real world examples of the impact of resolution in the mountains. Over hilly or mountainous terrain the improvement from higher resolution can be significant. On flat terrain the influence is less, although if there's convergences or storms it may still be important.

Unless you're <4km or so you won't meaningfully resolve wave, convergences, or
convection. Instead you need to parameterize [ https://en.wikipedia.org/wiki/Parametrization_(atmospheric_modeling) ] the cumulus and convection, which has drawbacks.

I can't offer a comparison to NAM 3km or 12km, but I can show you the predicted tracks versus the route taken by Jim last week with their record flight (https://www.facebook.com/SkySight.io/photos/a.1388218147953276/2806515619456848/?type=3&theater)

[Tango Eight]

The model that is working best this season in the Northeastern USA is GFS. Winds, top of lift, cu predictions all right on the mark. Skysight not horrible or anything, just not as good as GFS. Dr Jack NAM totally broken: dew points much too high, surface forecast highs much too low. Sun on the ground, boundary layer top, winds work.

HRRR has never been useful for soaring here.

T8

On Saturday, June 13, 2020 at 3:22:25 PM UTC-4, wrote:
> On Saturday, 13 June 2020 20:54:15 UTC+2, Bret Hess wrote:
> > It seems that NAM forecasts (12 km and 3 km) on XCSkies are very helpful for soaring in Utah. I haven't seen a benefit of HRRR over NAM. I'd be interested in hearing from other pilots who fly in the mountains comparing NAM with HRRR or Skysight for thermal predictions.
> >
> > I'd also like to hear a discussion of why a 3 km resolution model is a great improvement for soaring over a 12 km model, for example. Higher resolution models generate more detailed forecasts, and maybe they show better what kind of variation you might see, but the variations at this scale don't seem pinned to the earth at an accuracy of 3 km.
>
> You'll see detailed explanations in my talk above about resolution (https://www.youtube.com/watch?v=Ez9Xjo4llT8) including real world examples of the impact of resolution in the mountains. Over hilly or mountainous terrain the improvement from higher resolution can be significant. On flat terrain the influence is less, although if there's convergences or storms it may still be important.
>
> Unless you're <4km or so you won't meaningfully resolve wave, convergences, or
> convection. Instead you need to parameterize [ https://en.wikipedia.org/wiki/Parametrization_(atmospheric_modeling) ] the cumulus and convection, which has drawbacks.
>
> I can't offer a comparison to NAM 3km or 12km, but I can show you the predicted tracks versus the route taken by Jim last week with their record flight (https://www.facebook.com/SkySight.io/photos/a.1388218147953276/2806515619456848/?type=3&theater)

[[email protected]]

Interesting T8, I have found the opposite to be true out west. HRRR, and Nam 12, see to be most accurate on the day of and 1 day before, GFS totally useless if not only for long term 3-4 days generalities.

CH ASW27

[Tango Eight]

On Sunday, June 14, 2020 at 1:30:47 PM UTC-4, wrote:
> Interesting T8, I have found the opposite to be true out west. HRRR, and Nam 12, see to be most accurate on the day of and 1 day before, GFS totally useless if not only for long term 3-4 days generalities.
>
> CH ASW27

Yesterday GFS and SkySight made similar predictions, GFS more optimistic. The win for the day goes to SkySight, especially for forecasting things winding down at 5pm, right on the money (GFS predicted 6pm). Blue day, sky revealed nothing except a significant inversion when you could get high enough to see it.

T8

[Bret Hess]

On Wednesday, June 17, 2020 at 10:44:53 AM UTC-6, Tango Eight wrote:

> GFS more optimistic.

Using XCSkies I also find a similar relation between GFS vs NAM: GFS is always too optimistic for thermals, and NAM is better.

[[email protected]]

It seems to me that GFS on xc skies always is high on its prediction of surface and altitude winds by about 5 to 10 mph high out west. Dont fly much back east and last year at Perry and Bermuda high I think I only looked at the Nam and Hrrr.

CH