SiriusXM Radar and Lightning Twice as Fast

[Txbyker]

I just got a flyer from Sirius about service improvements.

NEXRAD radar data now transmitting at 2.5 minute update rate versus previous 5 minutes.  2 Times faster than ADSB.

Precipitation data now at 2.5 minutes vs 5 minutes.  ADSB = 5 minutes.

Lightning data at 2.5 minutes versus none for ADSB.

I wonder if ADSB is hurting their subscriptions.  

Russ 

[PaulM]

I almost wondered if this was only the SirusXM signal.. and not the original XM signal that the older GDL69 systems receive..  (they are different, there is a new GDL69 SXM that receives all of the latest signals. )

The skeptic in me wonders.   I suppose we can look at the "age tag", and see if that regularly goes to 5 mins.. or resets between 2-3.. 

If the Nexrad source doesn't update at 2.5 mins, then this isn't really useful is it? 

I don't see anything here that says better than 5 mins:
https://en.wikipedia.org/wiki/NEXRAD

 

[EricJ]

If the source updates faster I think there's nothing stopping ADS-B from ultimately updating faster, either, or adding lightning.    They're flexible systems, which is part of the whole point of going that way.

Given the relative out-of-pocket cost comparison, I'm sure ADS-B is hurting Sirius/XM.   Plus ADS-B is pretty portable, I can get it on my tablet with my $200 Stratux no matter what I'm flying.

 

[Piloto]

The total scan is 5 minutes but the wx data update rate could be every 2.5 minutes. This means that the update does not includes the whole scan but recently refreshed sectors. Very much like what you see on actual wx radar display as the sweep goes along.

Myself I just have WX on the Samsung S7 cell phone using the Garmin Pilot app and the NOAA Aviation Weather website. Before leaving to the hangar I can plan a weather avoidance route that I can alter in-flight. Unlike ADSB or SiriusXM coverage is worldwide over land anywhere there is internet access.

José

Edited February 27, 2018 by Piloto

[TWinter]

I still keep my XM active on my 796 portable.. I've tried the SXAR1 and did not think the pixel was as clear as my 796 w/ the old GXM40 hockey puck...My GDL88 w/ ADSB on the 750 looks kind of blocky in my opinion as well. I've not tried the new XM add-on Sporty's is advertising. Was worried it would be like the SXAR1.

Still keep the XM for a while.

-Tom

[Hector]

I didn't cancel, but I downgraded my subscription to the $9.99 boaters subscription when I got ADSB-in.  I like having backups and the cheap boaters subscription still gets me all the METARS and low res radar and a few other things.  I think its worth the 9 bucks but would get the basic aviation version if they offered a cheaper package.

[LANCECASPER]

12 hours ago, Hector said:

I didn't cancel, but I downgraded my subscription to the $9.99 boaters subscription when I got ADSB-in.  I like having backups and the cheap boaters subscription still gets me all the METARS and low res radar and a few other things.  I think its worth the 9 bucks but would get the basic aviation version if they offered a cheaper package.

I do the same thing on a Garmin 496. Have you been able to get this subscription to work on any other Garmin device?

[Marauder]

I just got a flyer from Sirius about service improvements.
NEXRAD radar data now transmitting at 2.5 minute update rate versus previous 5 minutes.  2 Times faster than ADSB.
Precipitation data now at 2.5 minutes vs 5 minutes.  ADSB = 5 minutes.
Lightning data at 2.5 minutes versus none for ADSB.
I wonder if ADSB is hurting their subscriptions.  
Russ 

Russ - I think José may be onto what is happening. The actual radar & lightning image is a composite of a series of radar and spheric sweeps. That composite takes time to compile prior to uploading. So if the update rate is done sooner, it may be in fact a recycling of some of the previous composite imagery. Not sure if that is a good thing.


Sent from my iPad using Tapatalk Pro

[Piloto]

In the near future NEXRAD will be transitioning to phase array radar antennas. It replaces the mechanical rotating antennas making the sky scanning much quicker and reliable. Many onboard radars also use phase array antennas.

https://www.nssl.noaa.gov/tools/radar/mpar/

José

 

[PaulM]

18 minutes ago, Marauder said:

Russ - I think José may be onto what is happening. The actual radar & lightning image is a composite of a series of radar and spheric sweeps. That composite takes time to compile prior to uploading. So if the update rate is done sooner, it may be in fact a recycling of some of the previous composite imagery. Not sure if that is a good thing.

I'm not sure.. Does NWS send the partial data to 3rd parties?... We know how NEXRAD works, and they can have partial data, but do they let it out of their systems?.. when I checked a number of NEXRAD providers (windy.com)  Their composite updates at 5 mins.  Intellicast says 5mins (for individual sites in storm mode, not the composite) 

So, we should experiment.   Take a film of the NEXRAD from our GDL69 and see at what rate the mosaic actually changes.. if it is 5 mins, then the 2.5 is a "marketing" rate.. sending the same data every 2.5 mins isn't useful.  

they can "project" the data in the interval, but don't you think that should be disclosed to us?  Data may be projected/theoretical?   It would be better to know the real age of the data.. not transmitted 5 mins ago... but 1st radar sweep 10 mins ago.. last sweep, 5 mins. 

[Txbyker]

2 hours ago, Marauder said:

 

Russ - I think José may be onto what is happening. The actual radar & lightning image is a composite of a series of radar and spheric sweeps. That composite takes time to compile prior to uploading. So if the update rate is done sooner, it may be in fact a recycling of some of the previous composite imagery. Not sure if that is a good thing.


Sent from my iPad using Tapatalk Pro

 

A sneaky way of marketing perhaps.

[Txbyker]

I was intrigued by your responses and did some more research.  I read the AIM 4.5.9 that shows a table that FIS B broadcast of the NEXRAD image is now 2.5 minutes.  That made me think is Sirius just keeping up with the FIS B and wondering if they are actually getting the NEXRAD source faster than 5 minutes.  Are they just sending the same NEXRAD image every 2.5 minutes.  After waiting 20 minutes on hold with Sirius I got tech support that said the 2.5 minute transmitting was news to him that marketing had not informed them of that.  He gave me an email in support to ask about how this works.  I will post when they inform me.

Russ

[PaulM]

I just checked NOAA's source sites:

radar.weather.gov.

For base reflectivity: The sites that have no current storm activity updated on 10min cycles.  A site that had active storms within range had a 5-6min update cycle. 

The wide area mosaics were 10min cycles. 

WSI Intellicast only has15min animations on their website.   I can't tell if they have better products via their Aviation services. 

AOPA's mosaic data is 5mins

LockMart was 10mins for mosaics and 5 mins on individual base. 

A true 2.5min cycle with new data is unlikely.   2.5mins just gets you get data quicker after startup.   Still have to confirm that via experiment. 

[EricJ]

3 hours ago, PaulM said:

I'm not sure.. Does NWS send the partial data to 3rd parties?... We know how NEXRAD works, and they can have partial data, but do they let it out of their systems?.. when I checked a number of NEXRAD providers (windy.com)  Their composite updates at 5 mins.  Intellicast says 5mins (for individual sites in storm mode, not the composite) 

So, we should experiment.   Take a film of the NEXRAD from our GDL69 and see at what rate the mosaic actually changes.. if it is 5 mins, then the 2.5 is a "marketing" rate.. sending the same data every 2.5 mins isn't useful.  

they can "project" the data in the interval, but don't you think that should be disclosed to us?  Data may be projected/theoretical?   It would be better to know the real age of the data.. not transmitted 5 mins ago... but 1st radar sweep 10 mins ago.. last sweep, 5 mins. 

Unless they filter it themselves and then retransmit it.   This is done sometimes in other applications to allow marketing to claim a faster update rate:   take the same data at the low sample rate, upsample it and filter it (e.g., average the previous N samples, although that's a really crappy filter), and then retransmit at the new sample rate.

[Txbyker]

52 minutes ago, PaulM said:

I just checked NOAA's source sites:

radar.weather.gov.

For base reflectivity: The sites that have no current storm activity updated on 10min cycles.  A site that had active storms within range had a 5-6min update cycle. 

The wide area mosaics were 10min cycles. 

WSI Intellicast only has15min animations on their website.   I can't tell if they have better products via their Aviation services. 

AOPA's mosaic data is 5mins

LockMart was 10mins for mosaics and 5 mins on individual base. 

A true 2.5min cycle with new data is unlikely.   2.5mins just gets you get data quicker after startup.   Still have to confirm that via experiment. 

This is what Sirius Aviation support told me "It’s faster radar. The weather provider was able to change the way they process the information thereby improving the update rate."

I will try to test it.

Russ

[Marauder]

This is what Sirius Aviation support told me "It’s faster radar. The weather provider was able to change the way they process the information thereby improving the update rate."
I will try to test it.
Russ

I’m pretty sure a Doppler radar site samples at a fixed rate. Sounds like Sirius is making claims that I don’t believe can be true. The only way I could see this working is they used a live feed from one radar site and retransmitted it. The radar we see is a composite radar made of inputs from overlapping coverage.


Sent from my iPad using Tapatalk Pro

[Marauder]

Here is an example. The first photo is a single site radar sweep from central PA near State College. The composite is the second picture. The composite captures an area of multiple radar site returns, well beyond the range of the first site. Unless they found a way to just feed you the nearest radar, and that radar covers your intended direction of flight, I am pretty sure it won’t be a complete picture.






Sent from my iPad using Tapatalk Pro

[0TreeLemur]

AFAIK, the WSR-88D scan rate is still about 14-deg/s, requiring 27sec to do a 360-degree scan.  During weather, they do 14 of these scans.  The total time to do a volume scan is approx. 6 minutes.   Note that 14-deg/sec is really fast.   Increasing it is infeasible for lots of reasons.   Sirius is either doing some trickery or repeating the same image.   I don't work for NOAA but I work with a bunch of people that do.

[Piloto]

I think they are just transmitting the same data scan but at a faster update rate. This may make a difference on the device you are displaying WX, specially if its doing multitasking like showing a map also. On a very fast processing device/display may not make a difference.

José

[bradp]

Anyone use the storm app where you can select the reflectivity angle ?  It’s pretty cool and doesn’t seem to involve the composite time.  I’ll have to dig back but one day I was sitting in schenley Park in Pittsburgh as a line of storms was moving through.  I compared Mark 1 eyeballs to storm to foreflight and they updated in that order ...

Of course the future of radar for all the aircraft that don’t have onboard radar is whenever we get reliable internet in flight and can get hold of a live / raw radar stream.  

[Marauder]

8 hours ago, bradp said:

Anyone use the storm app where you can select the reflectivity angle ?  It’s pretty cool and doesn’t seem to involve the composite time.  I’ll have to dig back but one day I was sitting in schenley Park in Pittsburgh as a line of storms was moving through.  I compared Mark 1 eyeballs to storm to foreflight and they updated in that order ...

Of course the future of radar for all the aircraft that don’t have onboard radar is whenever we get reliable internet in flight and can get hold of a live / raw radar stream.  

I use it all the time. Great for a look see of current weather. Using both the reflectivity and velocity functions, you can get a sense of storm intensity and direction. Unfortunately, the angle doesn't translate into an actual defined altitude for cloud top or top of precipitation. Add in a ground StormScope view, no one shouldn't be surprised by the weather.

One recent change that I am not happy about is that they eliminate the ad free version. I paid for the original ad free version only to be told that ads were coming back.

[Scott Dennstaedt, PhD]

On 2/28/2018 at 9:56 AM, PaulM said:

I'm not sure.. Does NWS send the partial data to 3rd parties?... We know how NEXRAD works, and they can have partial data, but do they let it out of their systems?.. when I checked a number of NEXRAD providers (windy.com)  Their composite updates at 5 mins.  Intellicast says 5mins (for individual sites in storm mode, not the composite) 

So, we should experiment.   Take a film of the NEXRAD from our GDL69 and see at what rate the mosaic actually changes.. if it is 5 mins, then the 2.5 is a "marketing" rate.. sending the same data every 2.5 mins isn't useful.  

they can "project" the data in the interval, but don't you think that should be disclosed to us?  Data may be projected/theoretical?   It would be better to know the real age of the data.. not transmitted 5 mins ago... but 1st radar sweep 10 mins ago.. last sweep, 5 mins. 

Paul,

I know this is a somewhat older thread, but I did a presentation at AirVenture this past week that discussed this change to the SiriusXM (SXM) radar.  My response is not directed at you in particular, but someone who attended my presentation asked me to comment here...  

This "twice as fast" catch phrase as SXM is using affects both lightning and the composite reflectivity mosaic.  Just as a refresher, composite reflectivity examines the base reflectivity (dBZ, where Z is the reflectivity parameter) from every elevation scan in the volume coverage pattern (VCP).  It then extracts the highest dBZ in each column over the radar coverage region.  That could have been from the lowest base reflectivity elevation angle (lowest tilt) or the base reflectivity from one of the higher elevation angles. As I discussed in my presentation at AirVenture, the term "base" does not mean lowest as most pilots assume...every elevation angle has a base reflectivity product.  To create a mosaic, the composite reflectivity from each radar site must be stitched together (there can be some nasty assumptions here with this process).  

As you might imagine, the WDR-88D NEXRAD Doppler radars are asynchronous.  That is, one radar site may be scanning the atmosphere on the lowest elevation angle while a neighboring radar site may be scanning the atmosphere on the 4th elevation angle.  There are many different scanning strategies depending on the type of weather expected in the area (severe, drizzle, snow, etc).  Nevertheless, TWC (SXM's provider) doesn't care about the asynchronous aspect of the radars.  They simply grab the last known complete base reflectivity scan from whatever elevation angle and count backwards in time from there until they've grabbed an entire "volume scan's" worth of data.  For example, if the radar just completed the 4th elevation angle, they grab the base reflectivity from that scan and go back to elevation 3, 2, 1, 14, 13, ..., 7, 6, 5 to grab the base reflectivity from those scans.  Perhaps a neighboring radar was finished the 7th elevation angle, they would grab 7, 6, 5, 4, 3, 2, 1, 14, 13, ..., 9, 8 and so on.  In this case I'm assuming that every radar has 14 elevation angles per volume coverage pattern which is not always the case.   

So to get the twice as fast updates, they simply schedule this process above every 2.5 minutes.  They could do it once a minute.  But, both FIS-B and SXM are highly bandwidth challenged...you can only put so much data in the small pipe...and it has to be scheduled accordingly.  Of course, any particular pixel you see on your display could have been from the oldest elevation scan or the newest.  And yes there are delays in processing and uplink/downlink that occur.

For lightning.  At a scheduled time, TWC pulls all data collected in past 2 minutes and puts it into a file, which is then queued for SXM broadcast. Add 10 seconds of processing and queuing. All said, the lightning data a pilot would see could be anywhere from 25 sec to 2 min 55 sec old at reception in the cockpit. The age (I call this the virtual age) you see on some displays is based on the time of reception and not the age of the product.  The natural age of the product varies as stated above.  Of course, there could be other delays.  For example, the software vendor could choose to hold that data for a period of time before displaying it (why that happens is a whole different discussion).  Once received, you stare at that data for another 2.5 minutes until a new set of data is received.  Therefore, moments before the next update is received, the lightning could be 2 min 55 sec to 5 min 25 sec old (although the "age" in the display may only say 2 or 3 minutes).

Same is true of the radar mosaic.  The faster refresh rate cuts down that stare time from 5 minutes to 2.5 minutes.  The virtual age is based on the time of reception...once again, this could have been from the oldest elevation scan or the newest (you don't know).  In a perfect world, the average natural age of the radar mosaic is 3.5 to 5 minutes old and then you stare at that image for 2.5 minutes now instead of 5.   Also important to understand is that this radar mosaic is highly filtered.  It's designed to show only those returns that come from actual hydrometeors.  They attempt to filter out ground clutter and anomalous propagation, but sometimes it does get through.  And in the worst case scenario, they can add a manual gross filter in a region where precipitation is highly unlikely where it filters out ALL returns...so if they fail to remove it in a timely manner, I've have seen them filter out real precipitation returns...even severe storms in some cases.  It happens more often than they want to admit.  That's why I always have lightning (and storm tracks) turned on...it's not part of that gross filter.  

By the way, the NWS is experimenting with lowering the lowest elevation angles for some radar sites.  The lowest elevation angle is 0.5 degrees.  This test will include dropping the lowest elevation to -0.2, 0 and +0.2 degrees.  Since December 2017, the radar in San Francisco has already been operating at +0.2 degrees (MUX).  

Sorry for the length of this response, but I hope this helps clear a few things up...

  

Edited August 1, 2018 by scottd

[carusoam]

Welcome aboard, ScottD!

Thanks for stopping by.  Your experience is greatly appreciated.

Thanks for sharing with the Aviation community and the MS community specifically.

Best regards,

-a-

[PaulM]

Thank you Scott for the detail, 

   The information we needed to know was that TWC was getting data from the raw radar sweeps, and not the summary mosaic from NWS.   Given that the raw input data updates faster than every 5 mins, and that TWC was building its own mosaic we see that it is reasonable to have some data change, and therefore a different dataset to present on the 2.5m cycle. 

[Scott Dennstaedt, PhD]

On 8/2/2018 at 10:08 AM, PaulM said:

Thank you Scott for the detail, 

The information we needed to know was that TWC was getting data from the raw radar sweeps, and not the summary mosaic from NWS.   Given that the raw input data updates faster than every 5 mins, and that TWC was building its own mosaic we see that it is reasonable to have some data change, and therefore a different dataset to present on the 2.5m cycle. 

Yes, the TWC is doing a fair amount of processing of the raw data, so what you see with a new update is actually fresher than the prior one.  This is only for the composite reflectivity.  It would be less useful for the lowest elevation angle though since the fastest volume coverage pattern is every 4.5 minutes in VCP 12 or 212.  On average 1/2 the mosaic would update at 2.5 min and the other half would be the same as before - but it's be randomly distributed throughout the mosaic.  

Edited August 3, 2018 by scottd