How accurate are the OLC AGL traces?

[Nick Kennedy[_3_]]

Like most pilots who log flights on the OLC I watch the daily flights and I like to see whats going on out in the wild blue today.

Question: How accurate are the AGL statistics at the bottom of the traces?

Today for instance, There was a pilot of of Ely who got within 59 meters of the terrain and one out of Minden who got within 6 [! ] meters of the terrain, so says the OLC trace.
Are these numbers accurate? I see these very close to the mountain AGL numbers ALL the time.

Seeing the recent fatality in Ely due to hitting the mountain I think we might need to readjust our safety minimum's.

I don't know about the East but flying out West in the summer, I think we all need to be more careful in regards to flying close to the ground, being it a mountain, whatever.
What do you guys think?
Trying to fly safe in 2020
Nick
T

[MNLou]

Hi Nick -

When I want to analyze a flight from OLC, I download the IGC file and open it in See You.

I think the AGL numbers are See You are more accurate than from the OLC summary.

YMMV -

Lou

[kinsell]

On 7/15/20 7:34 PM, MNLou wrote:
> Hi Nick -
>
> When I want to analyze a flight from OLC, I download the IGC file and open it in See You.
>
> I think the AGL numbers are See You are more accurate than from the OLC summary.
>
> YMMV -
>
> Lou
>

The barometric sensor in the logger is going to be assuming 29.92 inches
of mercury for atmospheric pressure, any deviation from that is going to
introduce an error on top of the instrument accuracy. So the AGL
calculations can be off quite a bit no matter what database is used for
ground elevation..

[[email protected]]

On Thursday, July 16, 2020 at 6:28:25 AM UTC+2, kinsell wrote:
> On 7/15/20 7:34 PM, MNLou wrote:
> > Hi Nick -
> >
> > When I want to analyze a flight from OLC, I download the IGC file and open it in See You.
> >
> > I think the AGL numbers are See You are more accurate than from the OLC summary.
> >
> > YMMV -
> >
> > Lou
> >
>
> The barometric sensor in the logger is going to be assuming 29.92 inches
> of mercury for atmospheric pressure, any deviation from that is going to
> introduce an error on top of the instrument accuracy. So the AGL
> calculations can be off quite a bit no matter what database is used for
> ground elevation..

I would suspect OLC is using GPS altitude for AGL calculations, and probably using a fairly course terrain database which would explain the inaccuracies.

[[email protected]]

On Thursday, July 16, 2020 at 2:19:19 AM UTC-4, wrote:
> On Thursday, July 16, 2020 at 6:28:25 AM UTC+2, kinsell wrote:
> > On 7/15/20 7:34 PM, MNLou wrote:
> > > Hi Nick -
> > >
> > > When I want to analyze a flight from OLC, I download the IGC file and open it in See You.
> > >
> > > I think the AGL numbers are See You are more accurate than from the OLC summary.
> > >
> > > YMMV -
> > >
> > > Lou
> > >
> >
> > The barometric sensor in the logger is going to be assuming 29.92 inches
> > of mercury for atmospheric pressure, any deviation from that is going to
> > introduce an error on top of the instrument accuracy. So the AGL
> > calculations can be off quite a bit no matter what database is used for
> > ground elevation..
>
> I would suspect OLC is using GPS altitude for AGL calculations, and probably using a fairly course terrain database which would explain the inaccuracies.

Moreover, I've seen many examples on OLC when it is mistaken on the takeoff time and place, but still assumes the altitude at the "takeoff" is that of the stated airport. This happens if, e.g., the logger does not have a GPS fix at the time of the real takeoff, or is reset later. So OLC "adjusts" the logged altitudes for its display. This results in grossly underestimated altitudes for the early part of the flight. (It apparently interpolates towards the landing altitude later in the flight.)

[kinsell]

On 7/16/20 12:19 AM, wrote:
> On Thursday, July 16, 2020 at 6:28:25 AM UTC+2, kinsell wrote:
>> On 7/15/20 7:34 PM, MNLou wrote:
>>> Hi Nick -
>>>
>>> When I want to analyze a flight from OLC, I download the IGC file and open it in See You.
>>>
>>> I think the AGL numbers are See You are more accurate than from the OLC summary.
>>>
>>> YMMV -
>>>
>>> Lou
>>>
>>
>> The barometric sensor in the logger is going to be assuming 29.92 inches
>> of mercury for atmospheric pressure, any deviation from that is going to
>> introduce an error on top of the instrument accuracy. So the AGL
>> calculations can be off quite a bit no matter what database is used for
>> ground elevation..
>
> I would suspect OLC is using GPS altitude for AGL calculations, and probably using a fairly course terrain database which would explain the inaccuracies.
>

The large majority of traces submitted to OLC are going to have pressure
data but yes they do accept ones with only GPS altitude. It's possible
they always use GPS altitude, but it's less accurate and tends to read
higher in my experience.

Whichever they use, resolution is not the same as accuracy. Subtracting
two large, inaccurate numbers from each other does not provide
meaningful results. And yes people sometimes get too close to the rocks.

[Tim Newport-Peace[_6_]]

At 14:25 16 July 2020, kinsell wrote:
>On 7/16/20 12:19 AM, wrote:
>> On Thursday, July 16, 2020 at 6:28:25 AM UTC+2, kinsell wrote:
>>> On 7/15/20 7:34 PM, MNLou wrote:
>>>> Hi Nick -
>>>>
>>>> When I want to analyze a flight from OLC, I download the IGC file and
>open it in See You.
>>>>
>>>> I think the AGL numbers are See You are more accurate than from the
OLC
>summary.
>>>>
>>>> YMMV -
>>>>
>>>> Lou
>>>>
>>>
>>> The barometric sensor in the logger is going to be assuming 29.92
inches
>>> of mercury for atmospheric pressure, any deviation from that is going
to
>>> introduce an error on top of the instrument accuracy. So the AGL
>>> calculations can be off quite a bit no matter what database is used
for
>>> ground elevation..
>>
>> I would suspect OLC is using GPS altitude for AGL calculations, and
>probably using a fairly course terrain database which would explain the
>inaccuracies.
>>
>
>The large majority of traces submitted to OLC are going to have pressure
>data but yes they do accept ones with only GPS altitude. It's possible
>they always use GPS altitude, but it's less accurate and tends to read
>higher in my experience.
>
>Whichever they use, resolution is not the same as accuracy. Subtracting
>two large, inaccurate numbers from each other does not provide
>meaningful results. And yes people sometimes get too close to the rocks.
>
GPS Altitudes from Flight Recorders _should_ be referenced to the WGS84
ellipsoid, but this is not always the case. There is usually a difference
between the WGS84 ellipsoid and Mean Sea Level.

What is the OLC Terrain Database referenced to?
>

[Steve Leonard[_2_]]

Over flat lands, pretty good. Over mountains, well, it is "interesting". I have seen friends traces showing negative AGL when he was polishing his way up the mountains. I wouldn't put too much faith in the AGL reported by OLC for mountain flights where the pilots are working close to the mountains.

My two cents worth,
Steve Leonard

[2G]

On Wednesday, July 15, 2020 at 6:22:32 PM UTC-7, Nick Kennedy wrote:
> Like most pilots who log flights on the OLC I watch the daily flights and I like to see whats going on out in the wild blue today.
>
> Question: How accurate are the AGL statistics at the bottom of the traces?
>
> Today for instance, There was a pilot of of Ely who got within 59 meters of the terrain and one out of Minden who got within 6 [! ] meters of the terrain, so says the OLC trace.
> Are these numbers accurate? I see these very close to the mountain AGL numbers ALL the time.
>
> Seeing the recent fatality in Ely due to hitting the mountain I think we might need to readjust our safety minimum's.
>
> I don't know about the East but flying out West in the summer, I think we all need to be more careful in regards to flying close to the ground, being it a mountain, whatever.
> What do you guys think?
> Trying to fly safe in 2020
> Nick
> T

I don't think there is any question that Marak was flying close to the terrain - too close. But exactly how close will never be known for sure (let's hope that they are able to recover his flight data log). The issue is do you want to be the next test pilot who tries to see how close is too close?

Tom