GPS vs Altimeter Altitude

[Bob_Belville]

I note with interest how different the displayed altitudes are between that displayed on the altimeter with the nearest surface barometric pressure set and the GPS altitude which is "real" and independent of pressure or density. Almost 400' @ 9000' the other day.

 

Obviously, if everyone close to each other is flying with a similar barometric setting it is not important that we're all at 9350 instead of 9000.

 

I suppose the pressure altimeter is much more "accurate" when the plane is close to the terrain and that the discrepancy is variation from the standard lapse rate.?

[smccray]

It's my understanding that the GPS altitude is closer to true altitude than the barametric altitude.  If conditions are exactly standard (both temperature and pressure), then the altimiter will equal true altitude which will equal the GPS altitude.

[bumper]

GPS horizontal positioning will always be more accurate than altitude due to sat geometry. This from Garmin's website:

 

"The main source of error has to do with the arrangement of the satellite configurations during fix determinations. The earth blocks out satellites needed to get a good quality vertical measurement. Once the vertical datum is taken into account, the accuracy permitted by geometry considerations remains less than that of horizontal positions. It is not uncommon for satellite heights to be off from map elevations by +/- 400 ft. Use these values with caution when navigating."

 

This is why some GPS devices also include a pressure sensor for "massaging" or smoothing altitude data when that's important - - the GPS on my bicycle and the data logger in my glider are examples of this.

 

 

bumper

[Bob_Belville]

S McCray, while the altimeter might be "right" when pressure and temperature are standard, that's a moot point for an aircraft in flight since the barometric pressure is a number measured by instrumentation on the ground and adjusted to an equivalent sea level pressure. The pressure altitude the ground is only approximate since the standard lapse rate (the pressure drop with altitude) is just that, a standard, a convention. The actual lapse rate varies in real world conditions.

[Bob_Belville]

Bumper, I think, though I have not checked it a lot, that the GPS altitude is quite accurate when the plane in sitting on the airport. This fits with my sense that the lapse rate is the reason there's a bigger difference at a significant altitude AGL where the altimeter is dependent upon atmospheric pressure obeying standard lapse rate whereas the GPS couldn't care less what the pressure or temp is. When I bring the GPS 696 home and it's running on my desk it will claim accuracy with a couple of feet once it is seeing several satellites.

[Skywarrior]

If GPS altitude is so accurate, why do we have to use WAAS-enabled GPS units for GPS approaches?

[Bob_Belville]

WAAS is still just GPS. It does not input any pressure or temp info. http://www.faa.gov/about/office_org/headquarters_offices/ato/service_units/techops/navservices/gnss/waas/

[Skywarrior]

I never said that it *did* have anything to do with pressure or temp.

 

WAAS is more accurate than 'regular' GPS, both laterally and vertically.

 

Measured accuracy for both types, vertical only:

 

GPS - 4.7 m

 

WAAS - 1.3 m

 

 

The only reason we *ever* gave a sh*t about pressure and temperature was because *all we had* was pressure instruments.

 

The FAA still mandates that we gauge our altitude by those 'legacy' instruments.

 

So don't think of your GPS's altitude measurements as anything but 'advisory', unless you're using a WAAS unit.

[Bob_Belville]

Chuck. My GTN 750 agrees with my GPS 696. But you misunderstand the intent of my post. I am simply observing that with the advent of sophisticated GPS we can see how much the lapse rate varies from standard. And it would make a huge difference in mountainous terrain since that peak does not vary with a.press. I suppose that's part of why MEA in such terrain provides 2000' or more. Again it does not matter that when I was overflying Savannah GA (sea level) I was really at 9350 not 9000 as long as everyone else was equally offset.  

[bd32322]

It's my understanding that the GPS altitude is closer to true altitude than the barametric altitude.  If conditions are exactly standard (both temperature and pressure), then the altimiter will equal true altitude which will equal the GPS altitude.

 

 

No that is not the case - true altitude is defined as the actual altitude above terrain regardless of temperature, pressure what not ...

How does the GPS satellite constellation know where the mountains are to give you a true altitude? It doesn't. Some boxes might pretend to give obstacle clearance but they are based on data that is not as accurate as WAAS accuracy data being mentioned in this thread.

 

What is used for this fake GPS altitude is a simplistic model of the earth - which is some sphere. GPS - knowing your point in space - calculates your height above this imaginary sphere. This altitude is very different from barometric altitude and is not accurate at all when it comes to telling how high above obstacles or ground you are - because the world is not a sphere first of all and its definitely not smooth

 

 

Why you ask? How are there LPV approaches going down to 200 feet above airport elevation?

 

That's because GPS knows your point in space very accurately - like someone mentioned the accuracy numbers are less than a meter 90% of the time, more than a meter or so some other percentage of the time.

 

But knowing your position in space very accurately - doesn't mean you know your height above the ground. For that you will need to know the GPS co-ordinates of the terrain. Since no one has gone and mapped the entire terrain with GPS coordinates - you accuracy for height above the ground is not sub-meter accuracy !

 

But what about LPV approaches? Those terrain points HAVE been measured by GPS devices and their 3-D coordinates in the GPS satellite constellation space are known exactly. So when you shoot a LPV approach - the GPS box technically doesn't have your barometric altitude and hence your height above terrain. It does however know where in space you are, where in space the LPV glideslope is and where in space the runway end is.. That's how you get an accurate approach path without knowing your height above the ground.

 

So relying on GPS altitude shown on iPads, 430s what not - is inviting disaster unless you have no other option. There was a discussion on another thread recently

[bd32322]

Having said the above - if you go to point X and your WAAS GPS shows gps altitude 1000 and then later you are flying over the point and your WAAS GPS altitude is 3000 - then you are exactly 2000 feet over point X.

 

This is accurate because you mapped the terrain at point X and just took the difference at flying height of 3000 - which is accurate to within WAAS GPS accuracy

[FloridaMan]

http://www.youtube.com/watch?v=qGWx6rkkWdg&list=FL4yGM3xSuFZjMMLe-iu-zSw

[AndyFromCB]

No that is not the case - true altitude is defined as the actual altitude above terrain regardless of temperature, pressure what not ...

How does the GPS satellite constellation know where the mountains are to give you a true altitude? It doesn't. Some boxes might pretend to give obstacle clearance but they are based on data that is not as accurate as WAAS accuracy data being mentioned in this thread.

What is used for this fake GPS altitude is a simplistic model of the earth - which is some sphere. GPS - knowing your point in space - calculates your height above this imaginary sphere. This altitude is very different from barometric altitude and is not accurate at all when it comes to telling how high above obstacles or ground you are - because the world is not a sphere first of all and its definitely not smooth

Why you ask? How are there LPV approaches going down to 200 feet above airport elevation?

That's because GPS knows your point in space very accurately - like someone mentioned the accuracy numbers are less than a meter 90% of the time, more than a meter or so some other percentage of the time.

But knowing your position in space very accurately - doesn't mean you know your height above the ground. For that you will need to know the GPS co-ordinates of the terrain. Since no one has gone and mapped the entire terrain with GPS coordinates - you accuracy for height above the ground is not sub-meter accuracy !

But what about LPV approaches? Those terrain points HAVE been measured by GPS devices and their 3-D coordinates in the GPS satellite constellation space are known exactly. So when you shoot a LPV approach - the GPS box technically doesn't have your barometric altitude and hence your height above terrain. It does however know where in space you are, where in space the LPV glideslope is and where in space the runway end is.. That's how you get an accurate approach path without knowing your height above the ground.

So relying on GPS altitude shown on iPads, 430s what not - is inviting disaster unless you have no other option. There was a discussion on another thread recently

As I am here to always make friends let me try again. Everything you said above is wrong.

[bd32322]

http://www.youtube.com/watch?v=qGWx6rkkWdg&list=FL4yGM3xSuFZjMMLe-iu-zSw

Antares, not sure whether you posted that for general education or to refute my points, but the video agrees with what I am saying. The GPS system itself cannot tell you height above ground. The video mentions that all runway ends were surveyed according to some geodetic model, which is precisely what I was saying regarding WAAS approaches - the runway height has to be brought into the GPS co-ordinate system. But the rest of the terrain is not in the GPS geodetic system because all terrain has not been surveyed by measuring gps altitude at all points on the earths surface.

The video also mentions that you need barometric aiding in order to get real altitude over sea level.

Sorry if you were not refuting my point but this could be a joint response for astelmaszek - maybe he can clarify why I am wrong

[bd32322]

As I am here to always make friends let me try again. Everything you said above is wrong.

Would help if you explain why? Maybe I wasnt explaining correctly?

The only way GPS altitude will be close to true altitude is if the major terrain points have been surveyed for GPS co-ordinates and that new model has been loaded into receivers. I dont know whether that has been done.

[N201MKTurbo]

Atmospheric pressure has nothing to do with local terrain. An altimeter doesn’t know what is under you, it just knows how much air is above you.

 

I tried to find the relation between pressure and altitude one time, and found that it is very complicated and nonlinear. An altimeter is a wondrous mechanical system that converts a nonlinear effect to a linear reading that very closely correlates to your MSL altitude.

 

GPS gives you the distance from the center of the earth corrected for the earth surface model. The earth is not a sphere, it bulges at the equator because of its spin.

 

 

It is all very complicated. And I wouldn’t pretend to be an expert.

[bd32322]

Aah crap, for some reason I said true altitude is height above terrain which it is not ... Doh

I was mostly concerned with terrain clearance by looking at msl values of obstructions and terrain using charts - with a barometric altimeter you can get a very accurate msl figure which you can compare on a chart. Yes, neither altitude knows how much air is below you .. GPS is how far you are from the satellite constellation and barometric is how much air is on top of you like m201mkturbo said.

For True altitude - height above mean sea level - gps altitude is still not correct for the reasons m201mktturbo mentions.

[Bob_Belville]

Wow, there has been a lot of extraneous info posted to this thread!

 

Despite a lot of references to AGL, neither GPS nor a standard pressure altimeter will directly provide any info about AGL. (Unless the plane is on the ground.)

 

Both GPS and an altimeter provide altitude relative to MSL.

 

GPS uses math to triangulate the elevation of the radio relative to MSL.

 

OTOH, an altimeter knows only the weight of air above the instrument. This value is used to calculate an approximation of the elevation above MSL. The nearest station reporting the current barometric pressure helps to improve the approximation but the necessity to assume a standard lapse rate introduces significant error.  

[M016576]

The US has mapped and made available a "terrain mesh" of the world called DTED (Digital Terrain Elevation Data) using our satellite constellation. It's available to the general public as a relatively small computer file. That's how all these "synthetic vision" and TAWS displays know (estimate) where the actual ground is located. The better the DTED level, the better your terrain mesh, the better your ability to determine your actual AGL altitude using just a GPS coordinate. I think the best current DTED level is something like DTED 5, which is something like one point measured every meter.

All these new boxes and GPS's have DTED in them. Really, though, all that matters is that we have a system that is *relative* to all the aircraft and fields that are out there. That system is pressure based right now, but could easily be mandated to be GPS/DTED derived in the future. Or maybe it could be something new, something we don't even know about yet

[AndyFromCB]

WASS is not relative to anything. In all my years working with precision agriculture solutions, I have found WASS data to be within about 3 feet of true ECL 97% of the time, within 5 feet of true ECL the other 3% of the time. When I say mean sea level, I actually say mean earth center level. For all intents and purposes here, all that GPS calculates is height above center of earth (ECL). It then takes some minute amount of processing power to figure out where on earth you're actually are. It has nothing to do with terrain surveying and all to to do with basic shape of the earth. So actually relaying on GPS altitude to estimate ones hight above the terrain is considerably more accurate by a factor of at least .0025 than relaying on your altimeter to determine how high you are from slamming into Gannet peak. WASS as it stands today, blows alway CAT I approaches requirements with a chance of it being wrong at any giver time of being about 1 second per year.

 

We drive tractors/implements here in Iowa without a soul involved for 95 days a year in our test projects. At any given time I can tell you where the tractor is within 3 feet XYZ on the field. Never have I seen it be more than 5 feet off based on laser ring gyros and within an hours time, they are accurate to fraction of a an inch.

 

What I am trying to say is when you look at your altimeter, you could be between 500 feet below or above the actual MSL vs indicated when cruising at about 14,000 MSL. Your GPS WASS altitude will most likely be within 3 feet of actual never more than 5 feet and no, RAIM doesn't matter with WASS no matter what your friendly FAA fellow told you. Don't forget what drives EGPWS in turbine aircraft. It ain't the altimeter.

[Bob_Belville]

Astel Maszek wrote:

 

...What I am trying to say is when you look at your altimeter, you could be between 500 feet below or above the actual MSL vs indicated when cruising at about 14,000 MSL. Your GPS WASS altitude will most likely be within 3 feet of actual never more than 5 feet and no, RAIM doesn't matter with WASS no matter what your friendly FAA fellow told you. Don't forget what drives EGPWS in turbine aircraft. It ain't the altimeter.

 

That's what I was looking to say/hear. Thanks.

 

[bd32322]

WASS is not relative to anything. In all my years working with precision agriculture solutions, I have found WASS data to be within about 3 feet of true ECL 97% of the time, within 5 feet of true ECL the other 3% of the time. When I say mean sea level, I actually say mean earth center level. For all intents and purposes here, all that GPS calculates is height above center of earth (ECL). It then takes some minute amount of processing power to figure out where on earth you're actually are. It has nothing to do with terrain surveying and all to to do with basic shape of the earth. So actually relaying on GPS altitude to estimate ones hight above the terrain is considerably more accurate by a factor of at least .0025 than relaying on your altimeter to determine how high you are from slamming into Gannet peak. WASS as it stands today, blows alway CAT I approaches requirements with a chance of it being wrong at any giver time of being about 1 second per year.

 

We drive tractors/implements here in Iowa without a soul involved for 95 days a year in our test projects. At any given time I can tell you where the tractor is within 3 feet XYZ on the field. Never have I seen it be more than 5 feet off based on laser ring gyros and within an hours time, they are accurate to fraction of a an inch.

 

What I am trying to say is when you look at your altimeter, you could be between 500 feet below or above the actual MSL vs indicated when cruising at about 14,000 MSL. Your GPS WASS altitude will most likely be within 3 feet of actual never more than 5 feet and no, RAIM doesn't matter with WASS no matter what your friendly FAA fellow told you. Don't forget what drives EGPWS in turbine aircraft. It ain't the altimeter.

 

 

Yes that's what everyone is saying - its height measured from the center of the earth. But to arrive at an altitude value it does this:

GPS alt = (height above center of earth) - (height of model's sea level)

The model's sea level is not accurate because the earth has a complex shape. This is just another altitude - if everything were charted using GPS altitude we would be fine - however, stuff in charts etc is in MSL based on barometric altitude - comparing the two is when there is a problem.

 

Your example of tractors being accurate in X Y dimensions is understandable - since the horizontal direction is never an issue. Its the vertical terrain that is an issue.

 

You mention EGPWS - but EGPWS uses a radar altimeter to get the height above terrain directly below the aircraft. The GPS knows exactly where you are with respect to the center of the earth. It knows how far above the terrain the aircraft is from the radar altimeter and it knows GPS altitude (not MSL) with a high degree of precision.

 

From the terrain database it gets:

(height of obstruction ahead) - (height at current position) = (height of obstruction compared to current position)

Say a mountain ahead is 14,000 feet MSL, height at the current position listed in terrain database is 10,000 feet MSL. Then the terrain rises by 4000 feet.

 

 

Then the EGPWS gets your radar altimeter reading - say that is 1000 feet

So you are 1000 feet above the ground when directly over the current position. At this altitude your GPS altitude is reading say 11,100 feet above center of earth.

 

So amount you need to climb up by is 4000 feet terrain rise - 1000 feet = 3000 feet

 

GPS altitude you need to reach to clear mountain is 3000 feet + 11,100 = 14,100 feet

 

If you used your GPS altitude as your primary means of altitude clearance without the radar altimeter - you would undershoot by 100 feet and hit the peak at 14,000 feet.

 

 

If you used your GPS altitude to clear the peak

 

 

As I was saying, the GPS altitude has a very low error - so if you measure altitude at x and climb by 1000 feet - gps altitude will show x+1000 but it has no relation to terrain - its only a coordinate in the satellite constellation space

 

You need the radar altimeter to tie it down to earth or you need barometric pressure altimeter to tie it down.

 

I am curious if a GPS unit showing obstacle clearance for passenger carriers, was ever approved without a barometer tie-in.

[bd32322]

Then again - this is not a problem if the terrain database has GPS coordinates for X,Y,Z values for all terrain points. But from Antares's youtube video it seems they only surveyed runway ends and nav points ..

 

Once all obstacles etc are in GPS co-ords you can safely use just a GPS altitude to clear those obstacles

[N201MKTurbo]

Anybody trying to clear an obstruction by 100 feet at night or in IMC has a death wish no matter what altitude reference they are using.

If I remember my FARs if IMC you need to clear everything by 2000 feet for five miles either side of your course. That negates the difference between GPS and indicated altitude.

[carusoam]

Bob,

"Despite a lot of references to AGL, neither GPS nor a standard pressure altimeter will directly provide any info about AGL. (Unless the plane is on the ground.)"

WingX provides altitude AGL on the Ipad1 as a new feature! Turns red when you get too close. It also colors terrain out front in red at and below your own alt...

The government has been measuring and mapping surface/ AGL with great accuracy from sattelites for years.

Just sharing Bob's point.

Best regards,

-a-