Mapquest inaccuracies outside U.S. or secret SA?

[+worldtraveler]

This cache was just approved (and very quickly, I might add), but there is a significant discrepancy (~1 km.) between where I placed the cache and the place shown by Mapquest on the cache page. I'm wondering whether anyone else has experienced discrepancies of a similar magnitude.

 

Here's the background:

1. I placed the cache using a Garmin GPS V in waypoint averaging mode. After 1100+ counts and showing an accuracy of 11.2 ft., I set the waypoint. Also, the WGS 84 datum was used.

 

2. The cache page and the Mapquest map show the same coordinates as the cache waypoint on the GPSr.

 

3. The location "star" on the Mapquest page is ~ 1 km west and slightly north of the place I actually hid the cache. The actual location is inside the green park area shown on Mapquest, ~100 meters from the lower left corner (this is just a very rough estimate.)

 

To try to isolate the cause of the discrepancy, I've used 2 other "known" reference points from different sources:

 

4. I set a waypoint at a hotel 1.3 mi. north of the cache location last year using a Garmin GPS III+. I transferred the waypoint to the GPS V, and it was still accurate when I checked it on the day I placed the cache. This shows a consistency between the 2 GPS units over time. They agree on where they "think" they're located.

 

5. I punched the hotel waypoint coords into Mapquest and once again the "star" on the map was west and slightly north of the actual hotel location; but this time the discrepancy was ~0.6 km. This shows some consistency in the discrepancy between Mapquest and my 2 GPS units in Bangkok.

 

6. I punched my U.S. home coordinates (essentially the same on both the III+ and V) into Mapquest, and it accurately showed my actual home location.

 

7. I've used the GPS V to find 49 caches in the U.S. and 1 near Hong Kong, so I don't think the error can be attributed to the GPSr. It reads essentially the same as at least 50 other GPS units.

 

That seems to narrow it to "operator error", some glitch with Mapquest in/around Bangkok, or some kind of induced error that would cause the GPSr to read incorrectly at/around this particular location. Since I used waypoint averaging and got the displayed accuracy down to 11.2 ft., I'm pretty sure I didn't do anything to insert any error. Mapquest certainly appears to be accurate within the U.S., but I haven't used it much elsewhere; hence the request for input.

 

And then there's the third possibility... and possible intrigue. The park where I placed the cache happens to be in an area where there is a large concentration of embassies, including that of the U.S. Do you suppose the discrepancies could be due to the imposition of selective availability? If so, this may account for the greater discrepancy (~1 km) near the cache (and embassies) and somewhat lesser discrepancy further away at the hotel.

 

I guess we'll have our answer if this topic disappears soon after posting.

 

Worldtraveler

[+Glenn]

My guess is that it may have somehting to do with how MapQuest converts the map datum on their Hong Kong map.

 

Different countries, even different provinces, use different datums. Sicily uses a different datum than the rest of Italy.

 

I've been reading a little about coord systems and map datum and even though the two are related they are two different things.

 

When geocaching you should set the coord system for WGS-84 and the map datum for the country/area that you are caching in. If you don't you will still find the cache but your GPS generated maps will be off.

 

Here is link to a page about the Hong Kong grid.

 

Here is another page that talks about being aware of the different map datums outside of America.

 

quote:

---

And then there's the third possibility... and possible intrigue. The park where I placed the cache happens to be in an area where there is a large concentration of embassies, including that of the U.S. Do you suppose the discrepancies could be due to the imposition of selective availability? If so, this may account for the greater discrepancy (~1 km) near the cache (and embassies) and somewhat lesser discrepancy further away at the hotel.

---

 

My limited understanding of the GPS system is that if SA was to be reinterduced it would have to be done system wide. You can not make a just a small location, say just an embasy, read different than only a few blocks away. A satellite sends the same signal to everyone that is listening. It can not select or deselect a certian building, city block, or even city. If you would want to "black out" an area, even a very small one you would have to "turn off" all the satellites that visable overhead from that area at that time. It would effect a very large area.

 

...

[+worldtraveler]

quote:

---

Originally posted by the Lone Rangers:

My guess is that it may have somehting to do with how MapQuest converts the map datum on their Hong Kong map.

---

 

I thought about the possibility of differences in the map datums, but could they yield that much difference? And if so, what use is a Mapquest map outside the U.S. (for geocaching.com purposes) if it's not using WGS 84?

 

quote:

---

When geocaching you should set the coord system for WGS-84 and the map datum for the country/area that you are caching in...

---

 

You lost me here. My humble GPS V allows changing the map datum, but I leave it on the geocaching "standard" WGS 84 when hunting/placing caches. I'm not sure what you mean by "set the coord system for WGS-84". My unit only allows changing the map datum, not the "coord system".

 

Worldtraveler

[+worldtraveler]

And now to further thicken the plot (if there is one), the map has been removed from the cache page! Jeremy, I sure hope it was you or one of your helpers.

 

Worldtraveler

[+Glenn]

quote:

---

Originally posted by worldtraveler:

I thought about the possibility of differences in the map datums, but could they yield that much difference? And if so, what use is a Mapquest map outside the U.S. (for geocaching.com purposes) if it's not using WGS 84?

---

 

It is an easy error to make. They get a Hong Kong map with HK map datum that they assume it is WGS-84. They dont have anyone in Hong Kong to verify that the maps are correct so they do know they made a mistake.

 

Here is a random location in Sicily.

First in WGS-84 map datum:

37 17.748N

014 38.233E

 

Now in SICIL map datum:

37 17.815N

014 38.275E

 

If the SICIL map datum was to be mistaken for WGS-84 map datum it would make a difference. How much I'm not sure. Maybe someone can calculate the difference for me. Then let me know how you did it.

 

quote:

---

 

You lost me here. My humble GPS V allows changing the map datum, but I leave it on the geocaching "standard" WGS 84 when hunting/placing caches. I'm not sure what you mean by "set the coord system for WGS-84". My unit only allows changing the map datum, not the "coord system".

 

Worldtraveler

---

 

Germin probably has a different name for it. Magellan calls it coord system. Its where you change what coordinate system your GPS reports it posistion to you in.

 

On a mapping GPS you should be able to select the coordinate system and the map datum seperatly.

 

The coordinate system is the system that is used when the GPS tells you where it (and hopefully you) is at.

 

The map datum setting is used to tell the GPS what coordinate system is being used on the map that you loaded in to it. This way the GPS can accuratly place you the map.

 

So you can load a map in to your GPS with, for example, SICIL map datum and be able to view your coordinates in WGS-84 and your location on your map will be correct. Otherwise you would have to veiw your coordinates as SICIL for your location to be properly shown on the map and when you switched back to WSG-84 your location on the map would jump because your GPS will be mapping your WSG-84 coordinate location on a SICIL coordinate map. Confused?

 

...

[+worldtraveler]

Sorry, I'm not familiar with Magellan units. Only Garmin's proprietary maps can be loaded into their GPSr's, but there's nothing in their documentation (that I'm aware of) that requires or suggests changing map datum depending on the maps that have been loaded.

 

Before starting this topic, I experimented a little with changing map datums in the GPSr to see how much effect it had on the cache coordinates. They changed a little, but when I punched the new coords into Mapquest, the location shift was very slight - nowhere near the ~1 km. discrepancy previously noted. That's why I questioned whether different map datums could account for that much discrepancy.

 

Worldtraveler

[+worldtraveler]

quote:

---

Originally posted by Lone Rangers:

Here is a random location in Sicily.

First in WGS-84 map datum:

37 17.748N

014 38.233E

 

Now in SICIL map datum:

37 17.815N

014 38.275E

 

If the SICIL map datum was to be mistaken for WGS-84 map datum it would make a difference. How much I'm not sure. Maybe someone can calculate the difference for me. Then let me know how you did it....

---

 

I converted them to the decimal degree format and punched them into 2 Mapquest windows. The shift appears to be only about 100 meters.

 

So back to my question: can different map datums account for a shift of 10 times that amount?

 

Worldtraveler

[+worldtraveler]

quote:

---

Originally posted by Lone Rangers:

My limited understanding of the GPS system is that if SA was to be reinterduced it would have to be done system wide... It would effect a very large area.

---

 

My understanding of the GPS system is pretty limited too; for that and other reasons, I'd be pretty hesitant to assert dogmatically what the U.S. government is or isn't capable of doing with their toys - especially those originally developed for military use.

 

I'm not a conspiracy nut; I'd just like to hear from someone who DOES know what map datum Mapquest uses for their Bangkok map. Then if I have the same one on my GPS V, I can convert the coordinates to accurately show on the map and include the info on the cache page.

 

Worldtraveler

[Moun10Bike]

quote:

---

Originally posted by worldtraveler:

And now to further thicken the plot (if there is one), the map has been removed from the cache page! Jeremy, I sure hope it was you or one of your helpers.

 

---

 

Any time that you edit a cache page, it forces the map to be regenerated (a good idea just in case your editing included changing the coordinates). At these times, caches located within the US revert to the old Tiger-based map for a short time, while caches outside of the US display no map at all. The map should reappear shortly once your cache makes it through the map generation queue.

 

[+worldtraveler]

quote:

---

Originally posted by Moun10Bike:

Any time that you edit a cache page, it forces the map to be regenerated...caches outside of the US display no map at all. The map should reappear shortly once your cache makes it through the map generation queue.

 

---

 

Thanks, Moun10Bike! I learned something today. Since the map disappeared, I was about to go back and re-edit to remove my note about the map discrepancy; I'm glad you explained it before I did that.

 

Now if I could just find out about the datum Mapquest uses for that area...

 

Worldtraveler

[Kerry.]

quote:

---

Originally posted by worldtraveler:

 

And then there's the third possibility... and possible intrigue. The park where I placed the cache happens to be in an area where there is a large concentration of embassies, including that of the U.S. Do you suppose the discrepancies could be due to the imposition of selective availability? If so, this may account for the greater discrepancy (~1 km) near the cache (and embassies) and somewhat lesser discrepancy further away at the hotel.

---

 

I would doubt (very much) that even Selective Deniability (a little different to the old Selective Availability) could be that selective and really would eliminate that possibility altogether.

 

It might be a map issue, it could just be that 0.01% inaccuracy period that the system can/will/could do sometimes? The distance sort of doesn't point to a datum issue especially as there's a difference in offsets within short distances and the shift is a little too much?

 

Could also be local interference of some sort, could even be deliberate interference as that is certainly controllable over short distances.

 

But it does sound more like a map accuracy or similar related issue as just because the maps are presumeably accurate in the U.S doesn't mean they are elsewhere.

 

Cheers, Kerry.

 

I never get lost everybody keeps telling me where to go

[+worldtraveler]

quote:

---

Originally posted by Kerry:

...Could also be local interference of some sort, could even be deliberate interference as that is certainly controllable over short distances.

 

But it does sound more like a map accuracy or similar related issue...

---

Yes, "interference" is a more accurate term than "selective availability" to describe that part of my question. No one's spoken up yet to relate a similar discrepancy between their cache location and the map display, so maybe I'll just have to wait until someone tries to find the cache. I've already left Bangkok and probably won't return until next year, so I can't go back and recheck the coordinates myself.

 

Worldtraveler